Plaxis Advanced Computational Geotechnics-Singapore 2011 (Compiled) PDF

ADVA
A ANCE ED COU
C RSE
E
ON
O CCOMP
PUTAATIO
ONALL GE
EOTE
ECHN
NICS
S
SINGA
S APO
ORE

Ven
nue National University o
of Singaporee
Datte 23 to 25 N
November 2 2011

Asssociate Proffessor Tan SSiew Ann
Natioonal Universiity of Singapore
(Co
ourse leaderr)
Dr Shen Rui Fu
u Natioonal Universiity of Singapore
Dr William Cheeang Plaxiis AsiaPac, SSingapore
Org
ganised by
Plaxis AsiaPac Pte Ltd

L
1 of 607
CONTENTS
Lectures
CG1 Concepts of Plasticity and The Mohr-Coulomb Model 5
CG2 Density and Shear Hardening 26
CG3 Hardening Soil Model 49
CG4 Determination of Soil Parameters 49
CG5 Soft Soil Model 73
CG6 Soft Soil Creep Model 108
CG8 Drained and Undrained Analysis 171
CG9 Structural Elements & Modelling Excavations in Plaxis 196
CG10 Modelling of Groundwater 210
CG11 Slope Stability Analysis and Rain-fall induce failures 225
CG12 Phi-C reduction analysis in Plaxis 267
CG13 Introduction of Plaxis 3D 279
CG14 Geometry, Calculation and Output Mode 295
CG15 Modelling of Piled Foundations in 3D 321
CG16 Modelling of Tunnels and Tunnelling in 3D 367
CG17 Embedded Elements in Plaxis 3D 409
CG18 Modelling of Excavations in 3D 426
Exercises
SGE1 Simulation of laboratory triaxial test 468
SGE2 Simulation of laboratory oedometer test 474
SGE3 2D Excavation : Anchored Excavation 513
SGE4 2D Slope Stability: Stabilisation of an Embankment Using Geotextile 540
SGE5 3D Piled Foundation: Fleiden Piled Raft Analysis 559
SGE6 3D Excavation: 3D Effects in Strutted Retaining Wall Analysis 576
2 of 607
DAY 1
Time Module On Hard and Soft Soils

Concepts of Plasticity and The Mohr-Coulomb
09:15 10:00 CG1 Prof.Tan
Model
10:00 10:45 CG2 Density and Shear Hardening Prof.Tan
10:45 11:00 Tea Break

11:00 11:45 CG3 Hardening Soil Model
William
11:45 01:00 CG4 Determination of Soil Parameters
01:00 02:00 Lunch
02:00 02:45 CG5 Soft Soil Model
Prof.Tan
02:45 03:30 CG6 Soft Soil Creep Model
03:30 03:45 Tea Break
03:45 04:30 E1 Simulation of Laboratory Tests 1 Shen
04:30 05:30 E2 Simulation of Laboratory Tests 2 Shen

DAY 2
Time Module Application of Plaxis 2D

09:00 09:45 CG7 Consolidation Analysis Prof.Tan
09:45 10:30 CG8 Drained and Undrained Analysis Prof.Tan
10:45 11:00 Tea Break
Structural Elements &
11:00 11:45 CG9 Shen
Modelling Excavations in Plaxis
11:45 01:00 E3 Excavation Exercise Shen
01:00 02:00 Lunch
02:00 02:45 CG10 Modelling of Groundwater Gouw
Slope Stability Analysis and Rain-fall induce
02:45 03:30 CG11 Prof.Tan
failures
03:30 03:45 Tea Break
03:45 04:30 CG12 Phi-C reduction analysis in Plaxis William
04:30 05:30 E4 Slope Stability Exercise Gouw

3 of 607
DAY 3
Time Module Application of Plaxis 3D

09:00 09:45 CG13 Introduction of Plaxis 3D William
09:45 10:30 CG14 Geometry, Calculation and Output Mode William
10:45 11:00 Tea Break
11:00 11:45 CG15 Modelling of Piled Foundations in 3D Prof.Tan
11:45 12:30 CG16 Modelling of Tunnels and Tunnelling in 3D William
12:30 01:30 Lunch
01:30 02:15 CG17 Embedded Elements in Plaxis 3D Shen
02:15 03:30 E5 Exercise: Pile Analysis Shen
03:30 03:45 Tea Break
03:45 04:30 CG18 Modelling of Excavations in 3D Prof.Tan
04:30 05:30 E6 Exercise: Multi-strutted Excavation William

4 of 607
CG1 - Concepts of Plasticity and
Mohr-Coulomb Model
Ronald Brinkgreve
Plaxis BV /
Delft University of Technology
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
Contents
• Aspects of real soil behaviour
• Stresses and strains
• Stress paths in standard soil tests
• Standard drained triaxial test (CD-test)
• Oedometer test
• Consolidated undrained triaxial test (CU-test)
• Basic concepts of the Mohr-Coulomb model
• Elastic strains, plastic strains
• Yield function, plastic potential
• Parameters
• Possibilities and limitations of the M-C model
Singapore 2011 5 of 607

Aspects of real soil behaviour
• Elasticity (reversible deformation; limited)
• Plasticity (irreversible deformation)
• Failure (ultimate limit state or critical state)
• Presence and role of pore water
• Undrained behaviour and consolidation
• Stress-dependent stiffness
• Time-dependent behaviour (creep, relaxation)
• Compaction en dilatancy
• Memory of pre-consolidation pressure
• Anisotropy (directional strength and/or stiffness)
Stresses and strains

• Stresses
Cartesian stresses:
  xx  yy  zz  xy  yz  zx T
   '  w
 = total stresses
’ = effectieve stresses
w = pore pressure (isotropic):
 Hydrostatic (constant head)

 Non-hydrostatic (variable head  groundwater flow)
 Excess pore press. (undrained behaviour  consolidation)

• Stresses
Principal stresses:
 1  s * t * s* 1
2
 ' xx  ' yy 
 2  zz
 3  s * r * t* 1
4
 ' xx  ' yy 2   xy2
Stress invariants (p and q):

p   xx   yy   zz    1   2   3 
1 1
3 3
1
q ( xx   yy ) 2  ( yy   zz ) 2  ( zz   xx ) 2  6 xy
2 2
 6 yz 2
 6 zx
2

• Strains
Cartesian strains:   xx  yy  zz  xy  yz  zx 
T
Normal strains Shear strains

 ux  ux  u y
 xx   xy  
x y x
u  u y  uz
 yy  y  yz  
y z y
u u u
 zz  z  zx  z  x
z x z

• Visualisation of stresses:
-1 -1
p-axis p-axis
q-axis
1
Rendulic plane 2
q-axis
-3
2= 3 -32
-2
Principal stress space Rendulic plane

• Visualisation of stresses: -1
-1
p-axis
Deviator plane
-3
-2 -3
-2
Principal stress space Deviator plane (-plane)
(p = constant)

Stress paths in standard soil tests
• Standard drained triaxial test (CD test)
Stress-strain and strain diagram:
1 1-3
v
 1
-1
3 3 v
-1
Loose sand Dense sand


• Standard drained triaxial test (CD test)
Stress paths:
xy
-’1
Axial loading
-3
n
-3 -’3

• Consolidated undrained triaxial test (CU test)
Stress-strain diagram:
1
1-3
v  0
 1
-1
3 3
pw
-1


• Consolidated undrained triaxial test (CU test)
Stress paths:
-’1
Axial loading
-3
-3 -’3

• Oedometer loading test 
1
1
-1
 1
ln 1
-1

• Oedometer loading test
Stress paths:
-’1
Axial loading
-’3

• Simple shear test
xy
dxy
 dxy
xy
v
xy

• Simple shear test
Stress paths:
-’1
Shearing
-’3

Basic concepts of the M-C model
• Division of strains and strain increments:
  e   p (strains)
d  d e  d p (strain increments)
Strains (or increments) are divided into elastic strains

and plastic strains
A soil model relates increments of stress to

increments of strain

• Elastic strain increments: From Hooke’s law:
 d xx
e 
 1   0 0 0   d ' xx 
    
    
 d e    1  0 0 0   d ' yy 
 yy    
    
 e     0   d ' zz 
 d zz  1 
1 0 0
 
     
 e  E   
d xy   0 0 0 2  2 0 0   d ' xy 
    
    
e
d yz   0 0 0 0 2  2 0   d ' yz 
    
    
 d e   0 0 0 0 0 2  2   d ' zx 
 zx 

• Plastic strain increments:
g
From flow rule: d  p  d
 '
d = magnitude of plastic strains (multiplier)
dg/d’ = direction of plastic strains (vector)
g = plastic potential (function)
Classical plasticity: g = f (associated plasticity)

For soils in general: g  f (non-associated plasticity)
f = yield function

• When do plastic strains occur?
Determination on the basis of a yield function f =

f(’,)
• If f < 0 Pure elastic behaviour

• If f = 0 and df < 0 Unloading from a plastic
state
(= elastic behaviour)
• If f = 0 and df = 0 Elasto-plastic behaviour
• f>0 Non-acceptable stress state

• Yield function:
Can be represented as a contour in (principal) stress

space
1
f > 0 Not acceptable
f = 0 Plasticity
f < 0 Elasticity
3
2

• Mohr-Coulomb yield criterion:
’yy

xy c’ cos’
-s* sin’
’ t*
xy
’ c’
’xx -3 -1 -

 -s*
The condition   c’ - ’ tan’ must hold for arbitrary angles 

Equivalent expression: t*  c’ cos - s* sin’

• Mohr-Coulomb yield criterion and yield function:
t*  c’ cos’ - s* sin’
t* = ½(’3 - ’1) t * 4

1  '  '
xx yy2 2
  xy
s* = ½(’3+’1) s* 2

1  '  '
xx yy 
1
2  '3  '1   c' cos  '  12  '3  '1 sin  '
f  1
2
 '3  '1   12  '3  '1 sin  ' c' cos  '
Note: Compression is negative and ’1 ’2 ’3

• Mohr-Coulomb yield function:
f  1
2
 '3  '1   12  '3  '1 sin  ' c' cos  '
• Mohr-Coulomb plastic potential:

g  12  '3  '1   1
2
 '3  '1 sin  ' c' cos '

-1
• Yield directions in
deviator plane:
 = 30°
 = 0°
-2 -3

• Mohr-Coulomb parameters:
E Young’s modulus [kN/m2]

 Poisson’s ratio [-]
c’ (effective) cohesion [kN/m2]
’ (effective) friction angle [º]
 Dilatancy angle [º]

• Elasticity parameters: E , 
- d1 
d3 
- 1
d1
E
d1 E
d3 1
 - 1
d1 
1
3

dxy
• Alternative elastic parameters:
 dxy
Shear modulus:
d xy E
G 
d xy 21   
Bulk modulus: dp
dp E dv
K  
d v 31  2 
Oedometer modulus: - d1

 - d1
d1 E 1   
Eoed  
d1 1   1  2 

• Plasticity parameters: c’, ’
-’1
xy
’ b
1
c’ a
’n -’3
2c' cos  ' 1  sin  '

a b
1  sin  ' 1  sin  '

• Plasticity parameter: 
xyxy xy
xy xy
yy
xy
xy
p
d yy yy
 tan 
p
d xy 
xy
dilatancy 

• Plasticity parameter: 
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  '  p
d yy
d xyp  d  d  xy   d cos
 ' xy  t*  p
 tan 
d xy

• Meaning of M-C parameters in drained triaxial
test:
1-3 E’ ’3 = confining pressure
2c' cos  '2 '3 sin  '

1  sin  '
-1
v
2 sin 
1  sin 
-1
1-2’

• Meaning of M-C parameters in oedometer test:
-1 1
1
1- Eoed

-1
-3
Possibilities and limitations of M-

C
• Possibilities and advantages:
• Simple and clear model (elastic perfectly-plastic

model)
• First order approach of soil behaviour in general
• Suitable for many practical applications
• Limited number and clear parameters
• Good representation of failure behaviour (drained)
• Dilatancy can be included

Possibilities and limitations of M-
C
• Limitations and disadvantages:
• Isotropic and homogeneous behaviour

• Linear elastic behaviour until failure
• No stress-dependent stiffness
• No distinction between primary loading and
unloading or reloading
• Dilatancy continues for ever (no critical void)
• Undrained behaviour not always realistic
• No anisotropy, no time-dependency (creep)
NON-COHESIVE GRANULAR
MATERIAL
r
shear stress
reality
f=0 f ≈ r – As0.9
s
mean stress r
Models
f = r –As -B
A = sinφ
B = c cosφ s

WARNING FOR DENSE
SANDS
σ y
σx
εy
• Do not rely on high friction angles

• Post-peak softening
• Be conservative and use φ ≤ 35º
COHESIVE GRANULAR
MATERIAL : CLAY
We • wp < w < wl r
consider : • w = water
f=0
content
• wp = plastic
limit tension cut-off
• wl = liquid s
limit
test data : f ≈ r - As - B
• cohesion = (B/cosφ) depends on water content
• effective cohesion very small for w ≈ wl
• effective cohesion significant for w ≈ wp

ROCK AND CONCRETE
test data : f ≈ r - As0.9 - B

• cohesion = (B/cosφ) tends to be very large
• tensile strength tends to be small; tension cut-off
r
is needed
r
r = As0.9 + B Mohr-Coulomb
approximation
tension cut-off
s s
Thank you for your attention
Questions ?

Universität Stuttgart
CG2-Density and Shear Hardening
5th Asian Advanced Course 24-26 Nov 2011

By: Prof Harry - NUS
Lecture by Prof. Pieter A. Vermeer

University of Stuttgart, Germany
Shear hardening and density hardening

as used in HS-model
1. Introduction
2. Hyperbolic stress-strain curve
3. Shear strain contours
4. Unloading – reloading
5. Density hardening
6. Double hardening
Kuala Lumpur, 2009
Singapore 2011
26 of 607
Introduction
Stress path in oedometer loading - unloading
Real test & HS model Mohr-Coulomb model
Introduction
Settlements for strip footing
distance from centre line [m]

0 3 6 9 12 15
-0,2
0,2
s / smax [-]
0,4
0,6 Linear Elastic

Mohr Coulomb
0,8 Hardening Soil
1
Singapore 2011
27 of 607
Introduction
Example for vertical displacements behind a retaining wall
Typical vertical displacements behind a retaining wall

(sheet pile wall in clay)
120
Mohr Coulomb
100
vertical displacements [mm]
80
Hardening Soil
60
40
20
0
0 5 10 15 20
-20
-40
distance from wall [m]
Part 2 Hyperbolic stress-strain curve
Data from triaxial tests will be considered
Triaxial samples
The traditional triaxial sample is relatively long sample (H/D = 2) without lubricated endplates.
A high quality research sample is relatively short (H/D = 1) and it is tested with lubricated end plates.
Singapore 2011
28 of 607
Stress-strain curve from a standard drained triaxial test
1 σ 1
3 = constant
3
1
1 isotropic loading
σ 3
q  σ1  σ 3
dense soil
q f = failure value
loose
qf q cs
q cs = critical state value
1
due to isotropic loading

Hyperbolic approximation
q  σ 1  σ 3
Asymptote
q a = qasymptotive
Hyperbola
1
end of isotropic compression is taken as
undeformed reference state
qa q
ε1  
2  E 50 q a  q
Kondner & Zelasko (1963); “A hyperbolic stress-strain formulation for sands”,

Proceedings of the Second Panamerican Conference on Soil Mechanics and
Foundation Engineering
2  sin  a ,
qa  σ 3  a    a   asymptotic

1  sin  a
Singapore 2011
29 of 607
Definition of secant modulus E50
q  σ 1  σ 3 E50
1
50%
Hyperbola
50%
1
m
 σ 3  a 
E 50  E ref 
50  

 p ref  a 
E ref
50
= reference modulus for primary loading at 50% of strength
msand  0.5 ; mclay  1 qa q

ε1  
2  E 50 qa  q

Modification after Duncan und Chang (1970) Universität Stuttgart
q  σ 1  σ 3
Asymptote
qf
Hyperbola qf qa  R f  0.9
Rf
1
hyperbolic for q < qf otherwise q = qf
qf 2  sin 
qa  with q f  3  a   a  c   cot 
Rf 1  sin 
Duncan & Chang (1970) Non-linear analysis of stress and strain in soils. Journal of the Soil Mechanics
and Foundation Division (ASCE), Vol. 96 No. SM5.
Singapore 2011
30 of 607
Parameters for hyperbolic law
effective cohesion : c
exponent for stress  dependency of E50 : m
secant modulus for  3  pref : E ref

50
friction angle : 
sin  
sin  a  wi th R f  0.9
sin    R f 1  sin   

Data from drained tests on OC Frankfurt clay
300 kPa 3
3  300 kPa c   10kPa
1  3
   19 
200
 3 ’ = 200 E ref
50  15MPa
m  0.7
 3 ’ = 100
100
m
measured  σ 3  c cot  a 
hyperbolic E 50  E ref 
50  

 p ref  c cot  a 
0 ε 1 [%]
0 1 2 3 4
Data from: Amann, P.; Breth, H. & Stroh, D. (1975), Verformungsverhalten des Baugrundes beim
Baugrubenaushub und anschließendem Hochhausbau am Beispiel des Frankfurter Tons, Report 15 of
Geotechnical Institute of the University of Darmstadt.
Singapore 2011
31 of 607
Part 3
Shear strain contours
Part 3 Shear strain contours
Hyperbolic approximation of curve from standard drained test
q  σ 1  σ 3
Asymptote
qa
Hyperbola
qa q
1  
2  E 50 q a  q
3
γ  shear strain  ε 1  ε 3  ε1
2
3 qa q
γ  
4 E 50 q a  q
2  sin  a
q a  q asymptotic  a   asymptotic
 qa  σ 3  a  
1  sin  a
Singapore 2011
32 of 607
Shear strain contours in p-q-plane for c  0 Universität Stuttgart
  
q q
  0 .05
  0 .01
p´ p´
sands : m  0.5 clays : m  1

curved lines straight lines
m
3 qa q  σ 3  a  2 sinφ a
  E 50  E ref
  qa  (σ 3  a)
4E 50 q  qa 50 1  sin φ a
 p ref  a 
Data for Fuji river sand after Ishihara et al. (1975)
1
3
p´
´
R  σ 1 / σ 3 . An oedometer test is a R  constant test with R  1/ K 0 .

Ref. : Ishihara, Tatsuoka and Yasuda (1975). “Undrained deformation and liquefaction of sand under
cyclic stresses“. Soils and Foundations, Vol. 15, No. 1.
Singapore 2011
33 of 607
Measured contour lines versus hyperbolic law Universität Stuttgart
3
  2  3  1
2
´ ´
Calculated shear strain contours are shown by dotted lines. They are calculated by using,
m
3 qa q  σ 3  a  2 sinφ a
  E 50  E ref   qa  (σ 3  a)
4E 50 q  qa 50 
 p ref  a  1  sin φ a
a  0, φ a  38, 50  30 MPa,
Eref m  0.5
Shear strain contours are yield loci Universität Stuttgart
p'
´
measured shear contours measured yield loci
Ref. : Tatsuoka & Ishihara (1974). “Yielding of sand in triaxial compression“. Soils and Foundations, Vol. 14, No. 2.
Singapore 2011
34 of 607
Loading and unloading / reloading Universität Stuttgart
q Loading from plastic state

Stress point on yield locus. Stress increment is
directed out of the elastic region. This causes
plastic yielding, i.e. plastic straining and an
elastic region expansion of the elastic region. The material is
p´ said to harden.
q Unloading from plastic state

failure line
Stress point on yield locus. Stress increments
are directed into elastic region. These stress
increments produce purely elastic increments of
strain. Elastic strain increments are related to
p´ stress increments by Hooke’s law using Eur ,i.e.
unloading-reloading modulus.
q Unloading/reloading from elastic state

current yield locus Stress below yield locus. All possible stress
strain.
p´
Part 4
Unloading - Reloading
Singapore 2011
35 of 607
Part 4 Unloading - Reloading
Unloading – reloading in standard drained triaxial tests Universität Stuttgart
qf failure line
E50
deviatoric stress |1 - 3|
m
 σ 3  a 
E ur  E ref 
ur  

 p ref  a 
1
axial strain

Unloading – reloading in standard drained tests on sand Universität Stuttgart
loose : E ur  ( 3 - 5)  E 50 dense : E ur  ( 2 - 3)  E 50
Default in Plaxis code: E ur  3  E 50
Singapore 2011
36 of 607
Hooke’ law of isotropic elasticity Universität Stuttgart
1
Δε 1e   Δ σ 1  ν ur  Δ σ 2  ν ur  Δ σ 3 
E ur
1
Δε e2   ν ur  Δ σ 1  Δ σ 2  ν ur  Δ σ 3 
E ur
1
Δε 3e   ν ur  Δ σ 1  ν ur  Δ σ 2  Δ σ 3 
E ur
m
  a 
ν ur  Poisson' s ratio  0 .2 E ur  E urref  3  a  c´cot  ´
 p ref  a 
 
Part 5
Density hardening
Singapore 2011
37 of 607
Part 5 Density hardening
Typical results of triaxial tests Universität Stuttgart
qf qr = qcrit
1
n n loose
loose
ncritical
dense
dense
clay sand
1 1
n = porosity

Measured contour lines of volumetric strain for a NC-Clay Universität Stuttgart
Henkel (1960) : The shear strength of saturated remoulded clays. Research Conf. on Shear Strength of
Cohesive Soils, Boulder, Colorado (ASCE).
Singapore 2011
38 of 607
Measured contour lines for a kaolin clay Universität Stuttgart
q [MN/m²]
 vol  0.25
0.14 vol = 0.20

0.07
p [MN/m²]
Biarez, J. & Hicher, P.-Y. (1994), Elementary Mechanics of Soil Behaviour, Balkema - Publishers.
Die Proben bestanden aus einem gesättigten Kaolin-Ton und wurde allseitig bis auf unterschiedliche Werte von p’ konsolidiert.
Danach wurden sie bis zum Bruch unter undrainierten Bedingungen einaxial belastet.

Contour lines resemble ellipses Universität Stuttgart
q [MN/m²]
 vol  0.25
 vol  0.20
0.14
0.07
p’ [MN/m²]
Biarez, J. & Hicher, P.-Y. (1994), Elementary Mechanics of Soil Behaviour, Balkema - Publishers.
Singapore 2011
39 of 607
Ellipse as volumetric strain contour Universität Stuttgart
q
1
q2
M p   pP
M 2 p
peak point
6  sin  '
with : M 
pp p 3  sin  '
1
pp
2
Ellipse is used in Modified Cam Clay model.

Measured contour lines for loose Fuji River sand Universität Stuttgart
general formula for m  1:

1 m
p  p 
1  
ε vol   ref  p 
1  m K ref  p ref 
 
pP special case for m = 1:


p ref pp
ε vol  ε 'ref  ln
K ref p ref
Ellipse:
q2
p p  p 
Kref = reference bulk modulus M 2  p
Ishihara, Tatsuoka and Yasuda (1975). “Undrained deformation and liquefaction of sand under cyclic stresses“,
J. Soils and Foundations, Vol. 15, No. 1.
Singapore 2011
40 of 607
Loading and unloading / reloading Universität Stuttgart
q Loading from plastic state

Stress point on yield locus. Stress increment is
directed out of the elastic region. This causes
plastic yielding, i.e. plastic straining and an
elastic region
expansion of the elastic region. The material is
 p´ said to harden: expansion of plastic region.
pp
q Unloading from plastic state

Stress point on yield locus. Stress increments
are directed into elastic region. These stress
strain. Elastic strain increments are usually
 p´ related to stress increments by Hooke’s law
pp
using Eur ,i.e. unloading-reloading modulus.
q Unloading/reloading from elastic state

current yield locus
Stress below yield locus. All possible stress
strain.
 p´
pp

Volumetric strain or density hardening Universität Stuttgart
q 6sin  cs q
M 6sin 
3  sin  cs
1 3  sin 
1
β•pp 
elastic elastic
 
pp p´ pp p´
Modified Cam Clay Hardening Soil Model
NC
K 0 is controlled by β. On inputting realistic
no control of K NC value of one needs
K NC relatively large β-
0 0
values and one gets a relatively steep yield
cap
Volumetric-strain hardening is dominant for NC-clays and very loose sands.

Shear-strain hardening, as considered in another lecture, is dominant for OC-clays and dense
sands.
Singapore 2011
41 of 607
Part 6
Double hardening
Part 6 Double hardening
Shear-strain hardening & volumetric-strain hardening Universität Stuttgart
q
q
elastic elastic
shear-strain hardening p´ volumetric-strain hardening p´
q HS-model
double hardening
pp p´
Shear-strain hardening is dominant for OC-clays and sands.

Volumetric-strain hardening is dominant for NC-clays.
Singapore 2011
42 of 607
Four stiffness zones
q
E50 & Eoed
combined hardening
E50
Eoed
shear hardening
cap hardening
Eur
elastic
pp
p
Appendix
Singapore 2011
43 of 607
Density Hardening of Soils Part 3: Contour lines in p-q-plane
The expression for vol is relatively simple
q general formula for m  1:

M  p 
1 m
1 p  
1 ε vol   ref  p 
1  m K ref  p ref 
 
special case for m = 1:

p ref pp
ε vol  ε 'ref  ln
K ref p ref
pp p
Ellipse:
Kref = reference bulk modulus q2
p p  p 
M 2  p

Pp is computed from p Universität Stuttgart
OCR POP
p p
 
Singapore 2011
44 of 607
ref
E 50
can be estimated from E ref
oed
50  E oed
E ref ref
50  2  E oed
E ref ref
NC-Clays:

ref Universität Stuttgart
E oed
is a tangent stiffness modulus
E ref
sec
m m
  1  a    3v  a 
E oed  E ref   E ur  E ref  
oed  p a ur  p a
 ref   ref 
Singapore 2011
45 of 607
HS-model does not include hysteresis Universität Stuttgart
q = 1 - 3
1
The unloading-reloading modulus Eur is the secant stiffness in large unloading-reloading loops. The modulus
E0 at small strain is a manifold of Eur.

Unloading – reloading in standard drained tests on sand Universität Stuttgart
σ 3v  100 kPa
e 0  0.54
Data for ε 1  3%
: E ur  (1 .9 - 3.3)  E 50
Singapore 2011
46 of 607
Complex reloading provides new yield point Universität Stuttgart
q q
new yield point new yielding

last yielding
p'
1
The last yield point and the new yield point are on a so-called yield locus. Soil
behaves essentially elastic as long as the stress remains below the yield
surface.
Yield loci of sand after Tatsuoka & Ishihara (1974) Universität Stuttgart
p'
Many samples of Fuji River sand were tested to obtain pairs of yield points and thus segments of yield loci by
Tatsuoka & Ishihara (1974): “Yielding of sand in triaxial comprtession”, Soils and Foundations 14 (2)
Singapore 2011
47 of 607
Data for a loose sand after Ishihara et al. (1975)
p ( kg / cm 2 )
1 kg / cm 2  100 kPa
Compression states : axial stress is major compressive stress

Extension states : axial stress is minor compressive stress
Shear strain contours in 3D stress space


1

3
9  I 1a I a2 / I a3
 sin 2 m 
2 1  I 1a I a2 / I a3
remember : I 1a  1a   a2   a3 I a2  1a  a2   a2  a3   a3 1a I a3  1a  a2  a3

 a
1  1  a ,  a
2   2  a ,  a3   3  a
Singapore 2011
48 of 607
Hardening Soil & Hardening Soil with
Small Strain Overlay
William Cheang
Notes partly by: Dr Ronald Brinkgreve
Plaxis b.v. and TU Delft
Hardening Soil model
Characteristics:
1. Stress-dependent stiffness behaviour according to a power law

2. Hyperbolic stress-strain relationship in axial compression
3. Plastic strain by mobilising friction (shear hardening)
4. Plastic strain by primary compression (compaction hardening)
5. Elastic unloading / reloading
6. Failure behaviour according to the Mohr-Coulomb criterion
7. Small-strain stiffness (HS-small model only)
49 of 607
Hyperbolic stress-strain relationship in (tri)axial loading:
(Duncan-Chang model)
q
 1 qult
q
1 / E0  1 / qult E0
Rf qult
q / E0 Eur
 1 
1  q / qult
E0 = initial stiffness 1
qult = asymptotic value of q (related to strength)
Rf = ‘failure ratio’ (standard value 0.9)

Hyperbolic stress-strain relationship in (tri)axial loading:
2c cos   2 '3 sin 

R f qult 
1  sin 
m
 '  pref = 100 kPa (1 bar)
E0  E  ref3
ref
0 
p 
m
 '  Unloading / reloading
Eur  Eurref  ref3 
p 
50 of 607
Shear hardening in the HS model
Elastoplastic formulation of hyperbolic q-1 relationship:
Yield function: f fric  f *   p

(non-associated)
 
1 q 2q
f*    p
  2 1p   vp
E50 1  q qa Eur
m m
 c cot  ' ' 3   c cot  ' ' 3 
E 50  E ref
  Eur  E ref
 
 c cot  ' p  c cot  ' p
50 ref ur ref
 
f f  12 ( 3   1 )  12 ( 3   1 ) sin  'c cos  ' (MC failure)

Yield function: f fric  f *   p

(non-associated)
 
1 q 2q
f*    p
  2 1p   vp
E50 1  q qa Eur
m m
 c cos  ' '3 sin  '   c cos  ' '3 sin  ' 
E50  E ref
  Eur  E ref
 
 c cos  ' p sin  '   c cos  ' p sin  ' 
50 ref ur ref
f f  12 ( 3   1 )  12 ( 3   1 ) sin  'c cos  ' (MC failure)
51 of 607
Elastic
q MC failure line q MC failure line

plastic
m 3p,fric
2p,fric
1p,fric
p’ 

Flow rule: d vp , fric  d p , fric
sin  m with:
sin  m  sin  cv
sin  m 
q MC failure line 1  sin  m sin  cv

sin  ' sin 
m 3p,fric sin  cv 
1  sin  ' sin 
2p,fric
 '1  '3
sin  m 
1 p,fric  '1  '3 2c cot  '
p’
52 of 607
Flow rule: d vp , fric  d p , fric
sin  m with:
sin  m  sin  cv
sin  m 
q MC failure line 1  sin  m sin  cv

m>0 sin  ' sin 
m
cv sin  cv 
1  sin  ' sin 
m<0 (in principle)
Note:
m < 0 is not taken
into account
p’
Compaction hardening
q
in the HS model
MC failure line
Yield function (associated):
q2
f cap   p 2  pc2
2
Cap
Hardening rule:  pc
1 m
  pc  fc = 0
 vp ,cap   
1  m  p ref  pc p’
 is determined by K0 nc c 1
 is determined by Eoed
m
ref  c cot  ' '1 
Eoed  Eoed  ref 
 c cot  ' p  v
53 of 607
Compaction and Shear hardening
in the HS model
Cap
Cone
Compaction and Shear hardening

in the HS model
Relevance of Compaction hardening:

• Plastic compaction in primary loading
• Distinction between primary loading and unloading/reloading
Relevance of Shear Hardening:

• Decreasing stiffness (increasing plastic shear strains) in deviatoric
stress paths (principal stress differences, shearing)
54 of 607
Small-strain stiffness in the HS model
(HSsmall)
Strain(path)-dependent elastic overlay model:
G0
Gs 
1  0.385  /  0.7
G0
1  0.385  /  0.7 2
Gt   Gur
G starts again at G0
Gur after full strain reversal

(HSsmall)
Gt Cyclic loading
G0 leads to Hysteresis
Gs
-c  Energy dissipation

  Damping
+c
G0
G0
55 of 607
(HSsmall)
(Excel sheet)
G0 0.7
Gt Gs
Gur

(HSsmall)
Relevance of small-strain stiffness:

• Very stiff behaviour at very small strains (vibrations)
• Reduction of stiffness with increasing strain; restart after load reversal
• Hysteresis in cyclic loading:
• Energy dissipation
• Damping
Also relevant for applications like:
• Excavations (settlement trough behind retaining wall)
• Tunnels (settlement trough above tunnel)
56 of 607
Parameters of the HS(small) model
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 only)
0.7 Shear strain at which G has reduced to 70% (HSsmall only)
m Rate of stress dependency in stiffness behaviour
pref Reference pressure (100 kPa)
ur Poisson’s ratio in unloading / reloading
c’ Cohesion
’ 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
Parameters of the HS model

Parameters:
q
3=pref c 1=pref 1
qult
(, c)
E50ref qf=Rf qult
Eurref
0.5 qf 1
Eoedref
1 v
Triaxial test Oedometer test
57 of 607
Eoed [MPa] for NC-soils and ´ = 100 kPa
105
rock After Janbu (1963)
104
Janbu :
m
103
  
Eoed  Eref 
oed  

sandy gravel  pref 
102
sand
10
more general:
m
Norwegian    a 
Eoed  E ref
  
 pref  a 
clays oed
1
Mexico City Clay with a = c´ cot´

0.1
0 50 100
porosity n [%]
For normally consolidated clays (m=1):
ref
Eoed  1
2
ref
E50 Order of magnitude (very rough)
50000 kPa
ref
Eoed  Correlation with Ip for pref=100 kPa
Ip
500 kPa
ref
Eoed  Correlation by Vermeer
wL  0.1
ref
Eoed  p ref * Relationship with Soft Soil model
58 of 607
For sands (m0.5):
ref
Eoed  E50
ref
Order of magnitude by Schanz
Correlation by Lengkeek
ref
Eoed  RD  60 MPa for pref=100 kPa
For sands (m0.5):
Schanz (1998)
59 of 607
Eur , G0 and 0.7
Eurref  (3 to 5) E50
ref
Eurref
G ref
 (2.5 to10)G ref
where ref
G 
2(1   ur )
0 ur ur
 0.7  (1 to 2) 10 4
Vucevic & Dobry, 1991
60 of 607
Initial conditions for the HS model
Initial pre-consolidation stress pc based on c:
’yy0 c ’yy0 c
Over-Consolidation Ratio: Pre-Overburden Pressure:

OCR = c /’yy0 POP = c -’yy0
Initial stresses:
’yy
’yy0 follows from soil weight
Prestress and pore pressure
’c
Initial
CAP ’xx0 = K0 ’yy0
POP 1
 ur
’yy0 1 ur
Initial stress

K 0nc  '0yy  POP    ur
1  ur
POP
1 K0 
K0nc
 '0yy
 ur
’xx0 ’xx K 0  OCR K 0nc  OCR  1
1  ur
61 of 607
Initial stresses:
q
MC failure line
K0nc line
Output:
 pc Cap
pc
p’0, q0 ' OCR '  OCRiso 
p eq
peq0 pc,0 p p eq   p' 

2
 q2 / 2
Comparison HS model and MC model

Stress-strain development in different stress paths:
Hardening-Soil model: Mohr-Coulomb model:

E50ref 25000 kPa
Eoedref 25000 kPa E 25000 kPa
Eurref 75000 kPa  0.30
pref 100 kPa c’ 0.1 kPa
m 0.5 ’ 35°
ur 0.2  5°
c’ 0.1 kPa
’ 35°
 5°
Rf 0.9
K0nc 0.426
62 of 607
Isotropic compression test:
Custom
1000 MC
HS.vlt
900
800
700
p' [kN/m²]
600
500
400
300
200
100
0
0 0.01 0.02 0.03 0.04 0.05 0.06
v

Drained triaxial test at 3=100 kPa :
E1DS
MC
HS.vlt
200
| 1 -  | [kN/m²]
3
100
0
0 -0.01 -0.02 -0.03 -0.04 -0.05 -0.06 -0.07 -0.08
1
63 of 607
Drained triaxial test at 3=100 kPa :
Custom
0.009 MC
HS.vlt
0.006
0.003
v

-0.003
0 -0.01 -0.02 -0.03 -0.04 -0.05 -0.06 -0.07 -0.08

1

Undrained triaxial test at 3=100 kPa :
E1DS
500
MC
HS.vlt
400
| 1 -  | [kN/m²]
300
3
200
100
0
0 -0.01 -0.02 -0.03 -0.04 -0.05 -0.06 -0.07 -0.08
1
64 of 607
Drained / undrained triaxial test at 3=100 kPa :
E1DS
500
MC(u)
HS(u).vlt
MC.vlt
400 HS.vlt
| 1 -  | [kN/m²]
300
3
200
100
0
0 -0.01 -0.02 -0.03 -0.04 -0.05 -0.06 -0.07 -0.08
1

Drained / undrained triaxial test at 3=100 kPa :
PQ
500
MC(u)
HS(u).vlt
MC.vlt
400 HS.vlt
300
q [kN/m²]
200
100
0
0 -100 -200 -300
p' [kN/m²]
65 of 607
One-dimensional compression test (oedometer):
Custom
1000 MC
HS.vlt
900
800
700
' 1 [kN/m²]
600
500
400
300
200
100
0
0 0.005 0.01 0.015 0.02 0.025 0.03
1

S3S1
-1100
One-dimensional
MC
HS.vlt
-1000
compression test -900
(oedometer):
-800
-700
' 1 [kN/m²]
-600
-500
-400
-300
-200
-100
0
0 -200 -400
' 3 [kN/m²]
66 of 607
One-dimensional compression test (oedometer):
Stress state after unloading
HS MC
Possibilities and advantages compared to Mohr-Coulomb:
• Better non-linear formulation of soil behaviour in general (both soft

soils and harder types of soil)
• Distinction between primary loading and unloading / reloading
• Memory of preconsolidation stress
• Different stiffnesses for different stress paths based on standard tests
• Well suited for unloading situations with simultaneous deviatoric
loading (excavations)
• Large stiffness at small strain levels (vibrations) (HSsmall only)
67 of 607
Limitations and disadvantages:
• No peak strength and softening (immediate residual strength)

• No secondary compression (Creep)
• No anisotropy
• E50 / Eoed > 2 difficult to input
Which model in which situation?

Soft soil (NC-clay, Hard soils (OC-
peat) clay, sand, gravel)
Primary load. Soft Soil (Crp), HS / HSsmall
(surcharge) HS / HSsmall
Unloading + HS / HSsmall HS / HSsmall
deviatoric load
(excavation)
Deviatoric Soft Soil (Crp), HS / HSsmall
loading HS / HSsmall
Secundary Soft Soil Creep n/a
compression
68 of 607
Examples of parameter selection
ESTIMATING INPUT PARAMETERS, HS MODEL

Triaxial test results, Shaoli (2004)
Dense Hokksund sand at 40 kPa, n = 35.9% (initial) – 39.6% (end of test)
dense 40
D e v ia t o r ic s t r e s s , q [ k P a ]
200
150
dense 40
100
50
pref  a
E50ref  E50
0  'x  a
0 1 2 3 4 5
100kPa
Axial strain [%]  20000kPa  32MPa
40kPa

Triaxial test results, Shaoli (2004)
Dense 40
Axial strain [%]

-4
1  sin 5
Volumetric strain, [%]
-3 1-sin    1.2
-2
2sin  Dense 40 2 sin 4.2
sin  0.29
-1
0 1 2 3 4 5
0
1   17
69 of 607

Oedometer test dense Hokksund sand, n = 39% , (Moen, 1975)
Test data
Loading:
 1 ' a
Eoed  Eoed
0 ref
-0,2
Test data pa ' a
Vertical strain [%]
-0,4
-0,6
pa ' a
-0,8 ref
Eoed  Eoed
-1  1 ' a
-1,2
850kPa 100kPa
-1,4   53MPa
0 200 400 600 800 1000 1200 0.008 400kPa
Vertical effective stress [kPa]
 3 ' a pa ' a 850kPa 100

Unloading: Eur oed  Eur  Eurref Eurref  Eur oed   215MPa
pa ' a  3 ' a 0.0028 200
HS Material parameters for dense Hokksund sand from

fitting PLAXIS results to experimental data:
E50ref = 35 MPa (estimated 32 MPa)

Eoedref = 45 MPa (estimated 53 MPa)
=0 Eurref = 180 MPa (estimated 215 MPa)
pw = 0 m = 0.6
c = 1 kPa
= 430
 = 180
K0NC = 0.4
 ur = 0.2
symmetry
Axial
Triaxial tests by Shaoli (2004)
70 of 607
Triaxial test results and PLAXIS simulation,

200
180
Deviatoric stress, q [kPa]
160
140
120
Plaxis 40
100 dense 40
80
60
40
20
0
0,00 1,00 2,00 3,00 4,00 5,00
Axial strain [%]
Triaxial test results and PLAXIS simulation,

Axial strain [%]

-4
-3,5
Volumetric strain, [%]
-3
-2,5
-2 from PLAXIS 40
Dense 40
-1,5
-1
-0,50,00 1,00 2,00 3,00 4,00 5,00
0
0,5
1
71 of 607
Oedometer test and PLAXIS simulation dense Hokksund sand, n = 39% ,

(Tore Ingar Moen, 1975)
0
Test data
-0,2
Vertical strain [%]
-0,4 Plaxis
-0,6
-0,8
-1
-1,2
-1,4
0 200 400 600 800 1000 1200
Vertical effective stress [kPa]
72 of 607
CG5 Soft Soil Model with creep
- 1D model for oedometer states

- general 3D model
- validation on triaxial tests
- application on the Leaning Tower of
Pisa
Lecture CG20 by: Prof. Pieter A. Vermeer & Heiko P. Neher

University of Stuttgart, Germany
Idealised curves from oedometer tests on soft soil

void ratio e
1
Cs swelling
index
NC-Line = normal consolidation line
compression index Cc
1
log
ln 
A
1
1
Cc
Cs 2.3  1 e0 
strain ε
A
2.3  1 e0 

OCR = OverConsolidation Ratio = σp/σ0´
e
0  existing stress
p  preconsolidation stress
0
p
NC-Line
log
Many soft soil layers show σp - values that cannot be explained by

preloading.
Bjerrum (1967): increase of OCR due to creep
NC-Linie
creep
Cs
1
0 p
log
Secondary Compression: Creep is better wording

!
Bjerrum (1967): 7th Rankine Lecture, Géotechnique, Vol. 17, pp. 81-118.

1D-Consolidation and 1D- Creep
p p
t
0
e Consolidation Creep or secondary compression
ec
1
Cα = secondary compression index
log t
tc
tc = end of consolidation
Early literature on secondary compression
Buisman (1936) used ε instead of e:
t + t
 = c + CB log = c + CB log tc c = consolidation strain
tc tc tc = end of consolidation
Bjerrum (1967) and Garlanger

(1972):
 + t
e  ec - C log with C  (1+ e0 )  CB

  extra parameter, e.g. one day
Buisman (1936): Results of long duration settlement tests, Proceedings 1st International Conference on
Soil Mechanics and Foundation Engineering, Vol. 1, pp. 103-107
Garlanger (1972): The consolidation of soils exhibiting creep under constant effective stress,
Géotechnique 22 (1), pp. 71-78

24-hours load-stepping is common
e e
ee   Cs   log 
Δe
ec   Cc   log 
24 hours
(NCL)
1 week
log t log 
tc  30´ τ = 24 h
e  eelastic  ecreep  ee  ec
NC-line is usually a 24 hours line. This implies τ = 24

hours
Constant Rate of Deformation Tests by Marques (1996)

σ´ (kPa)
Strain rate
1.7 x 10-5 s-1

Fast loading combined with fast
2.0 x 10-6 s-1
drainage gives states beyond the
NC-line (see also Sällfors, 1975)
ε (%)
NCL
 e
N.B. 
1  e0
Marques (1996): Influencia da velocidade de deformacao e da temperatura no adensamento de argilas naturais

(in Portuguese), M. Sc. Thesis, Université Laval, Ste-Foy, Canada
Sällfors (1975): Preconsolidation pressure of soft high plastic clays, Ph. D. Thesis, Chalmers University of
Technology, Gothenburg, Sweden

e 24-hours e NCL
creep
unloading
p p 0
0
log log
usual case : 0  p unusual case : p  0

by creep or unloading fast loading + fast drainage
1D constitutive model that matches existing data

In differential form:
c c c s
de  c s    c  1    c
e   e e  e c  
dt ln10   ln10   p 
Creep implies a decrease of e and thus a continuous increase of σp. The

increasing
 log p   e /(Cc  Cs )
c
σp must be continuously updated by using
Cc  Cs
Typical soil data: Cs  Cc / 10 and C  Cc / 30   27
C
27
   1
It follows that the creep rate is proportional to   
  OCR27
 p
It follows that the creep rate is negligibly small for OCR >
1.4

The influence of the overconsolidation ratio on the creep rate
Typical ratios : Cs = Cc/10 C = Cc/30
Cc Cs
27
 C  ´  C
 0.005  ´   0.005 1
creep rate: e  c      
 ln10   day   day OCR27
 p  p
e
OCR ec/day (%) ec/year
1 0.5 -
1.1 0.04 0.146
1.2 0.0035 0.013
1.3 0.0004 0.0015
NC-line
1.4 0.00006 0.0002 with p  
1.5 0.000009 0.00003
log
Part 2: From oedometer states to general stress states
oedometer general
pe  p  q2 / M2p  equivalent stress

  oedometer stress
ppe  equivalent preconsolidation stress

p  preconsolidation stress
Cc
 logp   ec / Cc  Cs   lnppe   ec /    
ln10
27
 C      pe 
27
C
e c    e 
c   
 ln10    pe 
 p   p ln10
Definition of pe and constants λ and comes from Modified Cam Clay model

From principal stress space to p-q plane
Rendulic
1 1
plane: 2  3
1 q
q 2
3 p 3 2
2 3 3 2 p
1
mean stress: p  1  2  3 
3
1
deviatoric stress: q  1  2 2  2  3 2  3  1 2
2
Yield cap of Modified Cam-Clay model
q2
q pe  p 
CS - Line M2p
pe  ppe
6 sinCS
M 
3  sinCS
cap
CS  Critical State
 lnppe   ec /(  )

e
p´
p p

Elastoplastic MCC-model versus viscoplastic SSC-model
Both models have decomposition: e = ee + e c
ec  pp
e
Both with moving cap (density hardening):
q
MCC: cap is moved by primary loading
SSC: cap is moved by „time“,
cap stress well below cap: low creep rate

stress near the cap: high creep rate
p
ppe
MCC = Modified Cam Clay SSC = Soft Soil Creep
The 3D Soft Soil Creep Model
   e   c
q
 c pe
ε c  vol
 σ
27
σ
ε c e c *  pe 
 cvol  
1 e0   ppe 
p   pe / p
pe ppe

Part 3: Performance of Soft-Soil Creep Model on triaxial tests
q q
drained drained
failure
undrained undrained
p´ vertical strain 1
Undrained triaxial tests after Vaid et al. (1977) and simulation with SSC
model
undrained q
q
CS-Line
fast shearing
slow
p´
Vaid & Campanella (1977): Time-dependent behaviour of undisturbed clay, J. Geotech. Engng. Div., ASCE;
Vol. 103(7), pp. 693-709

Influence of strain rate on the undrained shear strength in triaxial
compression after Kulhawy and Mayne (1990)
Kulhawy & Mayne (1990): Manual on Estimating Soil Properties for Foundation Design
Options: effective cohesion and steep cap
“critical state line”

q Mohr-Coulomb failure line
1
Mcs Mmc
1
cap
p´
ppe
cćot ´
6 sin
Mmc 
3  sin
6 sincs
Mcs  Value of cs can be selected such that K0nc = 1 – sin
3  sincs
´
The more difference between Mmc and Mcs the steeper the cap. The above picture would
suggest the possibility of tensile stresses, but these can be omitted by using a „tension cut-
off“.

Input Parameters for Soft Soil Creep Model
 Cc
deformation parameters: *   modified compression index
1 e0 (1 e0 ) ln10
*  * / 10 modified swelling index

modified creep index
*  * / 30
Poisson´s ratio
ur  0.2 ( or 0.15 )
strength parameters: c’ ’ (for soft soil: ) 0
special parameters:
p
p ( or POP  p  0 or OCR  )

K 0  Knc
0  OCR 0  1  sin 
Knc 
Initial geostatic state for overconsolidated soils
POP  p  0
y
x input of OCR input of POP
0 p 0 p
y
p
Knc
0
1- vur POP
1
vur
y0
x
x0

Comparison of input data for two models
hardening soil model soft soil creep

model
E ref
ur
k *  3  pref  (1  2  ur ) / E ref
ur
 ur  ur
E ref
oed
*  pref / E ref
oed
c ´,  ´,  c ´, ´, 
p 0 p 0 ( input with OCR or POP )
E ref
50 -
- *
m always m = 1
q MC q MC
p´ p`
cap hardening + shear cap

hardening hardening
Part 4: application of SSC model on the Leaning Tower of Pisa
1173 Start of construction

1370 End of construction
1590 Galileo Galilei´s
fall tests

1370
1278
  5.5
1178
centre of gravity 58 m
22.5 m
wP = 28 % wL = 38 % w = 28 %
wP = 30 % wL = 70 % w = 52 %
wP = 13 % wL = 43 % w = 24 %
wP = 25 % wL = 51 % w = 38 %
Laval – Cc1
Laval – Cc2
Lancellotta e Pepe (1990)
Calabresi et al. (1993)
first analysis
last analysis

Deformation parameters * , * and *
Cc w 0.1
*   L
2.31 e 5
3 1 ur Cs
*   * / 10
2.3 1 ur 1 e
C
*   * / 30
2.31 e
z kPa
data Calabresi (1993) on p

[m]
data Lancellotta (1990) on p

depth
data from Laval samples on p
z0

first stage of construction (height 29 m)
average foundation pressure 310 kPa
settlement 0.5 m / inclination 0.2°
1178
maximum excess
pore pressure 74 kPa
second stage of construction (height 51

m)
aettlement 1.7 m / inclination 1.0°
1278
maximum excess
pore pressure 59
kPa

third construction stage: Bell Chamber (height 58
m)
Settlement 2.6 m / inclination 2.8°
1370
maximum excess
pore pressure 3
kPa

final analysis:
stiff clay layers and soft fill
1993
relative shear stresses
I MC
I MC
I MC
cu by Method A – Application for HK Soft Clay [1]

• Use a Hong Kong soft clay for study (Yeung & So, 2001)
• Measured cu (field Vane shear) and ′ (triaxial CU) available
• Undrained triaxial modelling using Method A – input of ′
• Three soil constitutive models investigated – Mohr Coulomb,
Modified Cam Clay and Soft Soil
• Two initial stress conditions investigated – K0 consolidated and
isotropically consolidated
• Objective: compare predicted cu with measured cu
• Pay attention to Effective Stress Paths (or excess pwp) predicted by
different soil models
34

cu by Method A – HK Soft Clay [2]
• In “drained” reclamation design involves soft clays, for initial

undrained loading stage designers have to
1. determine design undrained strength cu profile based on ground

investigation information
2. use field Vane shear, Cone Penetration Test (Piezocone) and triaxial
consolidated undrained tests on high quality samples
3. ensure soil model selected for analysis will predict the correct
design cu profile
35
Field Vane Shear vs. Triaxial

Compression & Extension
FV: Field Vane Shear
TC: Triaxial compression • Undrained strength cu
TE: Triaxial extension • Different field & lab results for
cu of an Italian soft clay
• cu of TC ≈ 1.1 × cu of FV
• cu of TE ≈ 0.7 × cu of FV
• Data by Ghionna, Jamiolkoeski,
Lacasse, Ladd, Lancellotta and
Lunne (1983)
36

• cu for Marine Clays from Field

Vane Shear tests (Yeung and So,
2001)
• A site offshore Lamma Island
• cu/′v0 ≈ 0.3 (from FV tests)
cu/′v0=0.3
• cu of TC ≈ 1.1 × cu of FV
• cu/′v0 ≈ 0.33 (inferred for TC)
• cu of TE ≈ 0.7 × cu of FV
• cu/′v0 ≈ 0.21 (inferred for TE)
• Soil model should predict
HK Marine Clays measured cu or design cu profile
37
p′=0.5*(′1+ ′3), q′=0.5*(′1- ′3)
c′=0 kPa, ′=30°
• Triaxial Consolidated Undrained tests with pwp measurement

• Hong Kong Marine Clays offshore Lamma Island
• Data by Yeung and So (2001)
38

• Undrained triaxial modelling – Mohr Coulomb (MC), Modified Cam
Clay (MCC) & Soft Soil (SS) models
• Use Method A with c′ = 0 and ′ = 30°
• Compare predicted cu with measured cu for two initial stress
conditions
1. ′h0 = 1-sin(′) × ′v0 [K0 consolidated]
2. ′h0= 1 × ′v0 [Isotropically consolidated]
• Select a particular ′v0, say 100 kPa
• Measured cu ≈ 0.33 × 100 ≈ 33 kPa (Tx. Comp.); Measured cu ≈
0.21 × 100 ≈ 21 kPa (Tx. Ext.)
39
• Oedometer tests for compression

index Cc of Marine Clays (Yeung
and So, 2001)
• Average Cc = 0.6
•  = Cc/2.3 = 0.26 [MCC input]
Ave. Cc=0.6 •  = /5 (assumed) = 0.052 [MCC
input]
• e0 = 2.0 (assumed)
• * = /(1+e0) [SS input]
• * = /(1+e0) [SS input]
•  &  (or * & *) affect effective
stress path in undrained shearing
40

• Plaxis triaxial (element)

a modelling
• Axisymmetric analysis
• Consolidated Undrained
(CU) test
• Compression shearing
r
(a ↑, r constant) –
beneath embankment
• Extension shearing (a
constant, r ↑) – near
embankment edge
41

′h0 = 1 - sin() × ′v0 [K0 consolidated]
140
120 q=2cu Mf=6sin()/3-
100 sin()
80 Measured q SS
q (kPa) =(1-3)
60
ult MCC MC
40
20 K0 Compression
0
-20 Extension
Measured qult
-40
-60
Mf=6sin()/3+sin(
-80
)
MC
-100
MCC Mf=6sin()/3-
-120 Soft Soil
-140 sin()
0 10 20 30 40 50 60 70 80 90 100 110
p' (kPa) =1/3×(′1+′2+′3)
42

Comparison of cu (kPa) – measurements & predictions
′h0 = 1 - sin() × ′v0 , ′v0=100 kPa [K0 consolidated]
Modified
Measur Mohr
Cam Soft Soil
ement* Coulomb
Clay
Tx.
33 40 30 33
Comp.
Tx.
21 28 30 24
Ext.
* For tx. compression cu/′v0 ≈ 0.33, for tx. extension cu/′v0 ≈ 0.21
43

′h0 = 1 × ′v0 [Isotropically consolidated]
140
q=2cu Mf=6sin()/3-
120
100
sin()
Compression
80 Measured q SS MC
ult
q (kPa) =(1-3)
60
40
MCC
20
0
K0
-20 Mf=6sin()/3+sin(
Measured qult
-40 )
-60
-80
MC
-100
MCC
-120 Soft Soil Extension Mf=6sin()/3-
-140
sin()
0 10 20 30 40 50 60 70 80 90 100 110
p' (kPa) =1/3×(′1+′2+′3)
44

Comparison of cu (kPa) – measurements & predictions
′h0 = 1 × ′v0 , ′v0=100 kPa [Isotropically consolidated]
Modified
Measur Mohr
Cam Soft Soil
ement* Coulomb
Clay
Tx.
33 61 35 42
Comp.
Tx.
21 43 35 30
Ext.
* For tx. compression cu/′v0 ≈ 0.33, for tx. extension cu/′v0 ≈ 0.21
45
Summary Method A Predicted cu HK Soft Clay [1]

• For isotropically consolidated soft clay (′h0=′v0), MC, MCC & SS
all over-predict cu in both tx. compression and tx. extension
• For K0 consolidated soft clay, ′h0=1 - sin()×′v0
1. in tx. compression MCC and SS predict cu close to measurement,
whereas MC over-predicts cu
2. in tx. extension MC, MCC and SS all over-predict cu
• For tx. compression the failure line in q-p′ space is
Mf = 6sin() / 3-sin() for MC, MCC and SS
• For tx. extension the failure line in q-p′ space is
Mf = 6sin() / 3+sin() for MC and SS
Mf = 6sin() / 3-sin() for MCC (single Mf value is input)
46

Summary Method A Predicted cu HK Soft Clay [2]
• Method A used with Mohr Coulomb model (input effective ′)
overestimates undrained strength cu – Not Safe!
• Method A used with Modified Cam Clay/Soft Soil appears to predict
reasonable cu in K0 consolidated tx. comp. (for this HK clay only).
But they will overestimate cu in situations of
1. soft clay is assumed isotropically consolidated
2. triaxial extension
3. soft clays showing highly anisotropic behaviour
• Method A used with advanced soil models must be validated by
measured cu using Vane shear/CPT/pressuremeter/lab CU tests
• Don’t know how to use advanced soil models → specify measured cu
to Mohr Coulomb, i.e. c′=cu, ′=0° (Method B/C)
47
Direct Input of cu Into Mohr Coulomb

q c = 0, ′ = ′
cu by Method A
Direct input cu
c = cu, ′ = 0°
ESP by Mohr Coulomb in
undrained shearing
P′
ESP for real soft clays
• For soft clays modelled by Mohr Coulomb, regardless of Method A

or Method B, the predicted Effective Stress Path (or excess pwp) is
always not correct – underestimate excess pwp
48

M.C. Vertical ESP in Undrained Test [1]
• Problem of Mohr Coulomb in effective stress modelling of soft, n.c.,
lightly o.c. clays – Method A with input of ′
• Mohr Coulomb has no compression cap (or yield locus)
No cap Compression cap
Mohr Coulomb Hardening Soil
49
• M.C. uses M.C. failure envelope as shear yield surface

• Before reaching failure envelope soil behaviour is isotropic elastic
• Hooke’s Law relates stress rates to elastic strain rates
Failure envelope
Elastic behaviour
50

• For isotropic elastic material, elastic response is
 p  1 / K  0  p
 
 q   0 1 / 3G q 
  Note: p = v
• No coupling between volumetric and shear effects

• Change in mean effective stress (p') generates volumetric strain
(p), but no shear strain (q)
• Change in shear stress (q) generates shear strain (q), but no
volumetric strain (p)
• But for real soils, both p' and q will cause p
• In undrained loading M.C. generates too small compressive epwp
51

• In undrained loading, change in excess p.w.p. (u) defined by
u  B 3  A1   3 
• For saturated soils, B ≈ 1

• For isotropic elastic soils, for example A = 1/3 in standard
compression test (2=3)
• Keep 3 constant (3=0) and increase 1 (1 = +ve)
• u = 1/3 1
• P = 1/3 1 (from p = 1/3 (1 + 2 + 3)
• P' = 0 (from p' = p - u)
• Real soils not elastic – ‘A’ = 0.5 to 1.0 for n.c. clays
u = (½ to 1)1, P′ < 0
52

• Soft, n.c. and lightly o.c. clays tend to contract in drained shearing
• In undrained shearing, the contractive tendency will manifest in
positive (compressive) excess p.w.p. (u)
• Effective stress path curves back from constant p' line
• The positive (compressive) U is affected by plastic straining before
reaching failure
u
Failure line
M.C. ESP
53

• Need a compression cap (yield locus) to model plastic straining
before failure
• For example, Modified Cam Clay yield locus
Yield locus
ESP
Muir Wood (1990)
• Plastic straining generates larger compressive excess pwp than elastic

straining (Mohr Coulomb)
• ESP bends to left with reduction in p', resulting in lower qult (cu)
54

• In Modified Cam Clay plastic straining on yield locus is
 p p 
 p 
    M 2   2 2 p
    Note: p = v
 q  p M 2   2    2
2

4 / M    q 
2 2

• Both p' and q will generate volumetric strain (pp), i.e. coupled
volumetric and shear effects → predict larger compressive epwp
-70 u = 65 kPa 1.0
0.9 MCC A = 0.95
-60 MCC
0.8
-50 0.7
u = 40 kPa
 u (kPa)
-40 0.6
A 0.5
-30 0.4 A = 0.33
-20 0.3
Mohr Coulomb
0.2
-10 Mohr Coulomb
0.1
0 0.0
0.00 -0.02 -0.04 -0.06 -0.08 -0.10 0.00 -0.02 -0.04 -0.06 -0.08 -0.10
a [-] a [-]
-ve u: compressive excess pwp
HK clay isotropically consolidated tx. comp.
55
Anisotropic Soil Model [1]

• Modified Cam Clay & Soft Soil are isotropic yield locus models
• Difficult to predict cu of soft clays showing highly anisotropic behaviour
Note:
q/p' ratio at peak
deviatoric stress < q/p'
ratio at critical state
56

Anisotropic Soil Model [2]
Isotropic Anisotropic
model model
Yield locus
Yield locus
Modified Cam-Clay
Sekiguchi-Ohta (1977)
type
57
Equation for cu in Mohr Coulumb [1]

• Equations for undrained strength cu predicted by Mohr Coulomb,
Modified Cam Clay & Soft Soil models
• Method A with effective friction angle (′)
• Mohr Coulomb (triaxial compression)
1 6 sin  
cu   
 pini
2 3  sin  
p′ini = initial mean effective stress
• Mohr Coulomb (triaxial extension)
1 6 sin  
cu   
 pini
2 3  sin 
58

Equation for cu in Mohr Coulumb [2]
• Mohr Coulomb (plane strain conditions)
Note: Effect
1
cu  K0  1sin   v0 of intermediate principal stress
2 ′2 not considered
′v0= initial effective vertical stress; K0 = 1 - sin ′
′ sin ′ K0 cu/’v0
15° 0.259 0.741 0.225

16° 0.276 0.724 0.238
17° 0.292 0.708 0.250
18° 0.309 0.691 0.261
20° 0.342 0.658 0.284
22° 0.375 0.625 0.304
24° 0.407 0.593 0.324
59
Equation for cu in Modified Cam Clay
• Modified Cam Clay (triaxial compression)
 
  p0 ini 
M f pini 
cu  
2  
 2 pini
Mf = (6 sin′)/(3 - sin′); Mf is input directly into Plaxis

p′0 ini = initial pre-consolidation pressure
q2 = Mf2p′(p′0 - p′); yield locus of Modified Cam-Clay model
q = deviatoric stress = (1 - 3)
60

Equation for cu in Soft Soil
• Soft Soil (triaxial compression)
* *
  M
M f pini 2
p0 ini  *
cu   2  
cap
2  M f  M cap pini
2
 
Mf = (6 sin′)/(3 - sin′); ′ is input into Plaxis

Mcap ≈ 3 - 2.8(K0), where K0 = 1 - sin′ ; or
Mcap from long equation in Slide 49
q2 = Mcap2p′(p′0 - p′); yield locus of Soft Soil model
61
Mcap & Mf Lines in SS & SSC Models
Mcap
Mf
• Mcap controls the steepness of cap (yield locus)

• Mcap has influence on ESP in undrained shearing
• Mf is the failure line – Mohr Coulomb failure criterion
• Soil fails on the Mf line
62

Mcap, Mf & K0nc in SS & SSC Models [1]
Soft Soil model Soft Soil Creep model

1. Input  1. Input 
2. Plaxis defaults K0nc=1-sin 2. Plaxis defaults
3. Plaxis determines Mcap from Mcap = Mf =
6sin / 3-sin[tx comp.]
4. Plaxis defaults
6sin / 3+sin [tx ext.]
Mf = 6sin / 3-sin [tx comp.]
3. Plaxis determines K0nc from
Mf = 6sin/ 3+sin [tx ext.]
63
Mcap, Mf & K0nc in SS & SSC Models [2]
Soft Soil model Soft Soil Creep model

• Users can change K0nc • Default Mcap = Mf
• K0nc > 1-sin, Mcap ↓ • Mcap cannot smaller than Mf
• K0nc < 1-sin, Mcap ↑ • User can only decrease K0nc
• Mcap ↑ predicts higher cu • K0nc ↓, Mcap ↑
• Mcap cannot smaller than Mf • Mcap ↑ predicts higher cu
• Mf always related to  • Mf always related to 
=24° =24°
64

Starting Creep Rate in Soft Soil Creep [1]
• At what creep rate Soft Soil Creep calculation starts?
• C constant b’cos in log10 scale

• Creep rate deceases with time
(secondary compression line is a
curve in normal time)
65
Starting Creep Rate in Soft Soil Creep [2]
• Starting creep rate defined by OCR

• By default, Plaxis specifies OCR=1, i.e. creep for only one day
• Only suitable for newly placed materials which exhibit large creep
deformation (high creep rate)
• OCR=1 not suitable for in-situ soils with long history
• Soils undergo creep deformation for hundreds/thousands of years,
i.e. OCR > 1
66

Specifying OCR to Soils with Long History
• Two ways to specify OCR

1. assign OCR in Initial Stress Conditions (K0-procedure); or
2. leave OCR =1 in Initial Stress Conditions, start calculation
with a “plastic nil” step and with a time interval t
• In (2) Plaxis will calculate stress state due to creep over t,
resulting in a certain OCR
• OCR = exp {(*/*-*)·ln(t)}
t = time in days elapsed since last primary loading step
67
Role of OCR in Self-weight Loading & Creep
Settlement [m]
0.00 Settlement
OCR0=2.0
of 10m
-0.05
thick soil
-0.10
OCR0=1.4 * = 0.10
-0.15 * = 0.02
* = 0.005
-0.20
ur = 0.15
OCR0=1.1 c′ = 0.0 kPa
-0.25
′ = 25°
-0.30 OCR0=1.0  = 0°
-0.35
K0nc = 0.577
0 200 400 600
Time (day)
800 1000 (1-sin ′)
68

References
1. Atkinson, J. H. & Bransby, P. L. (1978). The mechanics of soils – an introduction to
critical state soil mechanics. McGraw Hill.
2. Ghionna, V., Jamiolkowski, M., Lacasse, S., Ladd, C. C., Lancellota, R. and Lunne,
T. (1983). Evaluation of self-boring pressuremeter. Proc. Int. Symp. on Soil and Rock
Investigation by In Situ Testing, Paris, Vol.2, 293-301.
3. Lambe, T. W. & Whitman, R. V. (1969). Soil mechanics. Wiley.
4. Muir Wood, D. (1990). Soil behaviour and critical state soil mechanics. Cambridge
University Press.
5. Sekiguchi, H. & Ohta, H. (1977). Induced anisotropy and time dependency in clays.
Proc. 9th ICSMFE, Spec. Session 9, Tokyo, 229-238.
6. Skempton, A. W. (1954). The pore-pressure coefficients A and B. Geotechnique, 4,
143-147.
7. Vermeer, P. A. & Meier, C. P. (1998). Stability and deformations in deep excavations
in cohesive soils. Proc. Int. Conf. on Soil-structure Interaction in Urban Civil
Engineering. Darmstadt Geotechnics, Vol.1, No.4.
8. Yeung, A. T. & So, S. T. C. (2001). Geotechnical engineering properties of Hong
Kong Marine Clays. Proc. of 3rd Int. Conf. on Soft Soil Engrg., Hong Kong, 695-700.
69

CG6 The Soft Soil
Creep model
Dennis Waterman and Ronald
Brinkgreve
Plaxis BV
Contents
• Modified Cam-Clay model and
critical state
• Soft Soil model
• Soft Soil Creep model
• Parameters
• Creep and undrained analysis
• SSC compare with MC Model
108 of 607
Modified Cam-Clay
q Modified Cam-Clay:
Yield contour is an ellipse
M
1
plastic
elastic
pp p’
q2
Yield function: f   p'  p'  p p 
M2
Modified Cam-Clay
Elastic deformation is generated according to:
 p' 
ee  e0 e   ln  0  unloading/reloading
p 
Total deformation is generated according to:
 p' 
e  e0   ln   primary compression
 p0 
e = void ratio
 = swelling index
 = compression index
109 of 607
Modified Cam-Clay
Generally we prefer notation in strains:
 p'  
 ve   ve 0   * ln  0  ,  * 
p  1 e
 p'  
 vp   vp 0  ( *   * ) ln   ,  *

0
p  1 e
εv = volumetric strain
* = modified swelling index
* = modified compression index
Modified Cam-Clay
q
q2
εs f  2  p'  p'  p p 
M
dεs dε
M dεv
1
p’
pp εv
The Cam-Clay model has an associated flow rule, hence
the yield function f equals the plastic potential function g;
That is: f=g
110 of 607
Modified Cam-Clay
Plastic volumetric strains
 p'  p 
 vp   vp 0  ( *   * ) ln  0 
 ( *   * ) ln  0p  
p   pp 
 
dp p
d  vp  ( *   * ) 
pp
Plastic shear strains

2 pq
d  sp   d  vp (Burland, 1965)
M p q
2 2 2
Hardening rule
Modified Cam-Clay
Right from the M-line (“wet side”): q < M p’
d  vp  0, d  sp  0 (compaction, hardening)
Left from the M-line (“dry side”): q > M p’

d  vp  0, d  sp  0 (dilatancy, softening)
M
1
On the ellipse top: q = M p’
d  vp  0, d  sp   Failure!
CSL f=0
Therefore the M-line is referred to as the “Critical State Line”
111 of 607
Soft Soil
q
M
1
M K0nc
1
p’
pp
Soft Soil
Oedometer test
d xx  d zz  K0NC  d yy , d xy  0
d  xxe  d  zze , d  xye  0
d  xxp  d  zzp , d  xyp  0
Gives a relation between K0nc and M:
 * 
(1  K0 )(1  2 )  *  1
M 3
(1  K 0 ) 2
  
(1  2 K0 ) 2 *
(1  2 K 0 )(1  2 )  (1  K 0 )(1   )
*
M  3.0  2.8K 0
(Brinkgreve, 1994)
112 of 607
Soft Soil
q
M K0NC
1
α
p’
pp
M determines both the strength and the amount of
plastic strains generated when following the K0NC path
0.4  K 0NC  0.6  1.3  M  1.9
6sin( )
tan( )  M   32    46
3  sin( )
15    30  0.6  K0NC  0.9
One can have the correct K0NC or the correct strength,

but not both!
Soft Soil
q M
1 MC-line
K0NC
α
p’
pp
Solution in the Soft Soil model:

Add a Mohr-Coulomb failure surface to deal with the
strength and use the CSL only for determining K0NC
113 of 607
Soft Soil Creep
Model based on soil behaviour in oedometer test
• Creep decreases logarithmic with time
• Creep increases with load step, hence it increases with
effective stress level.
Soft Soil Creep model

Characteristics:
• Comparable behaviour with Soft Soil model:
• Logarithmic stress-strain relationship (stiffness
linearly dependent on p’)
• Elastic unloading / reloading
• Memory of preloading (preconsolidation stress)
• Irreversible volume strain upon primary loading
• Failure behaviour according to the Mohr-Coulomb
criterion
• Time-dependent deformations (secondary
compression)
• Irreversible strains by means of viscoplasticity (creep
strain) instead of plasticity 14
114 of 607
One-dimensional creep
modelling
One-dimensional compression:
B/C
d '1 C   '1 
d1  d1e  d1c   A 
 '1    p 
with:
c 
 p   p 0 exp   (development of preconsolidation stress)

 B 
A = 1D swelling index (elasticity)
A+B = 1D compression index (total strain)
C = 1D creep index
15
One-dimensional creep
modelling
16
115 of 607
3D model: Division of strains:
d  d e  d c  d p
d  e  D 1d  ' Elastic strains according to Hooke’s law
c dg c
d   d1 Creep strains (viscoplastic, time-dependent)
d '
dg f
d  p  d2 Plastic strains according to MC (failure)
d '
17
Soft Soil Creep

Soft Soil model describes the volume strain as sum of
instantaneous elastic and plastic volume strain:
 p'   p' 
d  v  d  ve  d  vp   *  0   ( *   * )  0 
p
  p 
It is tempting to “just” add a time-dependent creep part:
 p'   p' 
v  d  ve  d  vp  vc   *  0 
 ( *   * )  0   C  f ( p, t )
p  p 
Two problems:
• What is the “instantaneous plastic volume strain” in case of
undrained behaviour?
• Creep is also plastic behaviour – why the difference?
116 of 607
Soft Soil Creep
Therefore the Soft Soil creep model assumes:
 *  *
 p '   * p  eq *   vc
v         
e c *
 eq  , with p  p e
eq eq  *  *
 p' 
v v p p0
 pp 
*= modified compression index

*= modified swelling index
μ*= modified creep index
τ = load step period > preset to 1 day
Soft Soil Creep

Creep potential, equivalent isotropic and preconsolidation stress:
 *  *
 p '  p  eq *   vc
v           *  eq 
e c *
, with p  p e
eq eq  *  *
v v
 p' p  p p0
 p 
q M
1
p eq p eq p’
p
p p
117 of 607
Soft Soil Creep
Under constant load the cap expands in time and so
generates plastic volume strains: creep!
• Hence, the preconsolidation pressure Pp increases
in time and therefore so does the OCR.
A higher ratio between load and preconsolidation

pressure Pp means a higher creep rate and vice versa.
• If a load step is applied the creep rate suddenly
increases and then slowly decreases subsequently
• If an unload step is applied the creep rate suddenly
decreases and then keeps slowly decreasing.
Soft Soil Creep

Creep is always present; in a drained analysis,
during consolidation and “after” consolidation.
Hence, there is no such thing as “primary consolidation”
and “secondary consolidation”.
Creep never stops! It only goes slower with time…
118 of 607
Parameters
Parameters *, * and μ* are obtained from oedometer tests
μ* is the creep rate when OCR = 1
ε1=εv ε1=εv
ln p’ ln t
1
*
μ* 1
* 1
Rule of thumb (when insufficient data available):

* *
 5  10  15  30
* *
Parameters
Parameters *, * and μ* can be linked to internationally
normalized parameters:
Cc 2 Cr C
*  *  * 
2.3(1  e) 2.3 1  e 2.3(1  e)
Cc, Cr and C : 1-dimensional compression indices

*, * and μ* : isotropic compression indices.
1
Primary loading: dp  (1  K 0nc ) d yy
3
Unloading/reloading: no linear relation between dp and dyy
* is estimated assuming, on average, isotropic stress
during unloading
119 of 607
Parameters
 *   * primary compression plasticity
Creep ratio: 
* creep plasticity
CR < 5
Very creep-sensitive materials; fresh deposits, soft sediment
CR > 25
Materials with little creep.
One may consider using the Soft Soil model instead
Parameters
Overconsolidation pressure pp
• Important parameter: it determines the initial creep rate
• No direct input in Plaxis, but input by means of OCR or POP
OCR (Over-Consolidation Ratio)

• By default OCR=1 → only 1 day creep has occurred.
Hence, the initial creep rate will be (too) high than reality!
• Two possibilities to determine initial OCR:
• Simulation of time history
• Estimate with rule of thumb value of OCR
120 of 607
Parameters
OCR - Simulation of history
1. Perform K0 procedure with OCR=1 or
Gravity Loading in 1 day
2. Define a next calculation phase, plastic nil-step,
with a time interval equal to the estimated geological time in da
that the material has been at the site since deposition, that is, the
last primary loading step.
OCR - Estimation
Perform K0 procedure with OCR according to:
 *  *
ln( t )
*
OCR  e
Δt = elapsed time in days since last primary loading step
Creep and undrained

analysis
d   d   d   0  d   d 
v
e
v
c
v
e
v
c
v
Creep strains are always negative: volume loss.

This must be compensated by elastic volume increase.
p'
p '   ve , p '  0  ve  0  p '  0

p  p w  p '  0  p w   p '  p w  0
Hence, creep causes an increase of excess pore pressures
121 of 607
Creep and undrained
analysis
Creep and consolidation
Creep
Increases excess pore pressures and consolidation
time; BUT Decreases effective stresses and so the
creep rate
Consolidation
Increases effective stresses and so the creep rate; BUT
Decreases excess pore pressures
Creep and undrained

analysis100 kPa  = = 17 kN/m
k =k = 1e-6 m/day
sat unsat
3
x y
B * = 0.025
* = 0.005
c = 1 kPa
φ = 28º
 =0º
1m
A
3 alternatives:
μ*= 0.0040 (CR = 5)
μ*= 0.0020 (CR = 10)
μ*= 0.0001 (CR = 200)
1) Apply load in 1 day

1m
2) Consolidate until
pw < 0.1 kPa
122 of 607
Creep and undrained
analysis
Excess PP [kN/m2]
-120
CR = 5
CR = 10
CR = 200
-90
CR=10: 104 kPa after 84 days

CR= 5: 107 kPa after 105 days
-60
CR=200: t= 4.8 year
CR= 10: t= 9 year
OCR0 = 1 CR= 5: t=24 year
-30
0
1 10 100 1e3 1e4 1e5
Time [day]
Creep and undrained

analysis
-Uy [m]
0
0.03 26 mm
CR = 5
CR = 10
CR = 200
0.06
77 mm
0.09
97 mm
0.12
1 10 100 1e3 1e4 1e5
Time [day]
123 of 607
Creep and undrained
analysis
Uy [m]
0
-0.02
-0.04
CR=10
-0.06
OCR = 1.5
-0.08 OCR = 1.2
OCR = 1.0
-0.1
1 10 100 1e3 1e4 1e5
Time [day]

Relationships with other compression indices:
 
Cam-Clay *  *  -
1 e 1 e
1 2 1
Dutch literature *  *  * 
C p' C p ( ontl .) 2.3 C's
Den Haan  * A  B  *  2A *  C
Cc Cs C
International lit.  * *  * 
2.3(1  e) 1 e 2.3 (1  e)
34
124 of 607
Comparison of SSC and MC
model
Stress-strain development in various stress paths:
Soft-Soil Creep model: Mohr-Coulomb model:
* 0.10
* 0.02 E 2000 kPa
* 0.005  0.30
ur 0.15 c’ 0.0 kPa
c’ 0.0 kPa ’ 25°
’ 25°  0°
 0°
K0nc 0.677 (1-sin ’)
35

model
Isotropic compression:
1000
p' [kN/m2] MC SSC
800
600
400
200
0
0 0.2 0.4 0.6 0.8
eps-v
36
125 of 607
SSC model in standard tests
Isotropic compression:
1000
p' [kN/m2]
100 1 1
* *
10
1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
eps-v 37

Loading 10, 20, 40, 80, 160, 320, 640 kPa + 100 d. creep
38
126 of 607
CID tri-axial test at different loading rates:
160
q' [kN/m2]
120
900 kPa/d
80
9 kPa/d
40
0
0 0.05 0.10 0.15 0.20 0.25 0.30
eps-1
39

CIU-test at different loading rates (only creep):
80
900 kPa/d
q' [kN/m2]
90 kPa/d
60
9 kPa/d
40
20
0
0 0.01 0.02 0.03 0.04 0.05
eps-1
40
127 of 607
CIU-test at different loading rates (only creep):
80
q' [kN/m2] 900 kPa/d
90 kPa/d
60
9 kPa/d
40
20
0
0 20 40 60 80 100
p' [kN/m2]
41

CID-test at different loading rates (creep + consolidation):
160
9 kPa/d
q' [kN/m2]
120
90 kPa/d
900 kPa/d
80
9000 kPa/d
40
00 40 80 120 160
p' [kN/m2]
42
128 of 607
model
The role of OCR in self-weight loading and creep:
Settlement
Settlement [m]
0.00 of 10m
OCR0=2.0
thick layer
-0.05
* 0.10
-0.10
OCR0=1.4 * 0.02
-0.15 * 0.005
ur 0.15
-0.20
c’ 0.0 kPa
-0.25
OCR0=1.1
’ 25°
 0°
-0.30 OCR0=1.0
K0nc 0.677
-0.35
0 200 400 600 800 1000 (1-sin
Time ’)
Soft-Soil Creep Model
Possibilities and advantages compared to Mohr-Coulomb:
• Better non-linear formulation of soft soil behaviour (NC-clay;

logarithmic compression)
• Distinction between primary loading and unloading / reloading
• Memory of pre-consolidation stress
• Stiffness parameters based on oedometer tests
• Primarily suitable for compressive stress paths
Course for Experienced Plaxis users
Geo-engineering
129 of 607
Soft-Soil Creep Model
Limitations and disadvantages:
• Not suitable for other types than soft soils

• Not recommended for excavations and pure unloading situations
• No peak strength and softening (directly to residual strength)
• No anisotropy
Geo-engineering
Summary
• Soft soil creep:
• Is a time dependent cam-clay based model
• Distinguishes between elastic and creep strains
where creep includes all plastic deformations
• Elastic deformations only depend on *
• Plastic deformations depend on *,μ* and
the initial OCR
• Parameter determination can be difficult,
simulation of a laboratory test is advised to
check model parameters.
130 of 607
Thank you
Questions?
Geo-engineering
131 of 607
COMPUTATIONAL GEOTECHNICS COURSE
CG7a CONSOLIDATION
CONSOLIDATION: OUTLINE
 Introduction
 Basic theory of groundwater flow
 Permeability
 Confined and unconfined problems
 Finite element formulation for consolidation
 Mechanical problem
 Hydraulic problem
 Global equations
 Time step
 Boundary conditions
Singapore 2011
132 of 607
TYPES OF ANALYSIS
 Drained
 Loading/Construction/ excavation: very slow (in relation to the soil
permeability)
 Undrained
 Loading/Construction/ excavation: very fast (in relation to the soil
permeability)
 Intermediate cases: consolidation analysis

 Both mechanical and hydraulic (flow) problems interact
 More complex computations: coupled analysis
EXAMPLE
Excess pore
water pressure
Consolidation
Initial state Undrained loading Final state
Singapore 2011
133 of 607
OTHER EXAMPLES
 Construction at intermediate rates
 Change of hydraulic conditions
BASIC THEORY OF GROUNDWATER FLOW

 Darcy´s law
h
q  k  k i
L
h 
q  k  k
dy dy
p
h y
w
q : flow h   : total head

y: vertical coordinate
p : water pressure
Singapore 2011
134 of 607
BASIC THEORY OF GROUNDWATER FLOW
 Permeability often anisotropic
 
qx  k x q y  k y
x y
 Equation of continuity (no net inflow in an

elementary area)
q x q y
 0
x y
 Excess pore pressure

p: (active) water pressure
p  p steady  p excess p steady : steady state pore pressure
p excess : excess pore pressure
PERMEABILITY
 Dependence on grain size
Soil k (cm/s)
Clean gravel >1
Clean sand 1 - 10-2
(coarse)
Sand mixture 10-2 - 5x10-3
Fine sand 5x10-2 -10-3
Silty sand 2x10-3 -10-4
Silt 5x10-3 -10-5
Clay 10-6 and less
Harr (1962)
Singapore 2011
135 of 607
PERMEABILITY
 Dependence on void ratio
PERMEABILITY
 PLAXIS allows consideration of change of permeability with void ratio
 k  e
log   
 k 0  ck
Default value for c k is 1015
 There may be large contrasts of permeability between different materials in

the same problem
 Too much permeability contrast may cause numerical difficulties
 The ratio between the highest and lowest permeability value should not
exceed 105
 To simulate an almost impermeable material (e.g. concrete), a value
lower by a factor 1000 is sufficient
Singapore 2011
136 of 607
TYPES OF FLOW PROBLEMS
 Confined flow  Unconfined flow
Domain defined Domain undefined
TRANSITION SATURATED/UNSATURATED

qx   K r kx
x

qy  K r ky
y
Kr 1 saturated zone
K r  10 4 unsaturated zone
4h
K r  10 4 h hk log( K r )  
hk
hk  0.7m (PLAXIS)
Singapore 2011
137 of 607
FINITE ELEMENT FORMULATION FOR CONSOLIDATION (1)
 Effective stresses
 Constitutive law
 Discretization
 In terms of excess pore pressure

 same shape functions for displacements
and pore pressures
 Mechanical problem: equilibrium equation
Stiffness matrix
Coupling matrix
Incremental load vector
Singapore 2011
138 of 607
 Hydraulic (flow) problem: continuity equation
Flow matrix
Coupling matrix
Water compressibility matrix
 Global system of equations
 Step-by-step integration procedure
0 <  < 1 ; Generally, fully implicit)
Singapore 2011
139 of 607
 Time step
 Automatic time stepping is required
 Critical time step
 Consolidation analysis
 Prescribed time
 Maximum excess pore pressure
BOUNDARY CONDITIONS
 Flow boundary conditions

 Prescribed groundwater head
 Closed flow boundary
 Closed consolidation boundaries
Singapore 2011
140 of 607
BOUNDARY CONDITIONS
 Seepage surfaces Seepage surface

Groundwater head = y
BOUNDARY CONDITIONS
 Infiltration  Sources and sinks

Drain line element
Prescribed flow
Prescribed head
Potts & Zdravkovic (1999)
Singapore 2011
141 of 607
INITIAL PORE PRESSURE AND EXCESS PORE PRESSURE
Sand
Clay Δpw
Sand pw0
e.g. t=0 t=1 t = 10 t = 40 t = 100 days
pw  pw 0   pw p w 0  steady state pore pressure
 Steady state pore pressures have to be established

 Consolidation in PLAXIS V9 is in terms of excess pore pressures
FINITE ELEMENT FORMULATION - PERMEABILITY
 PLAXIS allows consideration of change of permeability with void ratio
 k  e
log   
 k 0  ck
Default value for c k is 1015
 There may be large contrasts of permeability between different materials in the same
problem
• Too much permeability contrast may cause numerical difficulties
• The ratio between the highest and lowest permeability value should not exceed
105
• To simulate an almost impermeable material (e.g. concrete), a value lower by a
factor 1000 is sufficient
Singapore 2011
142 of 607
1D CONSOLIDATION - NUMERICAL SIMULATION
Parameters used in numerical simulation
Soil Parameters Vertical Drain Parameters

HS Model s m 1.9 Spacing
Parameters Clay de m 2 Influence Diameter (triangular pattern)
Type - UnDrained l m 10 Length
γunsat kN/m3 15 axb mm2 100x4 Dimension of Drain
γsat kN/m3 16 dw mm 60 Drain Diameter
kx m/day 4.32 x 10-4 AxB mm2 120x50 Dimension of Mandrel (rectangular)
ky m/day 4.32 x 10-4 dm mm 88 Mandrel Diameter
E50ref kN/m2 4300 ds mm 220 Smear Diameter
Eoedref kN/m2 1800 kh m/day 4.32 x 10-4 Soil
Eurref kN/m2 14400 kv m/day 4.32 x 10-4 Soil
cref kN/m2 15 khaxi m/day 4.32 x 10-4 Undisturbed Zone (axisymmetry)
φ degree 27 kvaxi m/day 4.32 x 10-4 Undisturbed Zone (axisymmetry)
Ψ degree 0 khps m/day 9.98 x 10-5 Undisturbed Zone (plane strain) (Indraratna)
υur - 0.2 kvps m/day 4.32 x 10-4 Undisturbed Zone (plane strain)
pref kN/m2 100 khps m/day 1.15 x 10-4 Undisturbed Zone (plane strain) (CUR 191)
m - 0.9 kvps m/day 4.32 x 10-4 Undisturbed Zone (plane strain) (CUR 191)
K0nc - 0.546 khps m/day 4.32 x 10-4 Undisturbed Zone (plane strain) (CUR 191)
cincr kN/m3 0 kvps m/day 3.45 x 10-3 Undisturbed Zone (plane strain) (CUR 191)
yref m 0 kh’axi m/day 2.15 x 10-4 Smear Zone (axisymmetry) (Indraratna)
Rf - 0.9 kv’axi m/day 4.32 x 10-4 Smear Zone (axisymmetry)
T-Strength kN/m2 0 kh’ps m/day 4.99 x 10-5 Smear Zone (plane strain) (Indraratna)
kv’ps m/day 4.32 x 10-4 Smear Zone (plane strain)
Parameters for Soft Soil model are chosen accordingly, stiffness for MC
model represents oedometer stiffness of HS model in middle of layer
ISOCHRONES
permeable
Tv
D
permeable
degree of consolidation Ut
Isochrones: lines of excess pore pressures (u t) at a given time
Singapore 2011
143 of 607
drainage at top and bottom
Isochrones at 25% consolidation:

not symmetric for advanced models
Influence of constitutive model on consolidation behaviour
Singapore 2011
144 of 607
Influence of parameter B on consolidation behaviour
1D CONSOLIDATION - COMPARISON WITH ANALYTICAL SOLUTION
MC Model
This image cannot currently be display ed.
0,1 1 10 100 1000

0
0,1
0,2
Degree of consolidation
0,3
0,4
2D Axi WithOut PVD
0,5
2D Axi PVD
0,6
2D Axi PVD&Smear
0,7
Terzaghi WithOut PVD (Analytical)
0,8
Barron PVD (Analytical)
0,9
Barron PVD&Smear (Analytical)
1
time (day)
Singapore 2011
145 of 607
1D CONSOLIDATION - COMPARISON WITH ANALYTICAL SOLUTION
HS Model
This image cannot currently be display ed.
0,1 1 10 100 1000

0
0,1
0,2
Degree of consolidation
0,3
0,4
2D Axi WithOut PVD
0,5
2D Axi PVD
0,6
2D Axi PVD&Smear
0,7
Terzaghi WithOut PVD (Analytical)
0,8
Barron PVD (Analytical)
0,9
Barron PVD&Smear (Analytical)
1
time (day)
Singapore 2011
146 of 607
01223ÿ5617289ÿ
2698
6 6ÿ238ÿ
9ÿ11 89ÿ62ÿ861
9ÿ2
ÿ!"ÿ#$%&$'()
147 of 607
148 of 607
149 of 607
150 of 607
151 of 607
152 of 607
153 of 607
154 of 607
155 of 607
156 of 607
157 of 607
158 of 607
159 of 607
160 of 607
161 of 607
162 of 607
163 of 607
164 of 607
165 of 607
166 of 607
167 of 607
168 of 607
169 of 607
170 of 607
CG8 - Drained and Undrained
Analysis
Prof Harry Tan Siew Ann
Centre for Soft Ground Engineering
National University of Singapore
Some of the used material was originally created by:

Prof. Helmut Schweiger, Technical University of Graz, Austria
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES Drained and Undrained Behavior
Contents
• Definition drained / undrained
• Drained / undrained soil behaviour
• Typical results from drained and undrained triaxial tests
• Skempton‘s parameters A and B
• Modelling undrained behaviour with Plaxis
• In terms of effective stresses with drained strength parameters
• In terms of effective stresses with undrained strength parameters
• In terms of total stresses
• Influence of constitutive model and parameters
• Influence of dilatancy
• Undrained behaviour with Mohr-Coulomb Model
• Undrained behaviour with Hardening Soil Model
• Excavation Example
• Summary
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES 2
Singapore 2011
171 of 607
Drained / undrained
• Drained analysis appropriate when

• Permeability is high
• Rate of loading is low
• Short term behaviour is not of interest for problem
considered
• Undrained analysis appropriate when

• Permeability is low and rate of loading is high
• Short term behaviour has to be assessed
Drained / undrained
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
T t Eoed = Oedometer modulus
γ w D2 w = Unit weight of water
D = Drainage length
t = Construction time
T = Dimensionless time factor
U = Degree of consolidation
Singapore 2011
172 of 607
Undrained behaviour
Implications of undrained soil behaviour:
• Excess pore pressures are generated

• No volume change
In fact small volumetric strains develop because a finite (but
high) bulk modulus of water is introduced in the finite
element formulation
• Predicted undrained shear strength depends on soil model
used
• Assumption of dilatancy angle has serious effects on
results
UNDRAINED BEHAVIOUR
 Implications of undrained soil behaviour

• excess pore pressures are generated
• no volume change
in fact small volumetric strains develop because a finite (but high)
bulk modulus of water is introduced in the finite element formulation
• predicted undrained shear strength depends on soil model used
• assumption of dilatancy angle has serious effects on results
q
e
lin
advanced lu re
fai
models
advanced
models
elastic-perfectly
cu,3 plastic models
cu,1 cu,2
pc’ p’
Results from undrained triaxial tests using simple and advanced constitutive models
Singapore 2011
173 of 607
UNDRAINED BEHAVIOUR
q [kN/m 2 drained
q [kN/m2] ]
350 Mohr-Coulomb / Soft Soil
300
undrained
250
Mohr-Coulomb
200
undrained
Soft Soil
150
100
50
0
0 -50 -100 -150 -200 -250 -300 -350
p' [kN/m2]
p' [kN/m2]
Undrained shear strengths predicted by Mohr-Coulomb and Soft Soil Model
Triaxial test (NC) – drained / undrained

Typical results from drained (left) and undrained (right) triaxial tests on normally consolidated soils
(from Atkinson & Bransby, 1978)
Singapore 2011
174 of 607
Triaxial test (OC) – drained / undrained
Typical results from drained (left) and undrained (right) triaxial tests on overconsolidated soils
Undrained triaxial test– NC / OC

Typical results from undrained triaxial tests on (a) normally consolidated and (b) overconsolidated clay (from
Ortigao, 1995)
Singapore 2011
175 of 607
Skempton’s parameters A and B
Skempton 1954:  p w  B   3  A   1    3 
- Fully saturated soil
- No inflow / outflow of pore water
- Bulk modulus of soil grains is considered to be very high
- Isotropic linear elastic material behaviour (Hooke´s law)
 vol , skeleton   vol , pore water

p' E´
 vol , skeleton  K' 
K' 31  2´ 
n p w
 vol , pore water 
Kw

Assuming triaxial compression:  1 ;  2   3
  1  2   3  3  pw K w
p w  
3K ' n
1  1 
leading to  pw 
nK '   3  3  1   3 
1  
Kw
pw  B  3  A 1   3 
1 1
with B A
nK ' 3
1
Kw
Singapore 2011
176 of 607
• Notes on parameters A and B:
• For Kw large compared to K´, parameter B ~ 1.0

(corresponds to pw = p > p´ = 0)
• Small amount of drapped air reduces parameter B
significantly (see next figure)
• Parameter A depends on stress path, even for elastic material
behaviour
• Parameter A cannot be determined a priori for complex
elastic-plastic constitutive models but is a result of the model
behaviour for the stress path followed
Dependence of pore pressure parameter B on degree of saturation
Singapore 2011
177 of 607
Undrained behaviour with PLAXIS
PLAXIS automatically adds stiffness of water when undrained material type is
chosen using the following approximation:
Kw Eu 2 G 1   u 
K total  K'   
n 31  2 u  31  2 u 
E' 1   u 
K total  assuming u = 0.495
3 1  2 u  1  '

Notes:
• This procedure gives reasonable B-values only for ´ < 0.35 !
• Real value of Kw/n ~ 1.106 kPa (for n = 0.5)
• In Version 8 B-value can be entered explicitely for undrained materials

Example 1:
E´ = 3 000 kPa, ´ = 0.3, u = 0.495

 K´ = 2 500 kPa, Ktotal = 115 000 kPa  Kw/n = 112 500 kPa
1
with B = 0.978 > reasonable value for saturated soil
nK '
1
Kw
Example 2:
E´ = 3 000 kPa, ´ = 0.45, u = 0.495

 K´ = 10 000 kPa, Ktotal = 103 103 kPa  Kw/n = 93 103 kPa
B = 0.903 > poor value for saturated soil
Singapore 2011
178 of 607
Method A (analysis in terms of effective stresses):
type of material behaviour: undrained
effective strength parameters c´, ´, ´
effective stiffness parameters E50´, ´
Method B (analysis in terms of effective stresses):

undrained strength parameters c = cu,  = 0,  = 0
Method C (analysis in terms of total stresses):

type of material behaviour: drained or non-porous
total strength parameters c = cu,  = 0,  = 0
undrained stiffness parameters Eu, u = 0.495

Notes on different methods:
• Method A:
• Recommended
• 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 avilable
• 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)
Singapore 2011
179 of 607
Undrained strength from Mohr circle
Consider fully undrained isotropic elastic behaviour
(Mohr Coulomb in elastic range)
pw = p > p´ = 0
 centre of Mohr Circle remains at the same point
cu  
1 'o

   'yo sin '  c' cos '
2 x
Mohr Circle for evaluating undrained shear strength (plane strain)

UNDRAINED STRENGTH: MOHR COULOMB vs ADVANCED MODELS
Parameters used for Hardening Soil model (HS)

Parameter Meaning Value
 [kN/m³] Unit weight (unsaturated) 16
r [kN/m³] Unit weight (saturated) 16
ϕ′ [°] Friction angle (Mohr-Coulomb) 24
c′ [kPa] Cohesion (Mohr-Coulomb) 0
ψ [°] Angle of dilatancy 0
ur [-] Poisson’s ratio unloading-reloading 0.20
E50ref [kPa] Secant modulus for primary triaxial loading 2 000
Eoedref [kPa] Tangent modulus for oedometric loading 1 000
Eurref [kPa] Secant modulus for un- and reloading 7 500
m [-] Exponent of the Ohde/Janbu law 1.0
pref [kPa] Reference stress for the stiffness parameters 100
K0nc [-] Coefficient of earth pressure at rest (NC) 1-sin(ϕ′)
Rf [-] Failure ratio 0.90
σTension [kPa] Tensile strength 0
Parameters for Soft Soil model and Mohr Coulomb model accordingly
Singapore 2011
180 of 607
UNDRAINED STRENGTH: MOHR COULOMB vs ADVANCED MODELS
Undrained triaxial test (numerical simulation - Method A)

Preconsolidation to 100 kPa > compression / extension
125
Advanced models - Method A
100
(HS / HSS / SS)
75
2cu
50
q [kN/m ]
2
25 Mohr Coulomb model

0
-25
-50
-75
-100
0.00 -25.00 -50.00 -75.00 -100.00 -125.00 -150.00
p' [kN/m 2 ]
Triaxial compression: Triaxial extension:

cu/v' = 0.47 (MC) cu/v' = 0.35 (MC)
cu/v' = 0.36 (HS) cu/v' = 0.26 (HS)
UNDRAINED STRENGTH: MOHR COULOMB vs HARDENING SOIL MODEL
Undrained test plane strain (numerical simulation - Method A)

K0-consolidation to v' = 100 kPa > compression / extension
-75
-50
-25
cu
t [kN/m ]
2
Mohr Coulomb model

25
50
s = ½(x + y)
t = ½(x - y)
75
0.00 -25.00 -50.00 -75.00 -100.00 -125.00 -150.00
2
s' [kN/m ]
Plane strain compression:
cu/v' = 0.32 (MC)
cu/v' = 0.29 (HS)
Singapore 2011
181 of 607
UNDRAINED STRENGTH: MOHR COULOMB - HARDENING SOIL MODEL
Undrained triaxial test (numerical simulation - Method A)

Test 1: Preconsolidation to 200 kPa > compression / extension
Test 2: Preconsolidation to 200 kPa > isotropic unloading to 150 kPa > compression / extension
Test 3: Preconsolidation to 200 kPa > isotropic unloading to 100 kPa > compression / extension
200
150
Hardening Soil model 2cu (NC)
Test 1
100
cu = 0 (OC)
50 Test 2
Test 3
q [kN/m ]
2
-50
-100
-150
Mohr Coulomb model
-200
0.00 -25.00 -50.00 -75.00 -100.00 -125.00 -150.00 -175.00 -200.00 -225.00
p' [kN/m 2 ]
Undrained test plane strain (numerical simulation)

K0-consolidation to v' = 100 kPa > compression
Comparison METHOD A / METHOD B

-40
Mohr Coulomb
Method A
Hardening Soil cu (MC - HS)
-30 Method A Method A
Mohr Coulomb
Method B
t [kN/m ]
2
-20 (cu based on HS-Method A)
Hardening Soil
Method B
-10
0 s = ½(x + y)
0.00 -25.00 -50.00 -75.00 -100.00
s' [kN/m 2 ] t = ½(x - y)
Singapore 2011
182 of 607
Undrained test plane strain (numerical simulation)

K0-consolidation to v' = 100 kPa > compression
Comparison excess pore pressures - METHOD A / METHOD B

Difference in excess pore pressure
-20
for HS model when using Method A
Hardening Soil or B
Method A (Note: cu is the same)
EX_PP [kN/m ]
Mohr Coulomb
2
Method A
-10 Mohr Coulomb
Method B
(cu based on HS-Method A) Hardening Soil
Method B
(cu based on HS-Method A)
0
0.00 -10.00 -20.00 -30.00 -40.00
s' [kN/m 2 ]
Consequence: consolidation analysis following undrained analysis with method B by means of "model change"
(introducing effective
PLAXIS FINITE ELEMENT strength
CODE FOR SOILparameters) starts with incorrect pore pressures
AND ROCK ANALYSES
UNDRAINED STRENGTH: INFLUENCE OF DILATANCY
300
275
250
225
200
175
q [kN/m ]
2
150
125
100
75
M C non dil
50 M C dil
H S_1 non dil
25 H S_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
 1 [% ]
Simulation of undrained triaxial compression test – MC / HS model - q vs 1
Singapore 2011
183 of 607
300
275  > 0: unlimited strength

250
225
200
175
q [kN/m ]
2
150
125
100
75 MC non dil
MC dil
50 HS_1 non dil
HS_1 dil
25 total stress path
0
0.00 25.00 50.00 75.00 100.00 125.00 150.00 175.00 200.00 225.00 250.00
2
p' [kN/m ]
Simulation of undrained triaxial compression test – MC / HS model - q vs p´
100
90 M C non dil
M C dil
80
H S_1 non dil
excess pore pressure [kN/m ]
2
70 H S_1 dil
60
50
40
30
20
10
-10
-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
 1 [% ]
Simulation of undrained triaxial compression test – MC / HS model - pw vs 1
Singapore 2011
184 of 607
1.0
0.9 M C non dil
0.8 M C dil
HS _1 non dil
0.7 HS _1 dil
0.6
0.5
parameter A
0.4
0.3
0.2
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
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
 1 [% ]
Simulation of undrained triaxial compression test – MC / HS model - A vs 1
UNDRAINED STRENGTH: INFLUENCE OF K0nc
125
100 K0nc increased
75 K0nc decreased
50
q [kN/m ]
2
25
-25
-50
-75
-100
0.00 -25.00 -50.00 -75.00 -100.00 -125.00 -150.00
2
p' [kN/m ]
Soft Soil Model
Singapore 2011
185 of 607
UNDRAINED STRENGTH: INFLUENCE OF Eoed and K0nc
125
100
75 HS-TC-Eoed >
HS-TC
nc
HS-TC-K0 <
50
HS-TC-Eoed + K0nc >
q [kN/m2]
25 HS-TE-Eoed >
HS-TE
nc
HS-TE-K0 <
0
HS-TE-Eoed + K0nc >
CSL-TC
-25
CSL-TE
-50
-75
-100
0.00 -25.00 -50.00 -75.00 -100.00 -125.00 -150.00
p' [kN/m2]
Hardening Soil Model
INFLUENCE OF FLOW RULE
p
 sin 'm 
sin 
ˆ m  sin  m  
 sin ' 
e.g. SOREIDE 2003
consequence of flow rule in

undrained triaxial compression
stress path
Undrained shear strength often correlated to
effective vertical stress:
cu = '; with  ~ 0.25 – 0.35
Note: undrained shear strength of
 < ~ 0.28 with HS-model difficult to achieve
Singapore 2011
186 of 607
EXAMPLE - EMBANKMENT CONSTRUCTION
Influence of consolidation on stability

(only possible with Method A and advanced model)
influence of construction speed is investigated
"fast" construction: 2 days of consolidation per placement of 1 m embankment
"slow" construction: 3 days of consolidation per placement of 1 m layer embankment
Influence of consolidation on stability (Method A)

"slow": max. excess porewater
pressure: 86 kPa
"fast": max. excess porewater

pressure: 100 kPa
Singapore 2011
187 of 607

"slow": stable
"fast": failure

excess pore pressure [kPa]
ex c es s pore pres s ure [kN/m2]
C hart 1
-50
slow
fast fast
-40
-30
-20
slow
-10
0
0 4 8 12 16
Time [day]
time [days]
Singapore 2011
188 of 607
Simplified representation of stress path during embankment construction

t = ½ (1 - 3)
s = ½ (1 + 3)
s = ½ (1 + 3)
from Ortigao, 1995
COMPARISON METHODS A AND B

Practical example: deep excavation in soft clay0
schluffiger TON und feinschluffiger grober
SAND
-5 organischer TON
-10
breiiger mariner TON
UPPER MARINE CLAY
-15
schluffiger TON
-20
Depth [m]
-25 breiiger mariner TON
LOWER MARINE CLAY

-30
cu = 300 kN/m² organischer TON

-35
schluffiger TON
-40
altes Schwemmland feinschluffiger grober
SAND schluffiger TON
OLD ALLUVIUM
-45
0 20 40 60 80 100
cu-Profile according to geotechnical design parameter
c_u [kN/m²]
table
Singapore 2011
189 of 607
Practical example: deep excavation in soft clay
100
Cu (Mohr-Coulomb)
95 Cu (Geotechnical Design Parameter Table)

Method A:
90
Mohr-Coulomb effective (' and c')
Method B:
Method A
85 Mohr-Coulomb (cu)
Elevation, RL (m)
> significant difference in undrained

80 shear strength
Method B
75
70
65
60
0 20 40 60 80 100
Shear strength (kPa)

Practical example: deep excavation in soft clay
ABH 32 ABH 31
Fill Fill
E (U) E (U)
c’, ’ Method A with Mohr-Coulomb (!)

M (U) M (U) overestimates undrained shear strength for
~62mm Marine Clay
normally consolidated soft soils
F2 ~170mm
F2
cu > difference in calculated horizontal

M (L) displacements significant
M (L)
> bending moments differ by a factor of 2
Marine Clay
> strut forces differ by approx. 10%
E (L) > minor differences in earth
F2
OA pressures
F2 SW2
OA
SW2 OA
SW1
OA
SW1
OA CZ OA CZ
Singapore 2011
190 of 607
SUMMARY
 Undrained analysis should be performed in effective stresses and with

effective stiffness and strength parameters
 Undrained shear strength is result of the constitutive model

Check if you have data on undrained shear strength !
If uncertain use undrained parameters directly!
 Dilatancy angle should be set to zero!
 Note that for NC-soils in general

• factor of safety against failure is lower for short term (undrained)
conditions for loading problems (e.g. embankment)
• factor of safety against failure is lower for long term (drained)
conditions for unloading problems (e.g. excavations), but for very
soft soils it may be different
Real Excavation Case in Stiff Residual

Soils-What is likely Field Conditions?
cv * t
T
H2
Cv=66 m2/day
H = 15 m
Cv=125 m2/day
Cv=53 m2/day
Singapore 2011
191 of 607
What is West Coast Station Situation ???
• Cv= 50 m2/day
• H=15 m
• t = 100 days
• T=50*100/(15*15) = 22.2 >>> 0.4
• Situation on Passive Side is likely to be DRAINED
Condition
Excess PP at Formation Level for k=1e-7 and 1e-8 m/s
Singapore 2011
192 of 607
Cases of k=1e-7 to 1e-9 m/s
Displacements at Formation Level
Cases of k=1e-7 to 1e-9 m/s

BMs at Formation Level
Singapore 2011
193 of 607
Wall Deflection at B (15/83.85 – 1.65m above FL)
U x at B
Ux at B [m]
0.2
DRN
UND
0.15
k=1e-7
k=1e-8
0.1 k=1e-9
0.05
0
0 50 100 150 200 250
Time [day]
Heave at C(0/78.7 – 3.5m below FL)

U y at C
Heav e at C [m]
0.035
DRN
0.03
UND
0.025
k =1e-7
0.02
k =1e-8
0.015 k =1e-9
0.01
5e-3
-5e-3
0 30 60 90 120
Time [day ]
Singapore 2011
194 of 607
References
Atkinson, J.H., Bransby, P.L. (1978)
The Mechanics of Soils, An Introduction to Critical State Soil 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 remarks on pore pressure parameters A and B in undrained analyses with the Hardening Soil
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
Singapore 2011
195 of 607
Structural Elements &
Modelling Excavations in Plaxis
Based on original course note by Ronald
Brinkgreve, Plaxis B.V.
Uç
RF Shen
24 November 2011
Structural elements in Plaxis
• Plate element
• Anchor element
• Geogrids element
• Interface element
196 of 607
Application of structural elements
wall strip footing tunnel
geotextile wall ground anchor cofferdam
strut anchored wall
1. Plate Element
Overview:
• 3 or 5 noded line elements
(for 6‐noded or 15‐noded element mesh)
• 3 degrees of freedom per node
• Plates have:
– Axial forces
– Shear forces
– Bending moments
– Hoop forces (axisymmetry)
• Elastic or elastoplastic behaviour
• For modelling walls, floors, tunnels
197 of 607
Plate Element
Plates – elastic parameters
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
b
Plate Element
Plates – elasto‐plastic behaviour
Np
M
Mp
198 of 607
Plate Element
(Illustration: Mp‐Np.P2D):
1200 ‐100‐90 ‐80 ‐70 ‐60 ‐50 ‐40 ‐30 ‐20 ‐10 0 10 20 30 40 50 60 70 80 90 100
Envelope 0
Elasto‐plastic plate
1000
Elastic plate
‐5
800
Elasto‐plastic plate
‐10 Elastic plate
600
N
400 ‐15
200 ‐20
0
‐25
‐200 0 200 400 600 800 1000 1200 M
M
Plate Element
Effect on Global FOS by c/phi Reduction
CBP Elastic, CBP Elasto-Plastic

Failure with no Failure with
Plastic Hinge, Plastic Hinge,
FOS=1.75 FOS=1.40
• Elastic wall excludes possibility of wall plastic hinge; and over-estimate FOS=1.75
• Allowing for wall plastic hinge (Elasto-plastic wall) gave lower FOS=1.40 and
smaller soil yielded zone behind the wall
8
199 of 607
Plate Element
Plates – weight, in soil
Actual problem In the model
dreal
wreal =  concrete  d real wmodel =  soil  d real  w plate Below GT

 soil   sat
wmodel = wreal  w plate = (  concrete -  soil )  d real
Above GT
 soil   unsat
Plate Element
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  w plate = ( concrete   soil )  d real Above GT
2
 soil   unsat
200 of 607
Plate Element
Plates – connections
Spring data:
• Stiffness
6 8
• Min/Max moment
Rotation
spring
5 7
Hinged connection
Rigid connection
(default)
Illustration: Connection.P2D
Plate Element
Walls – thin wall vs. thick wall
• Thin wall
– Wall thickness << wall length
– No much end‐bearing, only friction
→ Plate element suffices
• Thick wall
– Wall thickness significant
– End‐bearing capacity needed
→ Use soil elements with material set represen ng wall material
→ In order to obtain structural forces a plate with fictitious
properties may be inserted
201 of 607
Plate Element
Walls – thick wall

• Soil elements with material set representing wall material
• Difficult to obtain structural forces from soil elements, therefore introduce
very flexible fictious plate within the solid wall elements:
– No influence on deformation: low stiffness, no weight
– Located in on the neutral line (usually the middle)
– Tight bonding to the concrete elements: no interfaces
(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
2. Anchor Element
Anchors – fixed‐end
• To model supports, anchors and struts
– Elasto‐plastic spring element
– One end fixed to point in the geometry,
other end is fully fixed for displacement
– Positioning at any angle
– Pre‐stressing option
Anchors – node‐to‐node
• To model anchors, columns, struts and rods
– Elasto‐plastic spring element
– Connects two geometry points in the geometry
– No interaction with the mesh along the anchor
rod
– Pre‐stressing option
202 of 607
Anchor Element
Anchors – material properties
Axial stiffness, EA (for one anchor) [kN]
Spacing, Ls (out‐of‐plane distance between anchors) [m]
Maximum anchor force for compression and tension,
|Fmax,comp| and |Fmax,tens| [kN]
Ls
Anchor Element
Anchors – pre‐stressing
• Defined in Staged construction phase
• Both tension (grout anchor) or compression (strut)
possible
Tension = positive
203 of 607
3.Geogrid Element
Geogrids
• 3 or 5 noded line element
• Elastic or elasto‐plastic behaviour
• No flexural rigidity (EI), only axial stiffness (EA)
• Only allows for tension, not for compression
Anchor Element + Geogrid Element
Ground anchors
• Combination of node‐to‐node anchor and geogrid
• Node‐to‐node anchor represents anchor rod (free length)
(no interaction with surrounding soil)
• Geogrid represents grouted part (full interaction with surrounding soil)
• No interface around grouted part; interface would create unrealistic failure
surface
• Working load conditions only – no pullout
• If pullout force is known this can be used by limiting anchor rod force
204 of 607
Ground anchors
Axial forces in ground anchors:
Input
geometry
Probable actual distribution of axial

forces in ground anchor
Generated
mesh
axial forces in geotextile element
Nrod <> Ngrout due to shared node

between anchor, geotextile and soil
4. Interface Element
Interfaces – material properties
• Soil‐structure interaction
– Wall friction
– Slip and gapping between soil and structure
• Soil material properties
– Taken from soil using reduction factor Rinter
Cinter = Rinter * Csoil
tan(φinter) = Rinter * tan(φsoil)
ψinter = 0 for Rinter < 1
= ψsoil Rinter = 1
σt,inter = Rinter * σt,soil
Ginter = (Rinter)2 * Gsoil
– Individual material set for interface possible
205 of 607
Interface Element
Interfaces – reduction factor
Suggestions for Rinter:
– Interaction sand/steel = Rinter ≈ 0.6 – 0.7
– Interaction clay/steel = Rinter ≈ 0.5
– Interaction sand/concrete = Rinter ≈ 1.0 – 0.8
– Interaction clay/concrete = Rinter ≈ 1.0 – 0.7
– Interaction soil/geogrid (grouted body) = Rinter≈ 1.0
(interface may not be required)
– Interaction soil/geotextile = Rinter≈ 0.9 – 0.5 (foil, textile)
With reference to BS8002:
Pile‐soil interaction ‐ NSF on pile
(Ref: PhD thesis by RF Shen (2008)
Net downdrag loads (kN)

0 400 800 1200 1600 2000 2400
0
-1 Top sand layer
-2
Clay
-3 Test data
-4 Back-analysis
-5
Interface angle=10°
-6
-7 Interface angle=22°
-8 No interface
Elevation (m)
-9 curve
-10
-11
-12
-13
-14
-15
-16
Rigid base
-17
-18
-19
-20
-21
When using MC model and MCC model, the friction angle of the interface
element should be defined as follows:  9 
 4 M 
2

  (1   ' )   'i nt  arctan  4 
 'i nt  arctan  
 9
 '  9  2 M  
2
(MC model)  4 (MCC model)
206 of 607
Illustration: Interface_slide.P2D
4m
Unit weight = 25kN/m^3 1m
Interface properties: Cint = 2.5kPa; inter=26.6
Expected sliding resistance = Block length (4m) * Cint + Block weight (100kN) * tan (inter) =
60 kN
Material Behaviour in Excavation
• Unloading due to excavation
– Vertical unloading at excavation bottom
– Horizontal unloading behind wall (may accompanied by
shear plasticity)
• Primary loading due to pre‐stressing
• HS‐small model is preferred
– Non‐linear elastic unloading/reloading behaviour
– Shear plasticity due to horizontal unloading
– High far‐field stiffness for better settlement trough
prediction
207 of 607
Material behaviour: Stress paths
Construction phases:
Kactive K0
I 1st excavation σv
II Pre-stressing anchor K=1
III Final excavation Point A
II III
I
Point A
Point B
Kpassive
Point B
σh
Expected material behaviour in excavation
Eur , E50
Eoed
Eur , E50
Eur
E0 (far field small strain behavior)
208 of 607
Material behaviour
• Mohr‐Coulomb: unrealistic deformations
– Use of single E fails to cater for the complex material at various zones
– Overestimation over bottom heave
– Sometimes heave of soil behind the wall
– Soils below excavation behaves with Eur, even soils behind wall behaves
between Eur and E50. Use of E50 is too conservative.
• Hardening Soil model: qualitative realistic deformations
– Soil stiffness for Isotropic loading, shearing and unloading‐reloading can
be catered for automatically in the model.
– More realistic bottom heave
– Improved settlement trough behind wall.
• HS‐small model: qualitative and quantitative realistic deformations
– Improved version over HS to take care of far field small strain behavior
– More realistic settlement trough behind the wall (narrower and deeper)
Hands-on session
E3 – Excavation Exercise
209 of 607
Modelling groundwater in PLAXIS
Ronald Brinkgreve
Plaxis BV / Delft University of Technology
Contents
• Pore pressures
– Generation
– Definition
– Steady-state
• Groundwater flow
– Flow in unsaturated soil
– Material data sets
– Boundary conditions
– Finite Element Modelling
Singapore 2011
210 of 607
Pore pressures - generation
• Hydrostatic
– Based on defined phreatic levels
– Takes only water weight into account
– Aquifers possible
– Simple, horizontal water levels – no flow
– Undisturbed subsoil – simple excavations
• Steady-state groundwater flow
– No flow field change in time
– Long term solution
– Constant water levels
– Most excavations, some embankments

• Transient groundwater flow
– Flow field changes in time:
• Constantly changing natural water conditions
• Relatively fast building process, pumping, wells
– Embankments with river changes, tides
– Reservoir impoundment and drawdown
– Precipitation problems
– Pore pressure relaxation problems
Singapore 2011
211 of 607
• Hydrostatic pore pressures
– PLAXIS 2D, PLAXIS 3D, (PlaxFlow 3D Tunnel, 3D Foundation)
• Steady-state groundwater flow
– PLAXIS 2D, PLAXIS 3D*, (PlaxFlow, 3D Tunnel)
• Transient groundwater flow
– PLAXIS 2D, (PlaxFlow)
*PLAXIS 3D 2011 or higher
Pore pressures - definitions

• Steady-state pore pressures
– Due to phreatic levels
– Result of either hydrostatic distribution or steady or transient groundwater
flow analysis.
• Excess pore pressures
– Due to undrained behaviour
– Generated during the calculation
• Active pore pressures
– Steady-state + Excess pore pressures
Singapore 2011
212 of 607
Pore pressures - modes
• Classical mode
– Semi-coupled analysis.
– Full coupling between deformation and excess pore pressures.
– Steady-state pore pressures are generated prior to calculation and are thus INPUT.
– Changes in steady-state pore pressures may change excess pore pressures and
deformations, but not v.v.
– Soil is either fully saturated or dry
• Advanced mode
– Fully coupled analysis.
– Full coupling between deformation, steady-state pore pressures and excess pore
pressures.
– Changes in steady-state pore pressures, excess pore pressures and deformations
influence each other.
– Soil can be partially saturated
Advanced mode was introduced in Plaxis 2D 2010. Earlier versions only have classical
mode.
Pore pressures – steady-state

Phreatic levels
• Pore pressure based on distance below phreatic level and water weight
• Should only be used for horizontal water levels
• Different soil layers can have different phreatic levels
p = h * w
Singapore 2011
213 of 607
Phreatic levels
• General phreatic level

– Physical water level
– Causes both pore pressures and external water pressures if outside the
geometry.
• Cluster phreatic level
– Phreatic level for just one cluster
– Used for instance for confined aquifers.
• Cluster dry
– Sets cluster as dry, hence resets pore pressures to zero
• Interpolate
– Vertically interpolate pore pressures between adjacent clusters or lines
– Simplification for groundwater flow through low permeable layer

Phreatic levels
Layer 1
General phreatic level
Interpolate Layer 2
Cluster phreatic level Layer 3
Singapore 2011
214 of 607
Steady-state flow
• Calculation based on:

– Boundary conditions:
• Prescribed water levels
• Closed flow boundaries (bottom, axis of symmetry)
• Wells and drains
• Interface elements (on=impermeable, off=permeable)
• Inflow / outflow
– Soil permeabilities
Phreatic level in the soil is being calculated for t=∞

Steady-state flow
3 28 29 6 9 30 31 2
4 8 11 5
General General
General
16 17
21 26
19 18
20 23 24 27
22 25
13 14 12
15 7 10
0 1
Singapore 2011
215 of 607
Transient flow
• Calculation based on:

– Time-dependent boundary conditions:
• Changing prescribed water levels
• Closed flow boundaries (bottom, axis of symmetry)
• Wells and drains
• Interface elements (on=impermeable, off=permeable)
• Inflow / outflow
– Soil permeabilities
• Until Plaxis 2D V9 only by means of the PlaxFlow program, starting

from Plaxis 2D 2010 integrated.
Groundwater flow – flow in unsaturated soil

Water content and permeability in unsaturated zone
k  k re l k sa t , k re l  f  h p , S 
 (h p )
S (h ) 
n
hp=Ψ => pressure head
Singapore 2011
216 of 607
Linear Model
krel
1 hp  0 Saturated

1 k rel  1  m h p 0  hp   Partially saturated
 hp   Dry

m
 0 hp
β
For numerical stability

hp = -ε
hp = 0
1 Ae
 
3 N int

van Genuchten model
Singapore 2011
217 of 607
van Genuchten model
1 g n

S ( h p )  S res  ( Ssat  S res ) 1  g a h p   
gn ( g )
n
2
  g n 1  
 g n   g n 
     S  Sres
krel  S    S e  l  1   1  S e n  
g g 1
 with Se 
    Ssat  Sres
   
 
Ssat,Sres: saturated and residual saturation

ga, gn and gl: curve fitting parameters

Approximate van Genuchten model
 1 if hp  0
 Linear in Saturation

 
hp
S hp  1  if hps  h p  0
 hps
 0 if hp  h ps

 1 if hp  0
 4h p Log-linear in Permeability
 h
krel  h p   10 pk 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=10-4
Singapore 2011
218 of 607
Groundwater flow - material data sets
• Parameters:
– Permeabilities (kx, ky)
– Void ratio (to calculate storage)
– 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)
– Maximum unsaturated zone height
• Soil classification
– Particle fractions
– Predefined series (Staring, Hypres, USDA) with Van Genuchten and
Approx. van Genuchten parameters.
– User-defined

Soil classification - Staring
Dutch soil classification system
18 upper soils data sets

18 lower soil data sets
Upper soils:
< 1m below soil surface
Lower soils:
all deeper soils
Singapore 2011
219 of 607
Soil classification: Hypres
Hydraulic Properties of
European Soils
Particle distribution:
• < 2μm
• 2μm - 50μm
• 50μm – 2mm
5 upper soils data sets

5 lower soil data sets
1 organic soil data set

Soil classification: USDA
United States Department

of Agriculture
Particle distribution:
• < 2μm
• 2μm - 50μm
• 50μm – 2mm
12 soils data sets
No difference between
upper and lower soils
Singapore 2011
220 of 607
Soil classification and Van Genuchten parameters
Relative permeability
Degree of saturation
Groundwater flow - boundary conditions

2
Boundary condition on line between points 1 and 2
1
Name User input Condition
Closed -- q1,2  0
Head General phreatic level (hw) p1,2  ( y1,2  hw )  w
Head(user-defined) h1 and h2, or p1 and p2 p1,2  ( y1,2  h1,2 )  w
Inflow q1 and q2
Outflow q1 and q2
Singapore 2011
221 of 607
2
1
Infiltration
q1, q2,    max  h  y   max
Ψ1,min, Ψ2,min ,Ψ1,max, Ψ2,max
 min     max  inflow/outflow
   min  h  y   min
Plaxis has a shortcut button “Precipitation” to apply infiltration

boundaries to all boundaries available for precipitation taking into
account the slope of those boundaries.

Internal conditions
2
1
Drain Head (h) h1,2  h

Well - extraction Qwell, Ψmin    min  Q  Qwell
- infiltration Qwell    min  Q  0
Screen (interface) Equal to closed flow boundary
Singapore 2011
222 of 607
Groundwater flow - Finite Element Modeling
• GWF calculation generally needs finer mesh than deformation analysis
• GWF calculation generally needs often more additional steps than
deformation analysis
• GWF calculation usually converges, but not always to the correct solution
from an engineering point of view. This can be due to:
• Mesh coarseness
• Distorted elements
• Large differences in permeabilities
Groundwater flow - Finite Element Modeling

• Evaluation of results:
• Qualitative:
Do we believe the flow field and/or phreatic level at all?
• Quantitative:
Heads, pore pressures compared to hydrostatic,
compare with measurements.
Singapore 2011
223 of 607
Singapore 2011
224 of 607
CG11a - Slope Stability by FEM
PROF Harry Tan Siew Ann

Department of Civil Engineering
National University of Singapore
CONTENTS
Part 1
Advantages of FEM cf LEM
 Cases of Slope without GWT effects
Part 2
 CUT Slope – Undrained (Method A or B)
 CUT Slope – Drained
 CUT Slope – Undrained and Consolidation

Advantages of FEM vs LEM
for Slope Stability
• FEM need no assumption of location and shape of
slip surface
• FEM do not use concept of slices, no assumptions
about slice side forces. Global equilibrium preserved
until failure is reached
• FEM computes deformation of slope if realistic soil
model and stiffness is used
• FEM model is able to track progressive failure as
slope is excavated or build up
Cases of Slope without GWT
• Case 1 – Homogeneous slope without foundation

layer (D=1.0)
• Case 2 – Homogeneous slope with foundation layer
(D=1.5)
• Case 3 – Undrained slope with a thin weak layer
• Case 4 – Undrained slope with a weak foundation
layer (D=2.0)

Case 1 - Homogeneous slope without
foundation layer (D=1.0)
c’ = 20*10*0.05 = 10 kPa  '  20 o

H = 10 m
2:1 Slope
c' /  H  0.05
DH = 10 m
Definition of Failure by c/phi reduction
FS=1.343 cf 1.38 by Bishop LEM
c ' f  c ' / FS
tan  ' f  tan  ' / FS
Use c’ and phi’ reduction until soil mass reached limit

equilibrium; with large continuous displacements of a
point on slope crest

Visualize Failure Mechanism
Deformed mesh
Incremental displacement vectors
Visualize Failure Mechanism

Shear Strains Plot
Total strains

Case 2 - Homogeneous slope with
2:1 Slope
H = 10 m
DH = 15 m
NO GWT
 '  20 o
c' /  H  0.05
Case 2 - Homogeneous slope with

FS = 1.343

Case 3 - Undrained slope with a thin weak
layer
cu1 =50 kPa phi=0

2H Case 3a cu2=50 kPa
2H
Case 3b cu2 =30 kPa
Case 3c cu2=10 kPa
H
cu1 2H
H cu2
u  0 o
Case 3 - Undrained slope with a thin weak layer

(deformed meshes at failure)
cu1 =50 kPa phi=0 cu2/cu1 = 1.0
Case 3a cu2=50 kPa FS=1.45
Case 3b cu2 =30 kPa

Case 3c cu2=10 kPa
cu2/cu1 = 0.6
FS=1.35
cu2/cu1 = 0.2
FS=0.47

Case 3 - Undrained slope with a thin weak layer
(Incremental displacements at failure)
cu1 =50 kPa phi=0 cu2/cu1 = 1.0
Case 3a cu2=50 kPa FS=1.45
Case 3b cu2 =30 kPa

Case 3c cu2=10 kPa
cu2/cu1 = 0.6
FS=1.35
cu2/cu1 = 0.2
FS=0.47
Case 3 - Undrained slope with a thin weak layer (Shear Strains

at failure)
cu1 =50 kPa phi=0
Case 3a cu2=50 kPa cu2/cu1 = 1.0
Case 3b cu2 =30 kPa FS=1.45
Case 3c cu2=10 kPa
cu2/cu1 = 0.6
FS=1.35
cu2/cu1 = 0.2
FS=0.47

Case 3 - Undrained slope with a thin weak layer (FoS after
Griffiths and Lane (1999)
cu2/cu1 = 1.0
FS=1.45
cu2/cu1 = 0.6
FS=1.35
cu2/cu1 = 0.2
FS=0.47
Case 4 - Undrained slope with a weak

2:1 Slope
cu1=50 kPa
cu2 = 30, 75 and 100 kPa

cu2/cu1 = 0.6, 1.5, 2.0

Case 4 - Undrained slope with a weak foundation layer
(D=2.0) Incremental Displacements
cu2/cu1 = 0.6
FS=0.97
cu2/cu1 = 1.5
FS=2.02
cu2/cu1 = 2.0
FS=2.08

(D=2.0) Shear Strains
cu2/cu1 = 0.6
FS=0.97
cu2/cu1 = 1.5
FS=2.02
cu2/cu1 = 2.0
FS=2.08

(D=2.0) FoS after Griffiths and Lane 1999
cu2/cu1 = 2.0
FS=2.08
cu2/cu1 = 1.5
FS=2.02
cu2/cu1 = 0.6
FS=0.97
UNDRAINED BEHAVIOUR WITH PLAXIS

Notes on different methods:
Method A:
 recommended
 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
(consolidation analysis will not change shear strength)
 cannot be used with the Soft Soil model and the Soft Soil Creep model
Method C:
 NOT recommended (should give deformation close to Method B)
 no information on excess pore pressure distribution (total stress
analysis)

UNDRAINED STRENGTH FROM MOHR CIRCLE
Consider fully undrained isotropic elastic behaviour
(Mohr Coulomb in elastic range)
pw = p > p´ = 0
 centre of Mohr Circle remains at the same point
cu 
1
2
 
 x'o   y'o sin  '  c' cos  '
cu  K 0  1 y'o sin  '  c ' cos  '

1
2
Fig.6 Mohr Circle for evaluating undrained shear strength (plane strain)
Factor of Safety of Cuts/Excavations

Critical FS is Long-term
unloading condition,
For permanent cuts
drained strength is key
parameter for safe
design
For temporary cuts,
need to consider if
undrained or partially
drained condition

Factor of Safety of Embankments
Critical FS is Short-term
loading condition,
undrained strength is key
parameter for safe
design
Example of Underwater CUT Slope

LIMIT EQUILIBRIUM ANALYSIS OF CUT SLOPES
The figures below show the results of SLOPE/W calculations of FS
for a underwater cut slope in the undrained and drained condition,
by Bishop's Simplified method.
Drained and Undrained Parameters
The drained parameters are c'=2 kPa, '=240, =16 kN/m3
The equivalent undrained parameters are obtained from:
u u 
cc c' cos φo 'm
c ' cos ,
 ' m'sin  '
sin
At top of clay; c u  2 cos 24 0  1 .83 kPa
K 0  1  sin '  1 - sin 24 0  0.59
 ,v
,
m  1  K 0   6/2 1  0.59   4.77 kPa/m
2
,
m sin '  4.77 sin 24 0  1.94 kPa/m

Bishop‘s FS for Drained CUT
Cut Slope in Clay (Drained)
1.403
26
24
22
Water Level
20
18
Elevation (m)
16
14 Description: Clay Water
12 Soil Model: Mohr-Coulomb
10 Unit Weight: 16
8 Cohesion: 2 1:2 Cut
6 Phi: 24
4
2
0
0 5 10 15 20 25 30 35 40 45 50 55 60
Distance (m)
Bishop‘s FS for UnDrained CUT

Cut Slope in Clay (UnDrained)
2.085
26
24
22
Water Level
20
18
Elevation (m)
16
14
Water
12
Description: Clay
10
Soil Model: S=f(datum) 1:2 Cut
8 Unit Weight: 16
6 C - Datum: 1.83
4 Rate of Increase: 1.94
2 Datum (elevation): 20
0
0 5 10 15 20 25 30 35 40 45 50 55 60
Distance (m)

PLAXIS Analysis Cases
 Drained Analysis with c’=2 kPa and ’=24o
 Method A (analysis in terms of effective stresses):
effective strength parameters c´, ´, ´
 Method B (analysis in terms of effective stresses):
undrained strength parameters c = cu,  = 0,  = 0
 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
Drained CUT, Plaxis c/phi FS=1.35 cf LE=1.40
Drained Analysis with

Effective strength parameters c´=2 kPa, ´=24, ´=0
Effective stiffness parameters E50´=15000 kPa, ´=0.2

Method A - UnDrained CUT plus Full Consolidation
Plaxis c/phi FS=1.37 cf LEM=1.40
Method A (undrained)
Effective strength parameters c´=2 kPa, ´=24o, ´=0o
Slip circle same as Drained Case
Method A - UnDrained CUT,

Plaxis c/phi (Ignore UnDrained) FS=2.75 cf LE=2.09
Method A (in terms of effective stresses, undrained)


Method A - UnDrained CUT,
Plaxis c/phi (UnDrained) FS=2.27 cf LE=2.09
Method A (in terms of effective stresses, undrained)

Method B - UnDrained CUT,

Plaxis c/phi (Ignore UnDrained) FS=2.13 cf LE=2.09
Method B (in terms of effective stresses, undrained)

Undrained strength parameters c=1.83 kPa, ∆c=1.94 kPa, =0, =0

Method B - UnDrained CUT,
Plaxis c/phi (UnDrained) FS=2.14 cf LE=2.09
Method B (in terms of effective stresses, undrained)

Undrained strength parameters c=1.83 kPa, ∆c=1.94 kPa, =0, =0
c/phi Analysis of Method A and B

MC-UNDRAINED
Sum-Msf
3
METHOD A (IGNORE UNDR)
A (Ignore Undrained) =2.75 METHOD A (UNDR)
METHOD B (IGNORE UNDR)
2.5 METHOD B (UNDR)
A (Undrained) =2.27
2
B (Ignore Undrained) =2.13
B (Undrained) =2.14
1.5
1
0 3e3 6e3 9e3 1.2e4
|U| [m]

SUMMARY OF FS FOR CUT SLOPES
Analysis Condition PLAXIS SLOPE/W
Drained 1.35 1.40
A+Consolidation 1.37 1.40
A (Ignore UNDR) 2.75 2.09
A (UnDrained) 2.27 2.09
B (Ignore UNDR) 2.13 2.09
B (UnDrained) 2.14 2.09
Compare Excess PP of Method A
Method A Method A, c/phi Method A, c/phi

- Ignore - Undrained,
Undrained, FOS FOS = 2.27
= 2.75
Exc PP Exc PP changed,

unchanged, but but FS nearly
FS not OK OK

Compare Excess PP of Method B
Method B Method B, c/phi Method B, c/phi

- Ignore - Undrained,
Undrained, FOS FOS = 2.14
= 2.13
Exc PP Exc PP changed,

unchanged, but but FS is OK
FS is OK
Conclusions
• FEM analysis for Slope Stability is better than LEM

as failure mechanism is determined automatically as
part of the stress equilibrium process
• FEM can handle undrained, drained and
consolidation effects on slope stability, provided we
use Mohr-Coulomb failure criteria for c/phi
reduction analysis

01223ÿÿ56789ÿ5
36
ÿÿ66ÿ
9ÿ69
66ÿ99
ÿ69ÿ
ÿ
9 ÿ7ÿÿ79ÿ

ÿ99
ÿ
29ÿ
5 2! ÿ
" #ÿ7 66
$!
ÿ ÿ
% %&772
'
$(

76ÿ(#9
ÿ7ÿ5879
#9
7
ÿ7ÿ"!
ÿ(%

ÿ5 87ÿ3 ÿ00)5ÿ
ÿ)*+,!ÿ
,!ÿ5-
.96!ÿ2/+20ÿ598ÿ$11/
244 of 607
245 of 607
246 of 607
247 of 607
248 of 607
249 of 607
250 of 607
251 of 607
252 of 607
253 of 607
254 of 607
255 of 607
256 of 607
257 of 607
258 of 607
259 of 607
260 of 607
261 of 607
262 of 607
263 of 607
264 of 607
265 of 607
266 of 607
GENERATION OF INITIAL STRESSES & PHI-C
REDUCTION ANALYSIS
Dr William Cheang Wai Lum

Plaxis Asia
some of the slides were originally created by
Dr Ronald Brinkgreve
Dr Thomas Benz
Mr Dennis Waterman
CONTENTS
1. Part 1 : Initial Stresses
a. Ko-Procedure
b. Gravity Switch On
2. Part 2 : Phi-C reduction (Shear Strength Reduction Analysis)
a. Using Phi-C reduction
b. Some Examples
c. Pointers
d. Appendix
e. References
Singapore 2011
267 of 607
Ko and Gravity Switch-On Procedure
PART 1: INITIAL STRESSES
INITIAL STRESSES
1. Initial stresses represent the equilibrium state 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
Singapore 2011
268 of 607
K0-PROCEDURE
1. Generation of initial stresses during input.
2. Require coefficient for earth pressure at rest Ko
3. Disadvantage: No equilibrium for inclined surface
4. Advantage: No displacements are generated, only

stresses.
 'h   'v K0
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  so K0 
1  1   
Singapore 2011
269 of 607
GRAVITY LOADING
1. Procedure
a. Initial phase
a. Skip K0 procedure, ΣMweight remains zero
b. Phase 1
a. Choose Plastic calculation, Total multipliers
b. Set weight multiplier ΣMweight = 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 behaviour in Phase 1 to prevent

the generation of unrealistic excess pore pressures
2. K0 procedure has been used first
1. In the Initial phase redo the K0 procedure, but with

ΣMweight = 0; this will reset all initial stresses to zero.
Singapore 2011
270 of 607
GRAVITY LOADING
Cases where gravity loading should be used instead of K0-procedure:
Phi-C Reduction in Plaxis
PART 2-SAFETY FACTOR ANALYSIS
Singapore 2011
271 of 607
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
deformation analysis
6. Powerful alternative approach
Safety factor
• Many possible definitions
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: safety factor on soil resistance
Singapore 2011
272 of 607
SAFETY FACTOR ANALYSIS
Phi/c reduction:
a. Reduction of strength parameters c and tan(f) until failure is reached.
b. The factor of safety :
tan reduced
Lowered incrementally
c tan 
Msf  
creduced tan reduced
Calculation procedure:
a. Create a phi/c reduction phase
b. Accept the default increment for Msf=0.1 from the multiplier tab-
sheet.
c. Calculate
d. Carefully examine ΣMsf vs. displacement
curve in Plaxis Curves
Singapore 2011
273 of 607
Notes:
a. Select control point within (expected)
failing body
b. Use sufficient number of load steps
c. Use a sufficiently fine mesh
d. Limit the maximum structural forces by
choosing elasto-plastic behaviour for walls,
anchors and geotextiles.
Safety factor analysis

Number of load steps
1.16 1.16
1.12 1.12
Sum-Msf
Sum-Msf
1.08 1.08
1.04 1.04
1.0 1.0
0.0 0.3 0.6 0.9 1.2 1.5 0.0 0.3 0.6 0.9 1.2
displacement displacement
Singapore 2011
274 of 607
Use different plots to check failure mechanism
1. Arrows of incremental 2. Shadings of incremental 3. Shadings of incremental

displacements displacements shear strains
EXAMPLE 1: UNDRAINED STABILITY OF A SLOPE:
2
1
H = 12m
D cu = 50 kPa
cu = 100 kPa Plaxis:
F = 1,35
cu 50 D (Taylor,1948)
Stability charts: F  N0  6.6  1.38 , N0  f ( , )
Pd 12  20 H
Singapore 2011
275 of 607
EXAMPLE 2: HOMOGENOUS SLOPE WITH AND WITHOUT FOUNDATION
LAYER
Model Slope G1 (753 elements-15n) :

FOS= 1.347 (1.4001, 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
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
cu = 60 kPa
Singapore 2011
276 of 607
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 (1.471, 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.Taylor (1937) (green line)
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(1.051) FOS= 0.470 (0.591, 1.304)
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. Always use the arc-length time stepping procedure within the Phi-C reduction
(default)
5. Use a small tolerated error (maximum should be the default error of 1% )
Singapore 2011
277 of 607
Appendix : 2D vs. 3D (Benz)
1. 3D analysis may yield substantially different results from 2D analysis
2. The advantage of FEM over classical design tools is obvious. I
3. n the example: stability of a bentonite slurry trench.
Plane strain (14 m) 3D (40 m) Axis symmetric (40 m)

FOS = 1.17 FOS = 1.90 FOS= 2.69
REFERENCES
1. Matsui, T. & San, K.C. (1992) Finite element slope stability analysis by shear strength reduction technique. Soils and
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. Ugai, K.(1989). A method of calculation of total safety factor of slope by elasto-plastic FEM. Soils and Foundation 29(2).
pp.190-195.
4. Farias, M.M., Naylor, D.J.(1998). Safety analysis using finite elements. Computer and Geotechnics.Vol 22(2) pp 165-181.
5. Griffiths, D.V., Lane, P.A. (1999). Slope stability analysis by finite elements. Geotechnique 49 (3), pp.387-403.
Singapore 2011
278 of 607
Introduction to Plaxis 3D
Ronald Brinkgreve
Plaxis bv
Plaxis 3D Input
General toolbar
Mode switches
Selection explorer
Drawing area
Model explorer
Mode toolbar
Command line
Singapore 2011
279 of 607
Plaxis 3D Input : Modes
Definition of soil stratigraphy Definition of structural

elements, loads
and boundary conditions
SOIL STRUCTURES
Creation of the FE mesh Definition of pressure Definition of construction stages

distribution
MESH WATER LEVELS STAGED CONSTRUCTION
Model and Selection explorer

• The Model Explorer provides a graphical overview of
the complete model and the objects that it contains.
• The Selection Explorer provides the same
functionality as the Model Explorer, but only for the
current selection of objects
• For managing any objects created in the model:

– shows number of materials, loads
– Showing, hiding or deleting model items
– Renaming model items
– Changing properties of model items
(load values, water height, material sets, …)
Singapore 2011
280 of 607
Command line
• All the actions carried out using either the mouse or the explorers are
translated into commands.
• Alternatively, PLAXIS 3D allows to carry out actions using keyboard
input by directly typing the corresponding commands in the command
line.
– The Session tab displays the commands executed in the active session
– The Model history tab displays all the commands executed in the project
Other functionality
• Selection by either clicking individual objects or at once by defining a selection
box in the draw area. Criteria can be applied to the type of items to be
selected.
• Group creation for fast model creation when the same operations have to be
undertaken over a large number of objects
Singapore 2011
281 of 607
Soil Mode
Soil mode
Borehole 4
• Definition of soil volumes and initial water levels Borehole 3
• Based on the concept of boreholes Borehole 1
• Offers import geometry facilities Borehole 2

• Boreholes
– Locations in the draw area where the
information on soil layering and location of
the water table is specified
– For multiple boreholes, PLAXIS 3D will
automatically interpolate the soil layers
between boreholes
– Each defined soil layer is used throughout
the whole model.
Singapore 2011
282 of 607
Borehole definition
• Defining soil layer heights in the Soil
layers tabsheet
• Defining water conditions in the
Water tabsheet:
– Specific Head
– Hydrostatic distribution,
– Interpolate from adjacent layers
– Dry
– User-defined pore pressures
• Defining Initial Soil conditions in the
Initial conditions tabsheet
– Specify OCR, POP, K0x and K0y for
the K0 procedure
Material Sets
• Definition of material sets and

parameters for representing the soil
layers and structural elements
• Same method of input as Plaxis 2D
Singapore 2011
283 of 607
Structure Mode
Defining the geometry

• Points
– Generated by clicking the Create point button.
– Used for Point loads, Point prescribed displacements and Fixed-end
anchors.
• Lines
– Generated by clicking the Create line button.
– Used to define Beams, Line loads, Line prescribed displacements, Node-to-
node anchors and Embedded piles.
• Surfaces
– Generated by clicking the Create surface button.
– Used to define Plates, Geotextiles, Interfaces and Surface loads.
– The first three created points define the surface plane by default.
– Existing surfaces can be edited from the Surface points dialog box of from
the pop-up submenu of Create surface button
Singapore 2011
284 of 607
Defining the geometry: Array
• Generate multiple copies of a selection, arranged in a rectangular pattern
Defining the geometry: Extrude

• Extrusion of Lines and surfaces in order to create surfaces and volumes
respectively:
– From the Extrude dialog box
– By dragging and dropping the bottom surface to the top surface location
Singapore 2011
285 of 607
Defining the geometry: Other functions
• Decompose into surfaces: creates outer surfaces of selected volumes
• Decompose into outlines: creates contour of selected surfaces
• Intersect: splits selected geometric objects along their intersection
• Combine: merges selected geometric objects of the same kind
• These functionalities are only accessible from the RMB context menu oin
the draw area
Loads
• Generated by clicking the Create load button or by right-clicking on
any geometric objects in the draw area
Point load
Line load
Surface load
→ In a similar way it is possible create Prescribed displacements.
Singapore 2011
286 of 607
Default Boundary Conditions
• Default boundary conditions are

– All displacements fixed for the
bottom surface of the soil contour
– Perpendicular displacement fixed
for lateral surfaces of the soil
contour
→ Default boundary conditions can be

overwritten by specifying
Prescribed displacement surfaces
on the soil contour
Structures
• Generated by clicking the Structure Fixed-end anchor

button
Beam
• Can also be created by right-clicking on
selected: Node-to-node anchor
– points Embedded pile
(fixed-end anchor)
– lines Plate
(beam, node-to-node anchor or Geogrid
embedded pile)
– surfaces Positive interface
(plate, geogrid, interfaces) Negative interface
Singapore 2011
287 of 607
Importing Geometry
• Possibility to import from external sources in different formats like
– 3D Studio files (*.3DS)
– AutoCAD native (*.DWG)
– Interchange (*.DXF) file format
– Click to import surface
– Click to import volume
The Mesh Mode
Singapore 2011
288 of 607
Mesh density in Plaxis 3D
• Global coarseness:
– Defines an average element size based on model dimensions and relative
element size factor Re (Very coarse / Coarse / Medium / Fine / Very Fine)
– Reference element size = 0.05 * Re * (Model diagonal length)
• Local refinement (Fineness factor):
– Element size can be locally refined or coarsened
– Element size = (Fineness factor) * (Global coarseness) * (Ref. elem. size)
• Color code
– Depending on their degree of local refinement, geometric objects are
displayed in different color in the Draw area (gray for Fineness factor of 1
and green otherwise with darker colour when getting more refined)
Local mesh refinement
• By clicking the toolbar

buttons Refine mesh or
Coarsen mesh and
selecting the desired
items in the Draw Area
• From the Selection

explorer by entering the
desired fineness factor
value
• By invoking RMB context

menu after item selection
Singapore 2011
289 of 607
Generate Mesh
• Can be done by clicking the

Generate mesh button from the
Mesh Toolbar or the right mouse
button popup menu
• While the mesh is being

generated, the possibilities are
offered to either pause, resume
or cancel the process
The Water Levels Mode
Singapore 2011
290 of 607
Water Levels
• Generated water level is created by specifying a Head in the boreholes (in

the Soil mode) and is the default water level
– A single borehole can be used to create a horizontal water surface that
extends to the model boundaries.
– When multiple boreholes are used, a non-horizontal water surface can be
created by combining the heads in the various boreholes
– Non hydrostatic distribution in the soil may be specified in the Water
tabsheet of the Modify soil layers dialog box
• User water levels are available in the Water levels mode and can be used
as a alternative to the Generated water level.
Water conditions in soil volumes

• Water levels can be specified
for each individual volume
• Available options are:
– Global level (default)
– Custom level
– Head
– User-defined
– Interpolate
– Dry
• Can be done from
– The WaterConditions feature
in the Selection explorer.
– Right-clicking the mouse
Singapore 2011
291 of 607
The Staged Construction Mode
Calculation phase definition
Phase explorer
Change and/or (de)activate objects

per phase by means of the
Model explorer, Selection explorer
or directly in the Draw area
Singapore 2011
292 of 607
The Phase explorer
• For creating and editing the calculation phases
Insert phase Delete phase Define phase settings

Add phase
Calculation type indicator

K K0 procedure (initial phase)
Calculation status indicator
G gravity loading (initial phase)
to be calculated
P plastic
not to be calculated
D plastic drained
calculation successful
C consolidation
calculation failed
S safety
Defining Phase settings
• Calculation phase settings can be

edited:
– Calculation type
– Phase parameters
– Iterative procedure settings
Singapore 2011
293 of 607
Singapore 2011
294 of 607
Working in the Geometry Modes
of Plaxis 3D
William WL Cheang
Plaxis AsiaPac
Introduction
• The Geometry modes of Plaxis 3D comprises the Soil mode and the
Structures mode
• They are meant to fully define the model geometry in terms of:
– Soil stratigraphy
– Structural elements
– Soil structure interfaces
– Loads
– Boundary conditions
• The Geometry modes are indicated using blue tabsheets and precede the
Calculation modes (green tabsheets) when building up a model from
scratch
Singapore 2011 -3D
295 of 607
The Soil Mode
Introduction to the Soil Mode

• Meant to define soil
volumes and initial water
levels
• Based on the concept of
boreholes like in Plaxis 3D
Foundation
• Offers import geometry
facilities
Singapore 2011 -3D
296 of 607
Soil Mode Toolbar
• Adjust Soil Contour
Create Borehole
Import Soil …
Show Materials
Adjust Soil Contour

• The model contour can be adjusted by
Moving an existing points or lines
Inserting a control points
Deleting an existing points
Singapore 2011 -3D
297 of 607
Boreholes Borehole 4
Borehole 3
• Boreholes are locations in the draw
area at which the information on the Borehole 1
height of the constitutive soil layers
and location of the water table is Borehole 2
given
• If multiple boreholes are defined,
PLAXIS 3D will automatically
interpolate between boreholes and
derive the corresponding position
and height of the soil layers from the
available borehole information.
• Each defined soil layer is used
throughout the whole model contour
Creating Boreholes
Singapore 2011 -3D
298 of 607
Defining Water Conditions
• Water conditions can be specified
from the Modify soil layers dialog box
in the Water tabsheet
• Available options are
– Head
– Hydrostatic
– Interpolate
– Dry
– User-defined
Initial Soil Conditions

• Initial soil conditions can be
specified from the Modify soil
layers dialog box in the Initial
conditions tabsheet
• Values specified (OCR, POP, K0x
and K0y) will be using during
initial stage calculation based
on K0 procedure
Singapore 2011 -3D
299 of 607
Importing Top and Bottom Model Surfaces
• Definition of the top and
bottom soil layer surfaces can
be achieved as a result of
surface import operation
• Import formats include 3DS,
ITS, DWG, DWF and SLT
• Feature available for VIP
members only
Importing Soils
• The geometry of the soil
can be imported from
predefined files instead of
using the Borehole tool
• The same import formats
as for importing soil
surfaces are available
(3DS, ITS, DWG, DWF and
SLT)
• Feature also only
available for VIP members
Singapore 2011 -3D
300 of 607
Material Sets
• Definition of material model parameters
for constitutive soil layers and
structural elements
• Dialog box consistent among all Plaxis
products
• Available from many places in Plaxis
3D:
– Present in each model
– Many shortcuts from different dialog
boxes where material sets assignment
is required
• Entry point to the SoilTest facility
The Structure Mode
Singapore 2011 -3D
301 of 607
Introduction to Structure Mode
• Meant to define structural
elements and loading
• Structural objects are

created directly into a 3D
space using mouse and/or
assisting tools
Default Movement Limitation

• Defining 3D geometry using mouse input is cumbersome considering the
fact that the location in the direction perpendicular to the draw view
cannot be precisely defined. By default:
– mouse will enable movement in the – holding <Shift> while moving the
XY plane mouse will enable movement in Z
direction only
Singapore 2011 -3D
302 of 607
Changing Movement Limitation Settings
• Selecting one the six sides view
of the default views gives access
to the Movement limitation dialog
boxes
• The Movement limitation dialog

box can be closed by selecting
the Perspective view option of the
default views.
Structure Mode Toolbar
• Rotate • Points • Loads
Extrude Lines Displacements
Array Surfaces Structures
• Import surface • Import volume
Singapore 2011 -3D
303 of 607
Points
• Can be generated by cilcking the Create point button
• The following items can be assigned to a point
– Point load
– Point prescribed displacement
– Fixed-end anchor
Point load
Point prescribed displacement
Fixed-end anchor
Lines
• Can be generated by clicking the Create line button
• The following items can be assigned to a line
– Beam
– Line load
– Line prescribed displacement
– Node-to-node anchor
– Embedded pile
Singapore 2011 -3D
304 of 607
Surfaces
• Can be generated by clicking the Create surface buttom
• The first three created points define the surface plane by default
• Existing surfaces can be edited from the Surface points dialog box of
from the pop-up submenu of Create surface button
Mode points/lines
Insert points
Delete points
Rotate
• Rotate any selection around a rotation point compared to global axis
• Possibility to directly rotate the selected objects from the draw area
using Euler angles
Singapore 2011 -3D
305 of 607
Extrude
• Lines and surfaces can be extruded to create surfaces and volumes
correspondingly:
– From the Extrude dialog box
– By dragging and dropping the bottom surface to the top surface location
Array
• Generate multiple copies of a selection, arranged in a rectangular pattern
Singapore 2011 -3D
306 of 607
Further Geometrical Operations
• On top of the aforementioned geometrical operations (rotate, extrude and
array), Plaxis 3D also offers
– Decompose into surfaces: create outer surfaces of selected volumes
– Decompose into outlines: create outer lines (including points) of selected
surfaces
– Intersect: Split select geometric objects along their intersection
– Combine: Merge selected geometric objects of the same kind
• These functionalities are only accessible from the RMB context menu oin
the draw area
Loads
• Can be generated by clicking the Create load button or by invoking the
RMB context menu on any geometric objects in the draw area
Point load
Line load
Surface load
Singapore 2011 -3D
307 of 607
Prescribed Displacements
• Can be generated by clicking the Create prescribed displacements
button or by invoking the RMB context menu on any geometric objects
in the draw area is a very similar way as for Load definition
Point prescribed displacement
Line prescribed displacement
Surface prescribed displacement
Default Boundary Conditions

• Default boundary conditions are
– All displacements fixed for the
bottom surface of the soil
contour
– Perpendicular displacement
fixed for lateral surfaces of the
soil contour
• Default boundary conditions
can be overwritten by
specifying Prescribed
displacement surfaces on soil
contour
Singapore 2011 -3D
308 of 607
Structures
• Can be generated by clicking the Fixed-end anchor
Structure button
Beam
• Can also be created from the RMD
context menu after selecting: Node-to-node anchor
– points (fixed-end anchor) Embedded pile
– lines (beam, node-to-node anchor
or embedded pile) Plate
– surfaces (plate, geogrid, interfaces) Geogrid
Positive interface
Negative interface
Importing Geometry
• Possible to import from external sources in different formats like 3D
Studio files (*.3DS), AutoCAD native (*.DWG) and interchange (*.DXF) file
formats:
– Click to import surface
– Click to import volume
Singapore 2011 -3D
309 of 607
Working in the Calculation Modes of Plaxis 3D
William Cheang
Plaxis AsiaPac
Introduction
• The Calculation modes of Plaxis 3D comprises the Mesh mode, the Water
Levels mode and the Staged Construction mode
• They are meant to fully define the model geometry in terms of:
– The finite element mesh
– Changes in water pressure distribution
– Construction phases
– Calculation settings
• The Calculation modes are indicated using green tabsheets and follow the
Geometry modes (blue tabsheets)
Singapore 2011 -3D
310 of 607
The Mesh Mode
Introduction to the Mesh Mode

• Entered once the geometric
modeling is complete
• Meant to define the finite

element mesh used by the
kernel for project calculation
• Fully automated generation of

finite element meshes.
Singapore 2011 -3D
311 of 607
Mesh Density Definition in Plaxis 3D
• Global coarseness:
– Define an average element size based on model dimensions and relative
element size factor Re (Very coarse / Coarse / Medium / Fine / Very Fine)
– Ref. Elem. Size = 0.05 * Re * Model Diagonal Length
• Fineness Factor:
– Element size could be locally refine or coarsen
– Element size = Fineness Factor * Global Coarseness * Ref. Elem. Size
• Color code
– Depending on their Fineness Factor, geometric objects are displayed in
different color in the Draw Area (gray for fineness factor of 1 and green
otherwise with darker color as getting more refined)
Mesh Mode Toolbar
• Refine mesh
Coarsen mesh
Reset local coarseness
Generate mesh
View mesh
Select points for curves
Singapore 2011 -3D
312 of 607
Changing Local Mesh Density
• By clicking the toolbar
buttons Refine mesh or
Coarsen mesh and
selecting the desired
items in the Draw Area
• From the Selection

explorer by entering the
desired fineness factor
value
• By invoking RMB context

menu after item selection
Generate Mesh
• Can be done by clicking the
Generate mesh button from the
Mesh Toolbar or the RMB
context menu
• While the mesh is being

generated, the possibilities are
offered to either pause, resume
or cancel the process
Singapore 2011 -3D
313 of 607
View Mesh and Select Points
• The View mesh and the
Select points for curves will
open the Output Program
where
– mesh could be evaluated
– nodes and/or stress points
could be selected at the
location of which Output
needs to be generated
The Water Levels Mode
Singapore 2011 -3D
314 of 607
Water Levels
• Generated water level is created by specifying a Head in the boreholes (in
the Soil mode) and is the default water level
– A single borehole can be used to create a horizontal water surface that
extends to the model boundaries.
– When multiple boreholes are used, a non-horizontal water surface can be
created by combining the heads in the various boreholes
– Non hydrostatic distribution in the soil may be specified in the Water
tabsheet of theModify soil layers dialog box
• User water levels is available in the Water levels mode and can be defined
as a alternative to Generated water level
Presentation of the Water Level Mode
• Meant to define User

water levels in a similar
way as surface
• Enable preview of
generation of water
pressure
Singapore 2011 -3D
315 of 607
Assigning Water Conditions to Soil Volume
• Water levels can be specified
for each individual volume
• Available options are:
– Global level (default)
– Custom level
– Head
– User-defined
– Interpolate
– Dry
• Can be done from
– The WaterConditions feature
in the Selection explorer.
– The RMB context menu
The Staged Construction Mode
Singapore 2011 -3D
316 of 607
Introduction to the Staged Construction
Mode
• To define relevant
construction stages and
• Launch Plaxis 3D calculation
Staged Construction Toolbar

Activate
Deactivate
Show materials
Preview construction stage
Select points for curves
Calculate
View calculation results
Singapore 2011 -3D
317 of 607
Activate/Deactivate
• Indicates which model parts
should be active or inactive in
the each construction stage
• Can be defined through
– RMB context menu
– The selection or model
explorer
• Selection tool can be
advantageously used for fast
selection of large number of
objects
• Be careful with hidden objects
which might still be active
The Phase Explorer

• For defining and listing the defined phases for a calculation
• Accessible from all Calculation modes but only editable in the Mesh mode
Insert phase Delete phase Define phase settings

Add phase
Calculation type indicator

K K0 procedure (initial phase)
Calculation status indicator
G gravity loading (initial phase)
to be calculated
P plastic
not to be calculated
D plastic drained
calculation successful
C consolidation
calculation failed
S safety
Singapore 2011 -3D
318 of 607
Defining Calculation Stages
• Calculation phase settings can be
edited:
– Phase type
– Phase parameters
• Phase types are:
– Initial stress definition
– Plastic calculation
– Safety factor analysis (phi-c
reduction)
– Consolidation analysis (with closed
or open flow model boundaries)
Phase Parameters
• Advanced phase parameters
can also be controlled from the
Phases dialog box
– Maximum number of
calculation steps
– Reset displacement
– Updated Mesh
– Ignore undrained behaviour
– Parameters for automatic step
size
Singapore 2011 -3D
319 of 607
Singapore 2011 -3D
320 of 607
CG15 Deep Foundation
2D and 3D FEM
Prof Harry Tan

NUS
Outline
• Analysis of Single Pile in Axi-symmetry
– Influence of discretization
– Influence of interface
– Influence of dilatancy
• Case Study – Instrumented Spun Pile at
Butterworth 8 Condo Project
• Pile Raft Analysis
• Cases
– St Thomas Tower – large pile group
– River Gate Tower – Pile-raft jack-in piles
2

3

5

7

9
10

11
12

13
Singapore Case Study

Instrumented Spun Pile at Butterworth 8
tested to Plunging Failure at 2.2 WL
14

Instrumented Pile – Butterworth 8
Pile Instrumentation Load Transfer Results
15
Plaxis Model of Spun Pile
Fill
Upper MC
Inter Stiff Clay

Lower MC
Very Stiff Silty Clay
Hard Silty Clay
Rock
16

Plaxis INPUT
17
Compare Plaxis with Measured

Shaft Friction at Ultimate Load
Skin Friction (kN/m2)
0 50 100 150 200 250
0
-5
-10 MEASURED
PLAXIS
-15
-20
Depth (m )
-25
-30
-35
-40
-45
-50
18

Development of Shaft Friction
At 1*WL = 222 T At 2*WL = 444 T At 2.2*WL = 488 T
Limiting Limiting Limiting

Shear Shear Shear
Strength Strength Strength
Slightly Nearly Fully Fully
Developed Developed Developed
19
Plaxis compared to Load Test

Pile Load Test - PTP1 Butterworth 8
1.00E-02
0.00E+00 Loads (kN)

-1.00E+03 0.00E+00 1.00E+03 2.00E+03 3.00E+03 4.00E+03 5.00E+03 6.00E+03
-1.00E-02
-2.00E-02
S e t t le m e n t ( m )
-3.00E-02
-4.00E-02
-5.00E-02
QT-PLAXIS
-6.00E-02
-7.00E-02
Qb-PLAXIS
-8.00E-02 QT- Load Test
-9.00E-02 Qb-Load Test
-1.00E-01
20

UnDrained and Drained Loading
Total Pile Head Loads
UTP-1
SETTLEMENTS [m]
0
UNDRAINED
-0.05
DRAINED
-0.1
-0.15
-0.2
-0.25
-0.3
-0.35
0 1e4 2e4 3e4 4e4 5e4
TOTAL PILE HEAD LOADS (kPa)
21
UnDrained and Drained Loading

Pile Toe Loads
UTP-1
SETTLEMENTS [m]
0
UNDRAINED
-0.05
DRAINED
-0.1
-0.15
-0.2
-0.25
-0.3
-0.35
0 -10000 -20000 -30000 -40000 -50000
PILE TOE LOADS [kN/m2]
22

Chin Hyperbolic Plot for 2WL
Assume Qcorr = 0.7 Q Chin’s;

Corrected Qult=6113 kN (too
high!!!)
Rule of Thumb that Qult = 0.7 to 0.8 Chin’s Ultimate has
no merit., It can be dangerously WRONG; it is ROT!!! 23
Why Chin’s Plot is Incorrect?
Hyperbolic
Extrapolation
2*WL
Actual Pile
Plunge at
Chin’s Plot 2.2*WL
for 2*WL
24

Conclusions
• Chin’s Plot is Extrapolation Method
• It will grossly over-predict capacity if pile can fail
• FEM analysis is more realistic as it can model
plastic failure of pile when soil resistance is
exceeded
• Undrained loading will give lower pile capacity
• Undrained loading will show plunging failure
• Drained loading of frictional soil shows no limit to
end bearing and total capacity
25
Examples in Germany IGS
RF: Raft Foundation after Katzenbach, Moormann, Reul

PF: Pile Foundation
PRF: Piled Raft Foundation
s: Settlements at end of construction Ćommerzbank II Hochhaus´
1994-97
´MAIN TOWER´ PF, s = 2.1 cm
1996-99
PRF, s = 2.5 cm
´Helaba Éurotower´
Hochhaus´ 1974-77
1975-77 RF, s = 9 cm
RF, ´Japan Center´
ÉUROTHEUM´ s = 10 cm
1997-99 1994-96
Ćitibank´
PRF, PRF, s = 3.2 cm
1985-86
RF, s = 3.2 cm
s = 11 cm Ćommerz
bank I
Hochhaus´
1972-74
RF, s = 9 cm

Single pile vs Pile-Raft Behavior
• Paper by:
27
Single pile versus pile group versus piled raft
Resistance Q
Single pile
Pile group
Pile group Piled raft
Settlement s
After El-Mossalamy 1996

Piled raft foundation
28

Single pile versus piled raft
Resistance Q
C Single pile
I
E
Settlement s
Corner pile
Edge pile
Inner pile
Conclusion:
Group behaviour and pile-raft-interaction
reduce on the one hand the stiffness of
the piles and increase on the other hand
their bearing capacity!
After El-Mossalamy 1996
29
Bearing behaviour of a piled raft
Piled-Raft Foundation (PRF) after Hanisch, Katzenbach, König
Raft Foundation Pile Foundation

DIN 1054
DIN 4017 DIN 1054
DIN 4018 DIN 4014
DIN 4019 DIN 4026
DIN 4128
PRF
 PRF 
Q ( s ) P
Qtot ( s )
sPRF/sRF
30

 Q P + QR
after Hanisch, Katzenbach, König
Qtot = Stot
QP = Q b + Qs
QR =  (x,y) dA QP,1 (x,y) QP,j
Qtot     Stot D
with  = 2,00 for loading case 1

1,75 for loading case 2 Interaction between
Piled raft and subsoil
1,50 for loading case 3 (x,y)
z
Interaction influences:
 Pile-soil interaction
 Pile-Pile interaction s,j
 Raft-soil interaction s,1

e e
b,j
 Pile-raft interaction
31
Influence of pile-raft interaction
•For isolated single pile,

note progressive skin
friction failure with
increasing settlements
•For pile-raft, at larger
settlements, get larger
skin friction in upper
part of pile due to pile-
raft interaction
32

Influence of pile-raft interaction
•Pile leads to significant

decrease in raft contact
pressure next to pile
shaft compared to raft
only without pile
33
Influence of pile-raft and pile-pile interaction

• Example out of the literature by Katzenbach, Arslan,
Moormann
– Piles: diameter 1.5m
length 30m
– Model 1: quadratic raft with a width of 50m
64 piles with a distance of 3D
– Model 2: quadratic raft with a width of 50m
16 piles with a distance of 6D
Model 1 Model 2
I E
34

D = 1.5m
Load Q [MN]
0 10 20 30 40
Q single pile
L = 30m 0
Q inner pile
64 piles
e = 3D 5
Q edge pile
Q corner pile
10
Settlement [cm]
15
20
25
30
35
D = 1.5m
Load Q [MN]
0 10 20 30 40
Q single pile
L = 30m 0
Q inner pile M1
64 piles
e = 3D 5
Q inner pile M2
16 piles Q edge pile M2

e = 6D
10
Q corner pile M2
Settlement [cm]
15
20
25
30
36

37
Conclusions
• Bearing resistance failure will not occur until well above
WL
• Load sharing between pile and raft depends on
settlement. No linear relation between load sharing and
settlement
• Load in pile depend of pile position and spacing
• Pile-raft interaction leads to increasing vertical and
horizontal stresses in soils
• Increased load in piled-raft will mobilize larger skin
friction in upper part of piles
• Assumption of piles as an action on the raft of constant
value is unrealistic and lead to poor design
38

German Case History
39
Concept of Piled-Raft
40

Concept of Piled-Raft
Piled-Raft
41
Typical Frankfurt Clay
42

Examples in Frankfurt
RF: Raft Foundation after Katzenbach, Moormann, Reul

PF: Pile Foundation
PRF: Piled Raft Foundation
s: Settlements at end of construction Ćommerzbank II Hochhaus´
1994-97
´MAIN TOWER´ PF, s = 2.1 cm
1996-99
PRF, s = 2.5 cm
´Helaba Éurotower´
Hochhaus´ 1974-77
1975-77 RF, s = 9 cm
RF, ´Japan Center´
ÉUROTHEUM´ s = 10 cm
1997-99 1994-96
Ćitibank´
PRF, PRF, s = 3.2 cm
1985-86
RF, s = 3.2 cm
s = 11 cm Ćommerz
bank I
Hochhaus´
1972-74
RF, s = 9 cm
43
Frankfurt Cases
44

Frankfurt Cases
45
Case 1 Messe-Torhaus
• First case in Frankfurt of
piled-raft during 1982-84
• 400 MN,130m high, 30
storeys
• Two rafts 24.5x17.5m and
2.5m thick on 42 bored
piles each
• Piles 0.9m D and 20m long
• Without piles settlement
about 250 mm; with piles
reduce to 120 mm
46

47
80%
20%
48

• The results showed that design
assumption that ultimate pile resistance
of single piles is same for piled-raft piles
used as settlement reducers is
INCORRECT!!!
• There is a need for correct modeling of
piled-rafts to account for interactions
with aim of limiting settlement of
buildings, using fewer piles of greater
lengths
• This will result is better ratio of pile/raft
load share than 80:20
49
Case 1 Messe-Torhaus Piled-Raft
Messe-Torhaus after Katzenbach, Moormann, Reul
1982-1984
h = 130 m
30 floors
Normal force [MN] Shaft friction [kN/m2]
Inner pile
Depth [m] Depth [m] 50
Edge pile

Case 2 Messeturm Piled-Raft
after Katzenbach, Moormann, Reul
Messeturm
• Built in 1991
• Height = 256 m; 60 storeys
• Weight = 1900 MN
• Uplift = 200 MN
• Raft = 41x60m
• Raft alone estimate settle 400 mm
• Piled-raft estimate settle 200 mm
• Raft thickness = 6m centre and 3m edge
• 64 bored piles in 3 rings
51
52

• Load sharing
• Pile = 55%
• Raft = 45%
53
Good Design Strategy

 Piles spacing as large as
possible, and in 3 rings to
avoid group action as far as
possible
 Piles concentrate under
main loadings
 Outer rings diameter
chosen to minimise possible
tilt
 Different pile lengths to
equalise pile loads 54

Case 2 Messeturm Piled-Raft
Messeturm after Katzenbach, Moormann, Reul
1991
h = 256 m
60 floors
55
Singapore Case 1: St Thomas Walk

• Introduction
• Pile Load Test Model
• Foundation 3D Model
• Model Parameters
• Simulation Steps
• Simulation Results
• Summary & Conclusions
56

Introduction
• Plan of piled raft foundation
BH3
BH5
BH1
BH4
BH2
BH6
Main piled raft + 4 independent pilecaps

57
Introduction
• Aim of 3D study
– to predict total and differential settlements of piled
raft foundation
– to obtain likely raft forces and bending moments
• Pile load test model to obtain strength and stiffness
correlations with available SPT N-values
• Features in 3D model:
– Non-uniform soil profile by input of multiple
boreholes
– Unloading-reloading soil behaviour simulated
realistically by Hardening Soil models
– Large number of piles modelled using embedded
pile elements in Plaxis 3D Foundation
58

Pile Load Test Model
• Static load test by Cambrian Eng. Corp.
(June 2007) on 900-mm diameter pile with
44-m toe depth
• Soil profile estimated from nearby BH 2
BH5
BH1
BH4
BH2
BH6
Approx. test location

59

• Model geometry
Firm silty clay (N=9)
-9 m
Stiff clayey silt (N=14)
-21 m
V. stiff clayey silt (N=30)

-29 m Solid pile with
Hard clayey silt (N=100) interface elements
-33 m Pile properties:
E = 30 GPa; v= 0.2
Hard clayey silt (N=50)

-44 m
-50 m
60

• Simulated test up to 1×WL
• Top two layers (N=9 & N=14) modelled with
Hardening Soil (HS), lower layers with Mohr-
Coulomb (MC) models
• Back-calculated soil parameters based on N-values
(from trial simulations and curve-fitting):
– For HS models: Eoedref = 1.5N; E50ref = 3N; Eurref =
9N (all in MPa)
– For MC models: cref = 5N (φ = 0); Eref = 3N (in
MPa)
– For all models: interface factor Rinter = 0.8
– Above correlations within reasonable range for
OC silt
61

• Hence, HS parameters: MC parameters:
Firm silty Stiff clayey Very stiff Hard Hard
clay (N=9) silt (N=14) clayey clayey clayey
Type Undrained Undrained silt silt silt
(N=30) (N=50) (N=100)
γsat [kN/m3] 19 20
Type Undraine Undraine Undraine
E50ref [kN/m2] 27,000 42,000 d d d
Eoedref [kN/m2] 13,500 21,000 γsat 20 21 21
Eurref [kN/m2] 81,000 126,000 [kN/m3]
cref [kN/m2] 30 20 ν 0.3 0.3 0.3
φ [°] 28 28 Eref 90,000 150,000 300,000
[kN/m2]
ψ [°] 0 0
cref 150 250 500
νur 0.2 0.2
[kN/m2]
pref [kN/m2] 100 100
φ [°] 0 0 0
power (m) 0.8 0.8
ψ [°] 0 0 0
K0nc 0.531 0.531
Rinter 0.8 0.8 0.8
Rf 0.9 0.9
62
Rinter 0.8 0.8

• Comparison between simulated and field
settlements up to 1×WL:
• Linear settlement profiles show
Load (tonne) elastic soil behaviour up to 1×WL
0 100 200 300 400 500
0 • Field settlement = 4.5 mm at 1×WL
1
• Simulated settlement = 7 mm at
2 1×WL, overestimating field data
3 • Field soil seems stiffer than assumed
Settlement (mm)
4
parameters based on SPT N-values
5 Field test data • Hence use of assumed parameters
6
Simulated
tend to be conservative in settlement
predictions
7
63

• Comparison between simulated and field
skin friction at 1×WL :
• Simulated skin frictions reasonably
Skin friction (kPa) agree with field data
0 10 20 30 40 50 60 70
0 • A jump in skin frictions in depths of
5 30–35 m due to presence of thin hard
10 layer (N=100)
15
• Hence assumed soil strength
20
parameters and Rinter reasonably
Depth (m)
25
represents field soil properties
30
35
40 Field test data

Simulated
45
50
64

Foundation 3D Model
BH3
BH5
BH1
BH4
BH2
BH6
• Model includes main piled raft & 4 pilecaps; other

individual piles not modelled
65
Foundation 3D Model
BH3
BH5
BH1
BH4
BH2
BH6
• Non-uniform soil profile based on BH 1 – BH 6

66

Foundation 3D Model
• Modelled soil layers for BH 1 – BH 6 based on Soil
Report by Geospecs Pte. Ltd. (Jan 2007):
Depth of soil layer below ground surface RL122.2m
Soil layer (m)
BH 1 BH 2 BH 3 BH 4 BH 5 BH 6
Firm silty clay
0–3 0–9 0–3 0–5 0–3 0 – 12
(N=9)
Stiff clayey silt
3–9 9 – 21 3 – 12 5 – 17 3 – 28 12 – 16
(N=14)
Very stiff
clayey silt 9 – 20 21 – 29 12 – 17 17 – 18 28 – 30 16 – 18
(N=30)
Hard clayey
20 – 21 29 – 50 17 – 30 18 – 30 30 – 48 18 – 48
silt (N=50)
Hard clayey
> 21 > 50 > 30 > 30 > 48 > 48
silt (N=100) 67
Foundation 3D Model
Plate elements for raft & pilecaps,
with loading shown
Soil behind retaining wall

above excavation level is
represented by surcharge
Firm silty clay (N=9)

Stiff clayey silt (N=14)
Very stiff clayey silt (N=30) Complex non-uniform soil

profile based on 6 BHs
68

Foundation 3D Model
Embedded piles
with varying toe
depths based on
piling plan
69
3D Model in SPT>100 Geo-materials
70

3D Model in 50<SPT<100 Geo-materials
71
Foundation 3D Model
Line loads for column and shear wall loads

(arrow length denotes load magnitude)
72

Model Parameters
• Structural parameters
– Raft plate parameters (5 types of thickness):
Thickness, d (m) 1.0, 1.3, 1.5, 2.3, 2.5
γ [kN/m3] 24
E [kN/m2] 28×106
ν 0.2
– Embedded pile parameters (3 types of diameter):

Diameter (m) 1.0, 1.3, 1.4
γ [kN/m3] 24
E [kN/m2] 28×106
Max skin friction Layer dependant
Max toe resistance,
1,000* 73
Fmax [kN]
Simulation Steps
• Phase 0: Initial stress generation by K0 procedure
– K0 = 0.7 for top firm silty clay layer (over-
consolidated) & K0 = 0.531 for others (normally-
consolidated)
• Phase 1: Installation of bored (embedded) piles
– Undrained (short-term) simulation
• Reset displacements to zero
• Phase 2: Excavation to pile C.O.L (RL114.3m)
– Undrained excavation throughout model and addition
of surcharge on active side of retaining wall
(simplified)
• Reset displacements to zero
• Phase 3: Installation of rafts and foundation loads
– Drained (long-term) simulation 74

Simulation Results
• Short-term ground heave after excavation
Lower ground heave in raft
areas due to pile holding-down
Max ground heave = 10 mm
Ground heaving is realistically low due to simulation of

unloading-reloading behaviour in HS model
75
Simulation Results
• Long-term ground settlements under loading
76

Simulation Results
• Long-term raft settlements under loading
Max raft settlement = 28

mm
Max raft differential settlement = 16 mm
Column-to-column angular tilt << 1/500
77
Simulation Results
• Long-term ground effective stress under
loading
Effective vertical stress

beneath piled raft between 80
– 110 kPa
Area of piled raft = 310 m2; Total load on piled raft = 230,000 kN;
thus soil carries up to 15% of total load (=> pile design
adequate)
78

Simulation Results
• Raft shear force Q13
Max -Q13 = 4,560

kN/m
Max +Q13 = 7,600

kN/m
79
Simulation Results
• Raft shear force Q23
Max +Q23 = 4,100 kN/m

Max -Q23 = 4,250 kN/m
80

Simulation Results
• Raft bending moment M11
Max -M11 = 4,870 kNm/m
Max +M11 = 870 kNm/m
81
Simulation Results
• Raft bending moment M22
Max -M22 = 6,120 kNm/m

Max +M11 = 1,090 kNm/m
82

Pile no Max axial force (kN) Pile diameter (m) Max axial stress (kPa) Note
1 3,711 1.0 4,725 < 7.5 Mpa
2 3,668 1.0 4,670 < 7.5 Mpa
3 3,108 1.0 3,957 < 7.5 Mpa
4 6,814 1.4 4,426 < 7.5 Mpa
5 8,879 1.4 5,768 < 7.5 Mpa
6 8,809 1.4 5,722 < 7.5 Mpa
7 7,378 1.4 4,793 < 7.5 Mpa
8 7,119 1.4 4,625 < 7.5 Mpa
9 7,343 1.3 5,532 < 7.5 Mpa
10 6,425 1.3 4,841 < 7.5 Mpa
11 7,469 1.3 5,627 < 7.5 Mpa
45
46
12 6,515 1.3 4,908 < 7.5 Mpa
48
13 7,100 1.4 4,612 < 7.5 Mpa
14 9,295 1.4 6,038 < 7.5 Mpa
49
15 9,588 1.4 6,228 < 7.5 Mpa
44
47
16 8,965 1.4 5,824 < 7.5 Mpa
52
41
43 17 6,636 1.4 4,311 < 7.5 Mpa

18 6,884 1.4 4,472 < 7.5 Mpa
40
42
50
19 8,296 1.4 5,389 < 7.5 Mpa
39
51
20 9,618 1.4 6,248 < 7.5 Mpa
21 9,441 1.4 6,133 < 7.5 Mpa
22 8,429 1.4 5,476 < 7.5 Mpa
33
23
23 7,096 1.4 4,610 < 7.5 Mpa

17
38
24 8,442 1.4 5,484 < 7.5 Mpa
25 8,711 1.4 5,659 < 7.5 Mpa
32
27
22
26 8,776 1.4 5,701 < 7.5 Mpa

16
37
27 8,649 1.4 5,618 < 7.5 Mpa

28 7,693 1.4 4,997 < 7.5 Mpa
31
21
26
29 7,967 1.4 5,175 < 7.5 Mpa

36
15
30 9,095 1.4 5,908 < 7.5 Mpa

31 9,029 1.4 5,865 < 7.5 Mpa
25
30
20
32 8,492 1.4 5,517 < 7.5 Mpa

14
35
33 9,760 1.4 6,340 < 7.5 Mpa

34 7,802 1.4 5,068 < 7.5 Mpa
19
29
24
35 7,226 1.4 4,694 < 7.5 Mpa

13
34
36 7,372 1.4 4,789 < 7.5 Mpa

18
37 7,691 1.4 4,996 < 7.5 Mpa

28
38 9,343 1.4 6,069 < 7.5 Mpa

39 6,396 1.4 4,155 < 7.5 Mpa
40 8,764 1.4 5,693 < 7.5 Mpa
12
41 8,782 1.4 5,705 < 7.5 Mpa

8
4
10
42 7,212 1.4 4,685 < 7.5 Mpa

7
5
43 6,940 1.4 4,508 < 7.5 Mpa

11
44 3,335 1.0 4,246 < 7.5 Mpa

9
1
45 3,338 1.0 4,250 < 7.5 Mpa

46 3,201 1.0 4,076 < 7.5 Mpa
47 3,660 1.0 4,660 < 7.5 Mpa
2
48 3,921 1.0 4,992 < 7.5 Mpa

49 2,906 1.0 3,700 < 7.5 Mpa
50
51
6,597
5,885
1.3
1.3
4,970
4,434
83
< 7.5 Mpa
< 7.5 Mpa
52 6,244 1.3 4,704 < 7.5 Mpa
Summary & Conclusions

Max raft settlement 28 mm
Raft angular tilt (column-to-
<< 1/500
column)
Max raft shear forces Q13 7600 kN/m
Max raft shear forces Q23 4250 kN/m
4870
Max raft bending moment M11
kNm/m
6120
Max raft bending moment M22
kNm/m
• Soil parameters from simulation of pile load test

– Stiffness parameters tend to be conservative as they give
higher settlement under 1×WL than test pile
– Predicted raft settlements hence tend to be conservative
• 3D model uses 6 BHs for varied soil profile and HS model for
84
realistic unload-reloading soil behaviour

Case 2: Pile-Raft Study of 43 Storey River-Gate Tower
Pile load test Plaxis Axi-symmetric Model
85
Site Soil Profile Calibration of single pile load test
Block Model of Pile-Raft

Measured
Settlements
Composite reinforced soil Settlements Distribution

blocks to model piles 86

Embedded Piles in Pile-Raft Model (3DF Program)
87
CONCLUSIONS of River-Gate Pile-Raft Study

i. Pile-raft settlement predictions can be carried out through soil-pile
block model and embedded piles model. Both models can be
found to be comparable in settlement as shown in this case study.
However their results are still not close to the measured settlement
due to the difficulty to determine the actual undulating geological
formation. Erratic sub-soils condition has been encountered on site
in view of the presence of some very short piles which were not
able to be installed to the design piles penetration.
ii. Prediction of pile-raft settlement is very much dependent on

geological formation of the hard stratum from which the piles’ toes
are important to be properly socketed /supported.
iii. Stiffness of hard stratum governs the prediction of pile-raft

settlement. Accurate assessment of such stiffness can lead to
more accurate prediction of pile-raft settlement.
iv. Jack-in spun piles are generally safe even with short piles so long
they have been jacked to refusal with two times of working load of
88
piles

Favorable and Unfavorable Soils
for Piled Raft
Favorable : soil profiles with stiff clays or
dense sands
Unfavorable:
• Soft clays or loose sands near surface
• Soft compressible layers at relatively
shallow depths
• Soil likely to undergo consolidation or
swelling due to external causes
89
Points for practical design (Harry Poulos)

• More piles do not lead to good design
• Raft thickness affects differential settlements

and BMs, but little effect on load sharing or
maximum settlements
• For control of diff sett, best achieved by small

number of piles at strategic location rather than
large number of evenly distributed piles, or
increasing raft thickness
• Nature of applied loading is important for diff sett

and BMs, but less important for maximum sett or
load sharing ratio
90

Guidelines for economic design
(Hirokoshi and Randolph)
• For uniform loading, piles should be distributed
over the central 16-25% of raft area
• For concentrated loading, piles should be focus
under such loaded areas
• The total pile capacity should be designed for
between 40-70% of design load
• The degree of mobilization of pile capacity
should not exceed 80% to avoid significant
increase of differential settlements
91

Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
William WL Cheang
Principal Geotechnical Consultant
Plaxis 3D Plaxis AsiaPac
Lee Siew Wei
Modelling of Tunnels and Tunnelling

Associate Director
Geotechnical Consulting Group (Asia)
Contributions
Dr And Pickles
Mr Tom Henderson
Ms Carmen Choi
Mr Tse Sai Chung
Dr Ronald Brinkgreve
Mr Frank van Gool
Ms Wendy Swolls
Contents
1. Part 1- Modelling of Tunnels in Plaxis 3D

a. Input and construction of FE model
b. Conclusions
2. Part 2- Modelling of Tunnels in 2D or 3D
a. Modelling of Tunnels in 2D (Methods available)
b. Modelling of Earth Pressure Balance and Slurry Shields
c. Conclusions
3. Part 3- Cases
a. Tunnelling 1
b. Tunnelling 2
c. Tunnelling 3
Singapore Advanced Course 2011
367 of 607
Modelling of Tunnels and Tunnelling in 3D
1. Geometric modelling issues (CAD, Import or CM Line)
2. Construction stages
3. Modelling anchors
4. Modelling volume loss
5. Conclusions
Geometric modelling issues
Circular tunnel shapes (TBM tunnels) – Example
368 of 607
Geometric modelling issues (CM Line)
Circular tunnel shapes (TBM tunnels)
• Create cylinder using Cylinder command or using Import facility

cylinder 4 100 48
• Decompose cylinder volume into surfaces
• Apply plate and negative interface features to cylinder contour
Geometric modelling issues
Cross passages and entrance shafts – Example
Hint: Draw cross section surface and use Extrude command to create shafts
PLAXIS 3D will automatically create intersections
369 of 607
Geometric modelling issues (CAD)
Non-circular tunnel shapes

1. Using shape designer* to create
tunnel shape
2. Decompose tunnel volume into
surfaces
3. Assign Plate and interfaces features
to tunnel surface
* new in 3D 2011
Geometric modelling issues (Import)
Importing tunnel geometry using CAD model

• DXF triangulated surface model
- Model should be ‘cleaned’ before importing in PLAXIS 3D
• 3DS model
• Use Import command or corresponding tool in Structures mode
370 of 607
Construction stages (For sequence of events)
Creating geometry for construction stages

• Divide tunnel in excavation sections (top heading, bench, invert)
• Divide tunnel in longitudinal steps by defining cross section planes
• Intersect tunnel with excavation sections and cross section planes
• Remove unnecessary sub-surfaces around tunnel
Creating geometry for construction stages – Example (exploded view)
Modelling Anchors
Create anchors in the following way:

• Use Lineangles option to create end points for anchors, e.g.
• Repeat for different angles
• Remove lines but keep end points
• Use end points with Lineangles command to create new lines snapping onto
tunnel volume
Lineangles (x y z) 0 45 15
• Assign Beam feature to turn lines into anchors
• Alternatively, embedded piles can be used
Lineangles point_2 0 225 volume_1
371 of 607
Modelling Anchors
Example showing anchors and partial excavation
Modelling volume loss
1. Volume loss can be modelled by:

a. Defining Contraction in
Structures mode, or use
contraction tool or right-hand
mouse menu
b. Activate contraction in
Contraction Phase_Volume_1_1
Staged construction mode
2. Alternatively, volume loss can be

modelled by:
a. Applying Volumetric strain to
volume (Staged construction
mode)
b. Distinction and specific strain
components can be made
between xx, yy, zz
372 of 607
Conclusions
• PLAXIS 3D contains several features to model tunnels:

- TBM tunnels
- NATM tunnels
- Tunnel lining
- Anchors
- Construction stages
- Volume loss
- Deformation analysis
- Stability analysis (e.g. tunnel heading)
• Automatic intersection of objects
2D & 3D MODELLING OF
TUNNELLING
Part 2
373 of 607
Part 2‐Outlines
A. 2D modelling of tunnelling
1. influence of soil constitutive models
2. different methods modelling tunnel excavation
B. 3D modelling of tunnelling
1. background on pile response
2. progressive advance of tunnel face
3. response of piles & building to tunnelling
Tunnelling Observations
1. Tunnelling case histories in Hong Kong observed

• 1. Greenfield surface settlement profile fitted by Gaussian curve
with trough width parameter (K) of 0.5 in layered ground
• 2. With good workmanship achieved greenfield volume (or ground)
loss ratios were less than 1%
2. GCO (1985), Storry et al. (2001), Storry et al. (2003) & Hake & Chau
(2008)
374 of 607
Concept of Modelling Tunnelling in 2D
3D 2D 2D
Moller (2006)
• 3D arching around unsupported tunnel heading carries vertical load Pg by

transferring them around unsupported cut stretch
• 2D analysis cannot model 3D arching effect - this is compensated by including
an artificial support pressure Ps (can be a pressure- or displacement-controlled
approach)
Methods of Modelling Tunnelling in 2D
• Plaxis 2D provides
1. Lining Contraction Method
2. Stress Reduction Method (-method)
3. Applied Pressure Method (APM)
(from Grout Pressure Method by Moller & Vermeer, 2008)
375 of 607
Lining Contraction Method
1st Phase 2nd Phase
Vermeer & Brinkgreve

(1993)
Moller (2006)
Initial position
• 1St Phase: Lining is wished-in-place, soil elements inside tunnel deactivated –

tunnel heaves
• 2nd Phase: Lining is stepwise contracted until prescribed contraction % – radial
displacement towards tunnel center
• Tend to give unrealistic results for ground surface settlement & horizontal displacement
Stress Reduction Method ()
1 Pk 1
Pk = initial ground radial
= pressure
Pk ΣMstage = 1 - 
• 1St Phase: Soil elements inside tunnel deactivated,

2
internal support pressure = pk, net load acting on
unsupported perimeter = (1-)pk
• 2nd Phase: Lining activated, remove internal support
pressure & lining takes remaining load pk
Lining
 is Load Reduction factor, obtained from tunnelling
experience. Tend to give reasonable results
376 of 607
Applied Pressure Method (APM)
A. Based on Grout Pressure (Moller &

Vermeer, 2008)
B. Applied Pressure Methods vs beta-
method, difference is the profile of
internal support pressure
• 1st Phase: Soil elements inside tunnel deactivated, internal support

pressure manually specified which is Pcrown at tunnel crown, rate of
increasing with depth = grout (e.g. ~15 kN/m3)
• 2nd Phase: Lining activated & remove internal support pressure
FE Prediction of Greenfield Surface Settlement
1. Numerical analysis with simple constitutive model (LEPP) cannot replicate

measured greenfield (G/F) surface settlement curve
2. FE prediction improved by
1. Refining method of modelling tunnel excavation (TBM model)
2. Using advanced soil constitutive model
3. An exercise to investigate effects of these two factors
4. Case histories showed G/F surface settlement could be reasonably fitted with
Gaussian curve (reference check)
377 of 607
Modelling of Tunnelling in Hong Kong Soils
60m
0
Fill -3
Marine Deposits -6
20m
40m
Completely Decomposed Granite
6m Ø tunnel (CDG)
-
Rock 40
• Ground conditions: 3m Fill, 3m MD, 34m CDG & rock; GWT at surface
• Tunnel 6m diameter with axis at 20 mbgl; 2700 nos of 15-noded elements
Soils Modelled by Mohr Coulomb Model
 E ν c' / cu '
Soil
(kN/m3) (MPa) [-] (kPa) (Deg)
Fill 19 20 0.3 0 30
MD
16 6 0.3 15 0
(Undrained)
CDG 20 39 0.3 5 35
Soils Modelled by HS & HS‐small Models
E50ref & Pref G0

Soil Eurref m c' ' νur 0.7
Eoedref (kP (MP
(MPa) [-] (kPa) (Deg) [-] [-]
(MPa) a) a)
Fill
20 60 0.5 0 30 100 0.2 - -
(HS)
MD
(HS) 6 18 1 0 22 100 0.2 - -
(Und.)
CDG
5E
(HS- 39 117 0.5 5 35 200 0.2 200
-5
small)
378 of 607
Pre‐failure Stress‐strain Behaviour
1: Mohr Coulomb
1: Linear elastic, perfectly plastic

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
Initial Stress Equilibrium
• K0 = 1 – sin'
' = drained effective friction angle (Fill=30°; MD=22°)
Soil K0 CDG
Fill 0.5
MD 0.625
Schnaid et al.
(2000)
CDG 0.65 0.4 0.65 0.9
379 of 607
Details for Analyses
1. FE series 1-Mohr Coulomb

(i) Lining Contraction Method
(ii) Stress Reduction () Method
(iii) Applied Pressure Method - APM
2. FE Series 2-HS & HS-small
(i) Lining Contraction Method
(ii) Stress Reduction () Method
(iii) Applied Pressure Method - APM
• Compare greenfield surface settlement curves with a ground loss ratio (VL) of
1%.
Distance from tunnel centreline (m)

0 10 20 30 40 50 60
0 Results for Mohr Coulomb Soil Analyses
-2
Settlement (mm)
-4
Gaussian (K=0.5, VL 1%)
-6
Lining contraction - LC 1%, VL 0.32%
-8 Lining contraction - LC 1.7%, VL 1%
-10 Stress reduction - beta 0.68, VL 1%
-12 Applied pressure - Pcr 190 kPa, VL 1%
Mohr Coulomb Soil with Lining Contraction
380 of 607
Comparison of MC and HS & HS‐small Models

0 10 20 30 40 50 60
0 Mohr Coulomb
-2
Settlement (mm)
-4
-6
-8 Lining contraction - LC 1.7%, VL 1%
-10 Stress reduction - beta 0.68, VL 1%

0 10 20 30 40 50 60
0
-2
Settlement (mm)
-4
-6
-8 Lining contraction - LC 1.22 %, VL 1%
-10 Stress reduction - beta 0.66 , VL 1% HS & HS-small

Stress Reduction Stress Reduction vs. Applied Pressure

Applied Pressure Methods
HS & HS-small analyses, Greenfield

VL 1%
Exaggeration scale 100
Comparison of Radial Internal

Support Pressure
381 of 607
Summary of 2D Modelling of Tunnelling
A. Good prediction of greenfield surface settlement curve (Gaussian) in 2D requires
1. advanced soil constitutive model for nonlinear stiffness from small strains
2. refined method of modelling tunnel excavation
B. Tunnelling example investigated herein:
1. effect of advanced constitutive model is more significant than method of
modelling tunnel excavation
2. Applied Pressure Method gives steeper surface settlement curve, followed
by Stress Reduction Method & Lining Contraction Method
C. On realistic prediction of surface settlement curve & pattern of ground deformation
around tunnel:
1. Mohr Coulomb model + Lining Contraction Method gives unrealistic results
2. HS & HS-small models + Applied Pressure Method gives better results
TUNNELLING ADJACENT TO A BUILDING

SUPPORTED BY END-BEARING PILES
Part 3.1
382 of 607
Presentation Outline
1. Background of tunnel-pile interaction problem
2. Example of tunnelling beneath a piled building in HK
3. 3D finite element modelling of tunnel advance
4. Response of piles to tunnelling
5. Discussion on 3D FE analysis vs. analytical solution
Three‐Stage Building Damage Assessment Due to Tunnelling
1. 3D analysis for detailed evaluation

only
2. Give greater certainty on
requirements for protective measures
3. Saving in construction cost & time
justifies time spent on 3D analysis
4. 3D analysis has potential to add value
to tunnel design & construction
process
Burland (1995)
383 of 607
Tunnel‐pile‐soil Interaction
• A three-dimensional problem due to
1. progressive advance of tunnel face towards piles
2. movement of piles in 3D
3. oblique orientation of building relative to tunnel alignment
• Tunnelling induced ground movements can cause

1. increase/decrease in pile axial force (negative/positive skin friction) – relative pile/soil
vertical displacement
2. increase in pile bending moment – curvature of pile horizontal displacement
3. potential reduction in pile geotechnical capacity – reduction in soil effective stresses
4. distortion of building, e.g. angular distortion & horizontal strain
Zones of Influence
Zone C Zone B Zone A Zone B Zone C
Pile settlement
C B A
Depth
Selementas et al. (2005) 45º

45º
For pile toe located in

Zone A: pile head settlement > soil surface settlement; decrease in pile axial force
Zone B: pile head settlement ≈ soil surface settlement
Zone C: pile head settlement < soil surface settlement; increase in pile axial force
384 of 607
Analysis of Tunnel‐pile Interaction
• Typically use the combination of
1. empirical relationships/closed-form solutions to estimate
greenfield ground movements; and
2. boundary element methods to compute pile deformations and
stresses
• Suitable for preliminary assessment, with some limitations
• Alternatively, use 3D numerical analysis
Pros: model tunnelling, tunnel-pile-building interaction &
geotechnical entities in one single analysis
Cons: complicated, relatively long analysis time & require
advanced constitutive model for soil non-linear behaviour
Example of Tunnelling Below Piled Building
25m
25m
0 mbgl P4 P5 Rear P6
2m Pile cap
5 mbgl Fill 9m 10m
1m
4m
MD
10 mbgl P1 P2 Front P3
CDG 10m 6m Ø tunnel
1m
4m
20 mbgl Tunnel advance direction
Tunnel 2m Ø pile
6m Ø Pile design load 15MN (~5MPa)
30 mbgl
31.5 mbgl Rock 3m Ø bell-out
P1/P4 P2/P5 P3/P6
385 of 607
Information for Tunnel, Piles & Ground
1. 6 m diameter tunnel excavated by TBM, tunnel axis depth at 20 mbgl in

Completely Decomposed Granite
2. 15-storey building supported by 6 nos of 2 m diameter bored piles with 3 m
diameter bell-outs in rock at 32 mbgl
3. Each pile takes 15 MN design load (~5 MPa).
4. Building plan size is 25 m by 9 m, pile cap 2 m thick
5. Stratigraphy is 5 m Fill, 5 m Marine Deposits, 20 m CDG and rock. Groundwater
table at 2 mbgl
6. Tunnel constructed in between piles, tunnel edge to pile edge distances are 1 m, 4
m and 10 m
Soil Small Strain Non‐linear Stiffness
0.01% 0.1% 1%
Atkinson & Sallfors

(1991)
386 of 607
CDG Small Strain Non‐linear Stiffness
Laboratory small strain stiffness
• Hardening Soil + Small Strain
results for CDG samples
Overlay (HSsmall) constitutive
model to consider CDG small strain
Ng et al. (1998) non-linear stiffness
1600 Adopted line

1400 Triaxial_Upper
Triaxial_Low er
1200
HSsmall_Upper
1000
HSsmall_Low er
Gsec /p'
800 HSsmall_Baseline
600
400
200
0
0.0001 0.001 0.01 0.1 1 10
Shear strain (%)
3D Finite Element Model (Plaxis‐GiD)
Rear Load 15 MN
Building “Plate”
40m Pile cap
Bored pile
Front
Fill Tunnel face
Tunnel MD 149m
CDG
120m Rock TBM
length
Bell-out
43,000 elements Linings
Refined mesh around tunnel & building piles
387 of 607
Tunnel Confinement Pressure
A
PIII A. Face support pressure (PI to PII) =
PIV
hydrostatic pore pressure + overpressure
PI
Rear Front B. Along TBM shield, tunnel support pressures
vary to consider
6m Ø TBM shield 9m
1. Conical shape of TBM shield / over-
cutting
PII
2. Ground loss into tail void in rear
PVI C. Any combination of support pressure
PV profiles can be modelled
A
PIII
Pressure
PV increases
Section A-A with depth
Modelling of Tunnel Face Advance
1. Soil elements inside TBM shield are

Lining
Lining
TBM shield
deactivated
(elements nulled)
2. Apply tunnel support pressure profiles
3. For each face advance, shift tunnel
support pressures forward &
1.5 1.5m
correspondingly erect new lining behind
TBM
4. The process is repeated as tunnelling
Lining
Lining
TBM shield
(elements nulled) progresses
1.5 1.5m
Lining
Lining
TBM shield
(elements nulled)
1.5 1.5m
388 of 607
Modelling of Structures
1. Piles & pile cap modelled by solid elements
2. Interface elements along pile shafts & on pile cap vertical faces
3. Consider flexural stiffness (EI) & axial stiffness (EA) of superstructure by
incorporating a “Plate” structural elements on top of pile cap.
Superstructure EI estimated by (Potts & Addenbrooke, 1997)
1. Parallel Axis Theorem (bending about building neutral axis); or
2. Summation of EI for individual building storeys
4. Tunnel linings modelled by “Plate” elements
Prediction on Ground Surface Settlement
Overpressure 20 kPa, G/F VL 1.6% Overpressure 20 kPa
Fill -60 -40 -20 0 20 40 60

0 VL 0.31%
MD -4
Settlement (mm)
-8 VL 1.61%
-12
Mid-building
CDG
-16
Greenfield
Tunnel -20
Gaussian
-24
• Lateral spreading of • Gaussian curve with K = 0.45

displacements in MD layer • Close to K ≈ 0.5 from HK
• Settlement trough becomes wider tunnelling experience
389 of 607
Prediction on Pile Transverse Displacement
Overpressure 20 kPa
Transverse horizontal disp. (mm) +10D
-5 -4 -3 -2 -1 0
0
-2D
Front 5
Rear +2D
10
+2D
Depth (mbgl)
15
+10D Rear
20 1m P2
Front
25
30 -2D
35
Tunnel advance
Prediction on Pile Longitudinal Displacement
Overpressure 20 kPa
Longitudinal horizontal disp. (mm) +10D
-3 -2 -1 0 1 2
0
10 +2D
Depth (mbgl)
15 Rear
-2D 20 1m P2
Tunnel advance
Front Front
25
Rear
+2D
30 -2D
+10D 35
Tunnel advance
390 of 607
Prediction on Pile Settlement & Axial Force
Overpressure 20 kPa
Settlement (mm) Increase in axial force (MN)
0 -1 -2 0 1 2 3 4
0 0
P2 -2D P2 -2D
5 Front 5 Front
Rear Rear
10 +2D 10 +2D
+10D
Depth (mbgl)
Depth (mbgl)
+10D
15 7
15
20 20
25 25
A B
30 30
C
35 35
Pile toe
Prediction on Pile Bending Moment
Overpressure 20 kPa
Transverse moment (kNm) Longitudinal moment (kNm)
1500 500 -500 -1500 1500 500 -500 -1500
0 0
P2 P2
5 5
10 -2D 10 -2D
Depth (mbgl)
Depth (mbgl)
Front Front
15 15
Rear Rear
+2D +2D
20 20
+10D Tunnel +10D
25 25advance
30 30
35 35
391 of 607
Check on Potential Structure Damage
45 OP 10kPa Distance from tunnel centre (m) OP 10kPa

P2 OP 20kPa -10 -5 0 5 10 15
OP 20kPa
35 0.0 0.3 Cat. 4 &
OP 30kPa
5 OP 30kPa
Axial Force, N (MN)_
OP 40kPa
25 OP 40kPa
_
Bldg. settlement (mm)
-0.4 0.2
/L (%)
15
5 -0.8 Cat.
0.1
-5 =0.14 2 3
-1.2 mm 1
-15 0.0 0
-15 -10 -5 0 5 10 15 0 0.1 0.2 0.3
Moment, M (MNm) -1.6 h (%)
Pile N-M Interaction Diagram Building deflection Burland’s chart
Comparison with Closed Form Solution
Greenfield subsurface settle. (mm) Greenfield subsurface horiz. disp. (mm)
-35 -25 -15 -5 5 15 -15 -10 -5 0
0 0 0
Fill Fill
5 5
MD MD
10 10
Depth (mbgl)
Depth (mbgl)
15 CDG 15
CDG
20 20
25 25
Loganathan Loganathan
et al. (2001) 30 et al. (2001) 30
Rock
3D analysis Rock 3D analysis
35 35
Greenfield subsurface section corresponds to P2 location

3D analysis: Overpressure 20 kPa, G/F VL 1.61%
392 of 607
3D FEA vs. Analytical Solution
Issues 3D FEA Analytical Solution
Ease of use 1. Complicated 1. Relatively easy

2. Long analysis time 2. Less analysis time
Ground conditions 1. Layered soil 1. Homogeneous soil

2. Need realistic constitutive 2. Estimated greenfield
model deformation less good for
layered soil
Tunnelling 1. Model face advance 1. Only pile response in

progress 2. Pile response in transverse transverse direction
& longitudinal directions
3D FEA vs. Analytical Solution
Issues 3D FEA Analytical Solution
Ground loss, VL 1. Model confinement 1. Assume a certain VL
pressure & predict VL
Effect on 1. Model tunnel, piles, 1. Different boundary element
piles/building building & their interaction in programs for pile axial and
one single analysis lateral responses
2. Results from piles & 2. Specific analysis for pile
building used directly in group effect
structural check 3. Dedicated modification
factors account for building
rigidity
393 of 607
TUNNELLING BENEATH A BUILDING

SUPPORTED BY PILES
Part 3.2
Tunnelling Below Building on Frictional Piles
6m ø tunnel P8 2m
P7
P4
P6
P3 P5 P6 P7 P8
P5 23m
P2 P1 P2 P3 P4
Bldg. footprint
P1 33×11m
y x
3m y
z 2m ø piles
31 mbgl
Tunnel advance direction z x
6m ø tunnel
Plan view Vertical x-section
394 of 607
3D Model by Plaxis-GID
Pile cap + “Plate”
45m
Building 23m
Fill 3m Bored piles

MD 2m Ø
CDG
165m
Tunnel
Tunnelling
direction
136m
Tunnel 6m Ø
• 82,600 wedge/pyramid/tetrahedral elements
Pile
Settlement & Axial A B
P5 Force Pile P5 C
Rear
cap
Front
cap Settlement, Uy (mm) Pile axial laod, N (MN)
0 -5 -10 -15 -20 -25 -30 0 2 4 6 8 10 12
0 0
-5 -5
Depth (mbgl)
Depth (mbgl)
-10 -10
-15 -15 Initial
-20 -20
-25 -25
Front Rear +2D +15D Front Rear +2D +15D
395 of 607
Channel Tunnel Rail Link (UK) Measurement
Selementas (2005)
2m
Approaching
Final Initial
• Tunnel 8.15m dia. at 19m depth

• Driven cast-in situ pile 0.48m dia.
• Layered ground
Longitudinal Horizontal Displacement &
P5
Rear
Bending Moment Pile P5
cap
Front
Longitudinal hor. disp., Uz (mm) Longitudinal moment, Mz (kNm)
cap -1 0 1 2 3 -400 -200 0 200 400 600
0 0
-5 -5
Depth (mbgl)
Depth (mbgl)
-10 -10
-15 -15
Tunnel
-20 -20
advance
-25 -25
396 of 607
Transverse Horizontal Displacement &
P6
Rear
Bending Moment Pile P6
cap
Front
cap Transverse hor. disp., Ux (mm) Transverse moment, Mx (kNm)
-6 -5 -4 -3 -2 -1 0 -2000 -1500 -1000 -500 0 500
0 0
-5 -5
Depth (mbgl)
Depth (mbgl)
-10 -10
-15 -15
-20 -20
-25 -25
Tunnel
Greenfield Surface & Building Settlements
Reference distance (m)

-70 -50 -30 -10 10 30 50 70
0 VL 0.72%
Settlement (mm)_
-5
-10
-15
Greenfield
-20
Building
-25 VL 2.8%
-30 Building
397 of 607
Building Settlements > Greenfield Surface Settlements
Greenfield
surface
settlements
F: front; R: rear Bldg.

settlements
Sheung Wan GCO TN4/85
founded on 73 nos. of
0.457m Ø concrete piles
Sheung Wan 5.8m Ø

overrun tunnel built in
1980s
TUNNELLING NEAR A BUILDING

SUPPORTED BY LARGE NUMBERS OF PILES
Part 3.3
398 of 607
Tunnelling Near a Group of 48 Piles
0 mPD
Pile cap 1.5m
6m Ø tunnel Building footprint
13.6m
1m 0.6m Ø Franki piles @

3Ø spacing
10m -20mPD
1m
0.6m Ø Franki piles
Tunnel advance 6m Ø tunnel
Plan View Front View
3D Model by Plaxis 3D
Bldg. load
“Plate” modelling
superstructure EI & EA
Building
40m
Fill 1m 48 Franki piles
(Embedded Piles)
CDG Tunnel
advance
Tunnel 120m
140m
6m Ø tunnel
• Analysis by Plaxis 3D Release Candidate 2

• 69,951 nos. of 15-node wedge/13-node pyramid/10-node tetrahedral
elements
399 of 607
Effect of 3m Thick Annulus Grout
Grouted
annulus
21m
3m
Tunnel
• Grout modelled as Mohr Coulomb material

c’ = 100 kPa, ’ = 35°, E = 150 MPa
Effect of Fixed Pile Head Connections
Pile cap
“Plate” modelling
superstructure
Tunnel
• Place “Plate” modelling superstructure on underside of pile cap to form

fixed connections with “Embedded Pile” heads
400 of 607
Output of Results
Iso-surface of soil total displacements Pile group deformations
Isometric view Front view
Tunnel Tunnel
advance advance
Ho rizo
Horizontal ntal distance
distance fro mcentreline
from tunnel centreline (m)
(m)
-60 -40 -20 0 20 40 60
0
-2
Settlement (mm)
-4
-6
-8
Greenfield
-10
B uilding
Bldg.
-12
• For pile toes above tunnel, building settlements are greater than
greenfield surface settlements due to undermining below pile toes
401 of 607
Building Settlements > Greenfield Surface Settlements
Greenfield
surface
settlements
F: front; R: rear Bldg.

settlements
Sheung Wan GCO TN4/85
founded on 73 nos. of
0.457m Ø concrete piles
Sheung Wan 5.8m Ø

overrun tunnel built in
1980s
Comparison of Building Settlements
Ho distance
Horizontal rizo ntal distance fro centreline
from tunnel m centreline
(m)(m)
0 5 10 15 20 25
0
Building settlement (mm)
-2
Annulus grout
-4
-6
Fixed pile heads
-8 A 1- B aseline
A 2 - A nnulus gro ut
-10
A 3 - Fixed heads
-12 Baseline
(Pinned pile heads)
• Annulus grout reduces building settlements by half

• No significant difference (1 mm) between pinned & fixed pile heads
402 of 607
Transverse Horizontal Displ. of Closest Pile
P ile transverse Ux (mm)
-6 -5 -4 -3 -2 -1 0 1
0
-4
A 1- B aseline
Level (mPD)
(Pinned pile heads)
-8
A 2 - A nnulus
Gro ut
A 3 - Fixed -12
Heads Annulus grout
-16
-20
Tunnel
3D NUMERICAL MODELLING OF TUNNELLING

INTERSECTING PILES
Part 3.4
403 of 607
Tunnelling Intersecting Piles
17.8m
Building footprint
50 nos 0.6m Ø Franki 8.8m Plan View
piles @ 3Ø spacing
Tunnel 6m Ø tunnel
advance
0 mPD
Pile cap 1.5m
Front View
15 nos. pile toes to

be trimmed by ~ 3m
-23 mPD
3D Model by Plaxis 3D
Upper half
Full annulus grout
annulus grout
Bldg. load
“Plate” models
superstructure
Building
40m
Fill
CDG
Tunnel Grout
120m 3m
150m 12m
• Analysis by Plaxis 3D Release Candidate 2

• 79,404 nos. of wedge/pyramid/tetrahedral elements
404 of 607
Output of Results
Isometric view Front view
6m Ø tunnel
Exaggeration scale 100x
Horizontal distance from tunnel centreline (m)
-60 -40 -20 0 20 40 60
0
Settlement (mm)
-10
-20
-30 Greenfield
Building
-40
Bldg.
• Baseline analysis without grouting
405 of 607
Effect of 3m Thick Annulus Grout on Bldg. Settlements
Horizontal distance from tunnel centreline (m)

-16 -14 -12 -10 -8 -6 -4 -2 0 2 4
0
Building settlement (mm)
Full annulus grout

-10
Half annulus grout

-20
Baseline No grout
-30
Half grout
Full grout
-40
Effect of Annulus Grout on Pile Settle. & Axial Force
Critical pile
Tunnel 6m Ø tunnel
advance
Pile settlement (mm) Pile axial force (MN)
-40 -30 -20 -10 0 -1 0 1 2 3
0
Tension
-4
Level (mPD)
-8
Baseline -12
(No grout)
-16 Initial
Half grout
Baseline (No grout)
-20 Half grout
Full grout
-24 Full grout
406 of 607
Effect of Annulus Grout w.r.t Pile N‐M Capacity
Critical pile
Tunnel 6m Ø tunnel
advance
2500 Pile capacity
Baseline (No grout)
2000 Half grout
Full grout
Axial force (kN)
1500
1000
500
-500
-200 -150 -100 -50 0 50 100 150 200
Moment (kNm)
Summary
1. Details & results for 3D modelling of tunnel advance near a piled building are given:
a. pressure-controlled boundary on tunnel face & along TBM
b. integrated response of piles & building to tunnelling in 3D
c. bldg. on end-bearing piles: bldg. settlement < greenfield surface settle.
d. bldg. on friction piles (toes above tunnel): bldg. settlement > greenfield surface settlement
2. For detailed evaluation of identified critical buildings, 3D FEA:

a. offers a more detailed assessment of pile/structure behaviour than empirical/closed form
solution + boundary element method
b. considers layered ground, tunnel face advance, TBM support pressure, bldg. stiffness &
combined piles-cap-bldg. behaviour
3. 3D analysis adds value to tunnel design & construction process, e.g. assessment of
requirement for protective measures
407 of 607
References
1. Atkinson, J. H. & Sallfors G. (1991). Experimental determination of soil properties. Proc. 10th ECSMFE, Florence, Vol.3, 915-956
2. Burland, J. B. (1995). Assessment of risk of damage to buildings due to tunnelling and excavation. 1st Int. Conf. on Earthquake Geotech. Engrg., IS Tokyo.
3. Geotechnical Control Office (GCO) (1985). Technical Note T4/85 - MTR Island Line: Effects of Construction on Adjacent Property. Civil Engrg. Services
Dept., Hong Kong.
4. Hake, D. R. & Chau, I. P. W. (2008). Twin stacked tunnels - KDB200, Kowloon Southern Link, Hong Kong. Proc. 13rd Australian Tunnelling Conference, 445-
452.
5. Loganathan, N., Poulos, H. G. & Xu, K. J. (2001). Ground and pile-group responses due to tunnelling. Soils and Foundations, 41(1), 57-67.
6. Moller, S. (2006). Tunnel induced settlements and structural forces in linings. PhD thesis, University of Stuttgart.
7. Moller, S. & Vermeer, P. A. (2008). On numerical simulation of tunnel installation. Tunnelling & Underground Space Technology, 23, 461-475.
8. Ng, C. W. W., Sun, Y. F. & Lee, K. M. (1998). Laboratory measurements of small strain stiffness of granitic saprolites. Geotechnical Engineering, SEAGS,
29(2), 233-248.
9. Pang, C. H. (2006). The effects of tunnel construction on nearby pile foundation. PhD thesis, National University of Singapore.
10. Potts, D. M. & Addenbrooke, T. I. (1997). A structure’s influence on tunnelling-induced ground movements. Geotechnical Engineering, Proc. ICE, 125, 109-
125.
11. Schnaid, F., Ortigao, J. A. R., Mantaras, F. M., Cunha, R. P. & MacGregor, I. (2000). Analysis of self-boring pressuremeter (SBPM) and Marchetti
dilatometer (DMT) in granite saprolites. Canadian Geotechnical J., 37, 796-810.
12. Selementas, D. (2005). The response of full-scale piles and piled structures to tunnelling. PhD thesis, University of Cambridge.
13. Storry, R. B. & Stenning, A. S. (2001). Geotechnical design & contraction aspects of the Tsing Tsuen Tunnels – KCRC West Rail Phase; Contract DB320.
Proc. 14th SEAGC, Hong Kong, 443-448.
14. Storry, R. B., Stenning, A. S. & MacDonald, A. N. (2003). Geotechnical design and construction aspects of the Tsing Tsuen Tunnels – contract DB320 KCRC
West Rail Project. Proc. ITA World Tunnelling Congress, (Re)claiming the Underground Space, Saveur (ed.), 621-626.
15. Vermeer, P. A. & Brinkgreve, R. (1993). Plaxis Version 5 Manual. Rotterdam, a.a. Balkema edition.
408 of 607
Embedded Elements in Plaxis 3D
(A comparison of embedded pile against solid pile)
Based on original course note by

Prof. Harry Tan
By
RF Shen
25 November 2011
Embedded pile in Plaxis 3D
Embedded pile in Plaxis
3D can be used to
simulate piles to obtain
pile movement profiles,
internal forces readily.
HOWEVER, a good
understanding of the
behavior of embedded
pile in Plaxis 3D is critical
for the proper use of this
very useful element.
Illustration of a pile behind
excavation 2
Singapore Advanced- Dr RF Shen
409 of 607
An embedded pile = Beam element + Interface
element around the beam element to interact
with the surround soil elements.
As such, the definition of an embedded pile
element consists of 2 parts: properties of the
beam & properties of the interface element
(skin resistance and foot resistance).
For the definition of the
beam element part, it is
much the same as that
defined for the beam
elements.
Be cautious of non‐isotropic
pile (like H‐pile) with the
possibility of different pile
orientation, as will be
illustrated in the next slide!
4
410 of 607
In Plaxis 3D, the 1st local axis is red

indicated by a RED arrow align blue
with the embedded pile
green
direction; the 2nd local axis is
GREEN arrow and the 3rd axis is
BLUE arrow.
For non‐isotropic pile (like H‐pile)
the 3rd axis is the major axis with
assigned larger I3 in the input,
the alignment of the axis should
be adjusted accordingly for
lateral loading cases.
Illustration: Embeddedpile Orientation.P3D
For a higher resistance
to excavation‐induce
bending moment, the H‐
piles are oriented with
major axis to be bending
toward the excavation
side, has the embedded
pile orientation being
properly configured in
the right‐hand figure?
Illustration of a pile behind
excavation 6
411 of 607
An embedded pile can cross
a 10‐node tetrahedral soil
element at any place with
any arbitrary orientation,
introducing 3 extra nodes
inside the 10‐node
tetrahedral soil element.
An embedded pile crossing an
tetrahedral soil element
7
Interaction between
embedded pile and
surrounding soil element at
each node is based on:
embedded pile interacting with
surrounding soils
8
412 of 607
While interface element has been
provided along the embedded pile
shaft (good for correct simulation of
axial pile‐soil interaction), take note
that NO interface element provided
for the lateral soil sliding around the
embedded pile. As such, for laterally
loaded case, embedded pile only
works well in working condition
when there is not much pile‐soil
lateral sliding occurs, it can not be
used for simulation of ultimate embedded pile interacting with
lateral loading scenario. surrounding soils
9
Performance of embedded pile versus
solid pile in Plaxis 3D
• Single Bored Pile of 1m Dia and 20m L

• M-C soil of Cu=100kPa, =0, E’=40000
kPa, ’=0.3, Rinter=1
• Pile loaded in axial compression +Fz
• Pile loaded in axial tension -Fz
• Pile loaded laterally +Fx
10
413 of 607
3D FEM mesh for solid pile
The pile is simulated
by dia. 1m solid
cylindrical object with
surrounding interface
with Rinter=1
The dia. 1m solid
cylindrical object has
concrete elastic properties
with E=3.0E+7kPa 11
3D FEM mesh for embedded pile
Defining the beam
properties: The
embedded pile has the
structural properties
match exactly the dia.
1m bored pile
12
414 of 607
Calculated pile capacity
Defining embedded pile interface properties:
Skin resistance: Cu=100kPa, Rinter =1, Thus, Tmax =
3.14*1m*100kPa = 314kN/m.
End‐bearing resistance: qb = 9Cu = 900kPa, Fmax =
0.25*3.14*(1m)^2*900kPa = 706kN
So, Total shaft resistance Fshaft = 314kN/m * 20m = 6280kN
Total base resistance Fmax = 706 kN
13
Total pile resistance Ftotal = Fshaft + Fmax = 6990 kN
Axially loaded embedded pile
Fz =7000kN
=1000kN
=2000kN
=3000kN
=4000kN
=5000kN
=6000kN
14
415 of 607
Punching of embedded pile under
theoretical axial load
Fz =7000kN
15
Embedded pile toe resistance under axial loads
Fz =7000kN
16
416 of 607
Axial load transfer curves
Axial force (kN)
0 2000 4000 6000 8000
0
‐5
Elevation (m)
‐10
Embeddedpile Fz=1000kN
‐15
‐20 Embeddedpile Fz=5000kN
17
‐25
Shaft skin resistance of embedded pile
0 20 40 60 80 100 120
0
‐5
Elevation (m)
‐10
‐20
‐25
Shaft friction (kPa) 18
417 of 607
Axial load‐settlement behavior
0 1000 2000 3000 4000 5000 6000 7000 8000

0
10
20
Pile settlement (mm)
30
40
50
60
70
80 Embeddedpile
EmbeddedpileSolid pile
90
100 19
Axial load (kN)
Axial load transfer curves
Axial force (kN)
0 2000 4000 6000 8000
0
‐5
Elevation (m)
‐10
Solid pile (1000kN)
‐20 Solid pile (3000kN)
Solid pile (5000kN)
Solid pile (7000kN)
20
Solid pile (8000kN)
‐25
418 of 607
3D FEM mesh for embedded pile
0 20 40 60 80 100 120
0
‐5
Elevation (m)
Solid pile (1000kN)
Solid pile (3000kN)
‐20
Solid pile (5000kN)
Solid pile (7000kN)
Solid pile (8000kN)
‐25
Shaft friction (kPa) 21
Concluding remarks: Embedded pile performs
satisfactorily under axial loads and conform to
theoretically values, while solid pile exhibit too high end
bearing resistance and much stiffer pile response, and
thus needed to be used with cautions.
Can tension loading case eliminate the end‐bearing
difference?
Total shaft resistance Fshaft = 314kN/m * 20m = 6280kN
Self‐weight of bored pile = 0.25*3.14*(1m)^2 * 20m *
24kN/m^3 = 377 kN
So, expected total pull out resistance = 6280 + 377 = 6660kN
22
419 of 607
Can tension loading case eliminate
the end‐bearing difference?
‐50
‐40 Embededpile_use actual load
Embededpile_use actual load
Pile movement (mm)
Solid pile
‐30
‐20
‐10
0
0 1000 2000 3000 4000 5000 6000 7000 8000 23
Tension load (kN)
Load transfer curves under tension loadings
Tension force (kN)
‐8000 ‐7000 ‐6000 ‐5000 ‐4000 ‐3000 ‐2000 ‐1000 0
0
‐5
Elevation (m)
‐10
Embeddedpile Fz=5000kN ‐15
Solid pile (1000kN)
Solid pile (3000kN)
‐20
Solid pile (5000kN)
Solid pile (7000kN)
Solid pile (8000kN) 24
‐25
420 of 607
In general, embedded pile performs satisfactorily
under both axial compression loads and tension
loads and generally conform to theoretically
values, while solid pile exhibit too high end
bearing resistance and much stiffer pile response
under axial compression load and develop very
large suction at the toe of solid pile under tension
load which may not be so reliable, and thus
needed to be used with cautions.
How about the performances under lateral loads?
25
Estimation of lateral pile capacity assuming pile is rigid
enough and has sufficiently high strength, and failure occurs
in the clay
Brom's Theory:
Free head, L/d=20, e/d=0, Hu/cud^2=60
26
Hu=6000 kN
421 of 607
Lateral loading on pile in Plaxis 3D
27
Lateral load – movement curves
400
350
Embeddedpile
Solid pile
300 Solid pile
Lateral pile movement (mm)
250
200
150
100
50
0
0 1000 2000 3000 4000 5000 6000 7000
28
Lateral load (kN)
422 of 607
Take note that NO interface element
provided for the lateral soil sliding
around the embedded pile. As such,
for laterally loaded case, embedded
pile only works well in working
condition (FOS=2.0~3.0) when there
is not much pile‐soil lateral sliding
occurring, it can not be used for
simulation of ultimate lateral
loading scenario.
embedded pile interacting with
surrounding soils
29
BM under lateral loadings
14000
12000 Embeddedpile
Solid pile
Pile bending momnet (kNm)
10000
8000
6000
4000
2000
0
0 1000 2000 3000 4000 5000 6000 7000
30
Lateral load (kN)
423 of 607
Comparison of pile deflection profiles
‐20 0 20 40 60 80 100 120
0
‐2
‐4
‐6
Elevation (m)
‐8
Embeddedpile Fx=1000kN
‐10
‐12 Embeddedpile Fx=3000kN
‐14
Solid pile Fx=1000kN
‐16 Solid pile Fx=2000kN
‐18
‐20
Lateral pile movemetn (mm) 31
Comparison of BM profiles
‐2000 0 2000 4000 6000 8000
0
‐5
Elevation (m)
‐10
‐15 Embeddedpile Fx=3000kN
‐20 Solid pile Fx=2000kN
‐25
32
Lateral pile movemetn (mm)
424 of 607
Conclusions
• Embedded pile is a good model of single pile response
subjected to both vertical compression loads and
vertical pull out loads.
• Solid pile may give very large end bearing resistance
when subjected to vertical compression load, and
generate large suction force at the base when subjected
to pull out load, and thus must be treated with caution.
• Under working load condition with FOS=2~3 when there
is no much pile‐soil lateral slide occurring, embedded
pile give very reasonable pile deflection and pile bending
moment. However, embedded pile can not be used for
simulation of ultimate lateral loading scenario.
33
425 of 607
0123345
ÿ455667
ÿ789

ÿ ÿ

8 ÿ4ÿ2
ÿ
ÿ

23ÿ!"ÿ22
426 of 607
427 of 607
428 of 607
429 of 607
430 of 607
431 of 607
432 of 607
433 of 607
434 of 607
435 of 607
436 of 607
437 of 607
438 of 607
439 of 607
440 of 607
441 of 607
442 of 607
443 of 607
444 of 607
445 of 607
446 of 607
447 of 607
448 of 607
449 of 607
450 of 607
451 of 607
452 of 607
453 of 607
454 of 607
455 of 607
456 of 607
457 of 607
458 of 607
459 of 607
460 of 607
461 of 607
462 of 607
463 of 607
464 of 607
465 of 607
466 of 607
467 of 607
Simulation of laboratory tests
SIMULATION OF LABORATORY TESTS
Computational Geotechnics 1
468 of 607
2 Computational Geotechnics
469 of 607
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
• Simulation of laboratory tests
– 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.
470 of 607
Figure 1: Oedometer test results on sand
Figure 2: Development of horizontal and vertical stress in oedometer test
471 of 607
Figure 3: Triaxial test unloading-reloading (cell pressure = 100 kPa)
Figure 4: Axial vs. volume strain in drained triaxial test
472 of 607
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]
νur [-]
c’ [kPa]
ϕ0 [o ]
ψ [o ]
m [-]
K0N C [-]
With these data perform a triaxial test in the SoilTest program.
473 of 607
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
474 of 607
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]
νur [-]
c’ [kPa]
ϕ0 [o ]
ψ [o ]
m [-]
K0N C [-]
With these data perform an oedeometer test in the SoilTest program.
475 of 607
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 σ30 = 100 kPa a maximum value of approximately |σ10 − σ30 | = 400 kPa is reached at failure.
The Mohr-Coulomb failure criterium is:
1 0
2 |σ1 − σ30 | + 12 (σ10 + σ30 ) · sinϕ − c · cosϕ = 0
Considering it is sand we assume that the cohesion is zero and so the Mohr-Coulomb failure criterium reduces
to:
|σ10 −σ30 |
(σ10 +σ30 ) = sinϕ
Filling in σ30 = 100 kPa and σ10 = 500 kPa as obtained from the test we find for the
friction angleϕ0 = 420
Reference stiffness from triaxial test
The triaxial test stiffness E50 is the secant stiffness over the first 50% of the failure value for | σ10 − σ30 |. This is
indicated in red in the triaxial test graph of figure 8.
σ 0 =100 kP a 400
E503 = 0.013 = 30800 kP a
476 of 607
The triaxial test stiffness ,E 50 , is within the Hardening Soil model defined as:
m m
c cosϕ−σ30 sinϕ σ30

ref ref
E50 = E50 c cosϕ+pref sinϕ , c = 0 ⇒ E50 = E50 − pref
The reference stress pref is chosen equal to the cell pressure of this triaxial test then
ref σ 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.
σ 0 =100 kP a 400
Eur3 = 0.026−0.021 = 80000 kPa
Under the same assumptions as for the stiffness in triaxial testing counts:
ref σ 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
∆εv
sinψ = −2∆ε1 +∆εv
See figure 9 for details.

With ∆εv = 0.048-0.004 = 0.044 and ∆ε1 = -0.09-(-0.03) = -0.06 the dilatancy can be calculated as ψ=16o
Note: The Poisson’s 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 Poisson’s ratio needed is an unloading-reloading
Poisson’s ratio.
An acceptable value for the unloading-reloading Poisson’s ratio is νur = 0.2.
477 of 607
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 σy0 = 100 kPa en σy0 =
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.
σ 0 =100 kP a
y 320−0
Eoed = 1.4%−0.33% = 29900 kPa
σy0 =200 kP a 400−0
Eoed = 1.4%−0.47% = 43000 kPa
Within the Hardening Soil model the stress dependent oedometer stiffness is defined as:
m m
c cosϕ−σy0 sinϕ σy0

ref ref
Eoed = Eoed c cosϕ+pref sinϕ , c = 0 ⇒ Eoed = Eoed − pref
Choosing the reference pressure pref = 100 kPa gives
ref σ 0 =100 kP a
Eoed = Eoed
3
≈ 30000 kPa
The power m for stress dependent stiffness can now be determined as:
σ 0 =200 kP a m
y
σy0

Eoed 43000 200 m

ref
Eoed
= pref ⇒ 30000 = 100 ⇒ m = 0.5
478 of 607
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
∆σx ∆σ30 100
K0N C = ∆σy0 = ∆σ10 = 300 = 0.33
Alternatively one can use Jaki’s formula
K0N C ≈ 1 − sinϕ = 1 − sin(42o ) = 0.33
479 of 607
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 σ3 while the oedometer test results presented in figure 10 give the strain vs
the vertical stress σy (= σ1 voor oedometer testing).
σ 0 =100 kP a σ 0 =100/K0N C kP a σ 0 =300 kP a 400

Eur3 = Eur1 = Eur1 = 1.28%−0.91% = 108000 kPa
With pref = 100 kPa (pref refers to σ30 !) it follows that
ref σ 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
νur [-] 0.2
c’ [kPa] 0
ϕ0 [o ] 42
ψ [o ] 16
m [-] 0.5
K0N C [-] 0.33
480 of 607
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:
|σ1 −σ3 | σ1 +σ3

2 + 2 sinϕ − c cosϕ = 0
With p0 = (σ10 + 2σ30 )/3 and q = σ10 − σ30 under triaxial test conditions this can be rewritten as:
2p0 + 13 q

q 6sinϕ 0 6c cosϕ
2 = 2 sinϕ − c cosϕ = 0 ⇒ q = 3−sinϕ p + 3−sinϕ
Hence, the slope M of the Mohr-Coulomb line in p’-q plane is defined as:
6sinϕ 195
M= 3−sinϕ = 200 ⇒ ϕ = 250
From the intersection between Mohr-Coulomb line and the vertical axis where p=0 the cohesion can be determined:
6c0 cosϕ
q= 3−sinϕ = 0 ⇒ c = 0 kPa
481 of 607
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)-εv graph,
hence they have a relation of the form:
εy = εv = A · log(σy0 )
ε2 −ε1 0.370−0.270
A= log(σ2 )−log(σ1 ) = 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(σy0 )
εy = εv = A · ln(10)
dεy dσy0 ln(10)
dσy0 =A· 1
ln(10) · 1
σy0 ⇒E= dεy = A · σy0
The E modulus found is the oedometer stiffness can be rewrittens as:
σy0

ln(10)
E = Eoed = A · pref − pref
In the Hardening Soil model the oedometer stiffness is defined as (assuming c = 0) :
m
σy0

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:
482 of 607
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:
εy = εv = B · log(σy0 )
ε2 −ε1 0.427−0418
B= log(σ2 )−log(σ1 ) = log(120)−log(30) =0.0149
dσy0 ln(10)
Eur = dεy = B · σy0
However, the Eur in the Hardening Soil model is dependent on the smallest principal stress, which is σx0 in an
oedeometer test and not σy0 .
During the unloading process there is no linear relation between horizontal and vertical stress, as in the beginning
of unloading σy0 > σx0 where as after much unloading σy0 < σx0 . Therefore the assumption is made that during
unloading on average σx0 = σy0 .
0

ln(10) ln(10) ln(10) σx
Eur = B · σy0 = B · σx0 = B · pref pref
With the definition of Eur in the Hardening Soil model of
0
m
ref σx
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
Jaky’s rule:
K0N C ≈ 1 − sinϕ = 1 − sin(250 ) =0.58
Poisson’s ratio
The Poisson’s ratio for unloading and reloading is again estimated as νur = 0.2
483 of 607
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
νur [-] 0.2
c’ [kPa] 0
ϕ0 [o ] 25
ψ [o ] 0
m [-] 1.0
K0N C [-] 0.58
484 of 607
485 of 607
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.
486 of 607
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.
487 of 607
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.
488 of 607
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 σxx , σyy and σzz 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: τxy = 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 (∆σyy ) until the stress level where the unloading path should start.
Note that the horizontal stresses (∆σxx and ∆σzz ) 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
489 of 607
another unloading/reloading cycle is going to be made.
Figure 20: Unloading/reloading cycle in a triaxial test using the General option
490 of 607
Derivation of Soil Parameters from
Lab Test Results & Verification in
Plaxis SoilTest
by
RF Shen
23 Nov 2011
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
491 of 607
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
492 of 607
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 Ø x 76 mm a = q + r
Singapore 2011
493 of 607
450
400
350
Deviator stress (kPa)
300
250
200 3’ = 100 kPa

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

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

strain curve (Triaxial)
Singapore 2011
494 of 607
Another common lab test is Oedometer Test
Oedometer setups in NUS

Geotechnical Lab
Settlement dial gauge
Oedometer Cell
Sample: dia. =75mm
Protruded
lever arm Height = 20mm
Heavy dead weights
Singapore 2011
495 of 607
Typical sample size 75 mm Ø x 20 mm
Boundary
conditions
0
0.1
0.2
0.3
0.4
Vertical strain (%)
0.5
0.6
0.7 Test data
0.8
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 400
Vertical pressure (kPa)
Vertical stress ~ vertical strain curve

(Oedometer)
Singapore 2011
496 of 607
400
300
200
100
Test data
0
0 50 100 150 200
Lateral stress (kPa)
Vertical stress ~ lateral stress curve

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

above typical lab test data
Singapore 2011
497 of 607
Part 1: Strength parameters
C’=0 for sand
450
400
Since c’ = 0 for sand, it can be
350
simplified to:
300
250
200 3’ = 100 kPa

150
100 Test data
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
500  100
sin  '   0.67
Axial strain
500  100
 '  42
Singapore 2011
498 of 607
0.06
0.05
So,
0.048
0.04
Volumetric strain
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
Test data sin  0.27
0
0 0.02 0.04 0.06 0.08 0.1
0.03 0.09
  16
‐0.01
Axial strain
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
499 of 607
Part 2: Stiffness parameters
450
400 400
350
300
250
3’ = 100 kPa
200
150
100 Test data
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
0.013
Axial strain
400
E50ref   30800 kPa  30000 kPa
0.013
Singapore 2011
500 of 607
450
400400
350
300
3’ = 100 kPa

250
200
150
100 Test data
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
0.021 0.026
Axial strain
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
501 of 607
0
0.1
0.2
0.3
0.33
0.4
Vertical strain (%)
0.5
0.6
0.7 Test data
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 
 c cot  ' p 
Singapore 2011
502 of 607
0
0.1
320
  29900kPa  30000kPa
0.2 ref
Eoed
0.3
0.4
1.4%  0.33%
0.47
Vertical strain (%)
0.5
400
0.6 200 kPa
Eoed   43000kPa
0.7
0.8
Test data
1.4%  0.47%
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 400
Vertical pressure (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
Jaki’s formula:
K 0NC  1  sin  '  1  sin 42  0.33

400
300
200
100
Test data
0
0 50 100 150 200
Lateral stress (kPa)
 x ' 100
K 0NC    0.33
 y ' 300
Singapore 2011
503 of 607
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
504 of 607
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
505 of 607
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
Test data
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
506 of 607
CIU stress path
Gradient:
350
Test data
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
507 of 607
Typically less test points are available

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

report as logv’ ~  yy which is linear (unlike sand) which
must be dealt with cautions!
Singapore 2011
508 of 607
Oedometer test for clay
0
d y '
Test data
Eoed 
0.1
d y
Vertical strain (%)
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
Test data
0.1
ref
Eoed  2.3  100  6.02  1350 kPa
Vertical strain (%)
0.2
Eoed  y '
0.27 ref

0.3
Eoed pref
0.37 m
0.4
ref  c cot  ' '1 
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
509 of 607
Eur refers to when 3’ =
0
100kPa
Test data
0.1
During oedometer loading,

Vertical strain (%)
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
Vertical pressure (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 Jaki’s formula:

K  1  sin  '  1  sin 25  .58
NC
0
Poisson’s ratio ur = 0.2
80
3’ = 100kPa for
consolidation,
During shearing, 3 = 0
70
60
Excess pore pressure
50 accumulates during
40 shearing  3’  100kPa
3’ = 100 kPa
30
Typically for NC clay, E50ref
20
may be about
Test data 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
Axial strain
Can we use the CIU test 1 ~ q test data to

derive the E50ref ??
Singapore 2011
510 of 607
Summary of Hardening Soil Parameters
FEM simulation using Plaxis SoilTest Facility
Singapore 2011
511 of 607
Let’s call it a day!
See you tomorrow...
Thank you!
Singapore 2011
512 of 607
21/11/2011
E3: Excavation Exercise
Tied-back excavation using

secant bored piles (CBP) and
2 layers of ground anchors
Presented By
RF Shen
24 November 2011
Briefing of the Project
513 of 607
21/11/2011
Proposed secant bored pile wall (SBP)
Dia. = 1 m with 200mm overlapping
Secant wall
d = 1 m
I = 0.049 m^4/m
E= 2.70E+07 kPa
A = 0.79 m^2/m
c/c spacing = 0.80 m
so,
EA/m = 2.65E+07 kN/m Take c/c spacing of 0.8m
EI/m = 8.28E+05 kNm^2/m Take c/c spacing of 1.6m
3
weight = 19 kN/m/m
Proposed ground anchors
The ground anchors are made of 32mm dia. Steel
bars at c/c spacing of 1m. The steel bar have a
stiffness of Es=2.1*10^8 kPa and with ultimate
strength of 605kN per anchor.
The anchors will be pre‐stressed to 60% of the
ultimatee strength, namely 363kN/anchor.
The properties of the grout body can be ignored.
514 of 607
21/11/2011
Proposed ground anchors

Anchor free length (node‐to‐node element with Elastoplasticity ):
d = 0.032 m
E= 2.10E+08 kPa
A = 8.04E‐04 m^2/m
so,
EA/m = 1.7E+05 kN/m
Max axial force = 605 kN/m
Anchor grout body (geogrid element with Elastoplasticity):
d = 0.032 m
E= 2.10E+08 kPa
A = 8.04E‐04 m^2/m
so,
EA/m = 1.7E+05 kN/m
Max axial force = 605 kN/m 5
Soil & ground water condition
The upper 40m of the subsoil consists of a more
or less homogeneous layer of medium dense
sand. Typical soil parameters based on triaxial
tests are presented in the next slide.
Underneath this layer there is a very stiff layer of
gravel which can be acted as the bottom
boundary of the 2D FEM mesh.
The ground water table is very deep and does not
play a role in this analysis.
6
515 of 607
21/11/2011
Soil parameters
Simulation in Plaxis 2D
version 2010
516 of 607
Tied-back excavation using HSsmall
TIED-BACK EXCAVATION
Using the HSsmall model
517 of 607
518 of 607
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
519 of 607
520 of 607
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 =2.7•107 kN/m2 with a specific weight γ=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.
521 of 607
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 γunsat 18.0 18.0 kN/m3
Saturated weight γsat 18.0 18.0 kN/m3
ref 4
Drained triaxial test stiffness E50 2.0·10 2.0·104 kN/m2
Drained primary oedometer Eref
oed 2.0·104 2.0·104 kN/m2
stiffness
Unloading/reloading stiffness Eref
ur 8.0·104 8.0·104 kN/m2
Power for stress-dependent m 0.5 0.5 –
stiffness
Cohesion c’ 1.0 1.0 kN/m2
Friction angle ϕ0 35 35 º
Dilatancy angle ψ 5 5 º
ref 4
Small-strain shear modulus G0 10.0·10 10.0·104 kN/m2
Threshold shear strain γ0.7 1.5·10−4 1.5·10−4 –
0
Unloading/reloading Poisson’s νur 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.
522 of 607
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
Poisson’s ratio ν 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
523 of 607
524 of 607
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.
525 of 607
• 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.
526 of 607
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
527 of 607
528 of 607
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 it’s 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.
529 of 607
• 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
530 of 607
Figure 10: Phase 3: Second excavation Figure 11: Phase 4: Activation and
prestressing of 2nd anchor
Figure 12: Phase 5: Final excavation
531 of 607
532 of 607
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.
533 of 607
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 γp that specifies the maximum
shear strain level reach in the stress history.
534 of 607
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
535 of 607
536 of 607
APPENDIX A: DETERMINATION OF SOIL PARAMETERS FROM

TRIAXIAL TEST
Figure 18: Triaxial test for sand layer
Strength parameters
Fill in σ1 and σ3 in the Mohr-Coulomb criteria:
σ1 − σ3 = (σ1 + σ3 ) sin ϕ + 2c cos ϕ
Since the cohesion will be small, assume c = 0:
σ1 −σ3
σ1 +σ3
= sin ϕ
370−100
370+100
= sin ϕ
ϕ = 35o
ψ =ϕ − 30 = 5o
For reasons of numerical stability, use c = 1 kPa
537 of 607
Stiffness parameters
Since excavation is considered in this exercise, the input of Young’s modulus E should be
based on unloading, rather than on primary loading. For the same reason, Poisson’s ratio
should also be based on unloading, which results in a somewhat lower value.
The triaxial test has a cell pressure σ 3 = 100 kPa. This corresponds with reference pressure,
so E50 = Eref
50 .
ref ∆σv 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
γ0.7 = 1.5 · 10−4
538 of 607
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 π·(0.5)4
Iwall = D
= πr
4D
= 4·0.8
= 61.3 · 10−3 m4 /m
With Econcrete = 2.7 · 107 kN/m2 this gives
EA = (2.7·107 )(0.93) = 2.5·107 kN/m

EI = (2.7·107 )(61.3·10−3 ) = 1.67·106 kNm2 /m
And for the weight:
w = γ·A = 16 · 0.93 = 15 kN/m/m
539 of 607
Geotextile reinforced embankment
GEOTEXTILE REINFORCED
EMBANKMENT WITH CONSOLIDATION
540 of 607
2 Computational Geotecnics
541 of 607
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 Young’s modulus Eu . The effective Poisson’s ratio should be chosen
such that a realistic K0nc is obtained in one-dimensional compression (K0nc = ν 0 /(1 − ν 0 ) ≈ 0.5).
The effective Young’s modulus is calculated from the shear modulus E 0 = 2G(1 + ν 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 ϕ0 =
30° and c’ = 3 kPa. K0nc is assumed at 0.5. For the stiffness one should take E’ = 4000 kPa
and ν 0 =0.33.
542 of 607
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
SCHEME OF OPERATIONS
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
543 of 607
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.
544 of 607
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 (ϕ and c) as well as the elastic parameters
(E’ and ν’) 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.
545 of 607
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 γunsat 13.5 13.5 18.0 18.0 kN/m3
Saturated weight γsat 13.5 13.5 18.0 18.0 kN/m3
Young’s modulus E 2667 2667 4000 40000 kN/m2
Poisson’s ratio ν 0.33 0.33 0.33 0.33 –
Cohesion c 8.0 8.0 3.0 3.0 kN/m2
◦
Friction angle ϕ 20.0 20.0 30.0 32.0
◦
Dilatancy angle ψ 0.0 0.0 0.0 2.0
Permeability x-dir kx 1.0·10−3 1.0 1.0 1.0 m/day
Permeability y-dir ky 1.0·10−3 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.
546 of 607
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
547 of 607
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.
548 of 607
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.
549 of 607
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.
550 of 607
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.
551 of 607
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 → ΣM 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 ΣM sf .
Addicionally, figures 13 and 14 show the failure mechanisms with the lowest factor or safety
for both the undrained and consolidated case.
552 of 607
Figure 11: Curve generation window
Consolidated: ΣMsf=2.1
Undrained: ΣMsf=1.4
Figure 12: Safety factor curve for reinforced embankment
553 of 607
Figure 13: Incremental displacements, undrained (phase 4)
Figure 14: Incremental displacements, consolidated (phase 8)
554 of 607
SUGGESTION FOR EXTRA EXERCISE: NON-REINFORCED

EMBANKMENT
SCHEME OF OPERATIONS
• 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.
555 of 607
Figure 16: Incremental shear strains, undrained (phase 12)
Figure 17: Incremental displacements, consolidated (phase 16)
Figure 18: Incremental shear strains, consolidated (phase 16)
Consolidated: ΣMsf=1.4
Undrained: ΣMsf=1.1
Figure 19: Safety factor curve for non-reinforced embankment
556 of 607
557 of 607
SUGGESTIONS FOR THE DETERMINATION OF THE CLAY

PARAMETERS
qc
su ≈ 15 = 150
15
= 10 kPa
su = 12 σx,0 + σ ,0
0 ,0 ,0
y sin(ϕ) + c cos(ϕ) with σx = K0 · σy ≈ (1 − sin(ϕ)) · σ y
In the middle of the clay layer at about 2m below ground level:
σy,0 = h · (γsat − γwater ) = 2 · 3.5 = 7 kPa =⇒ σx,0 = (1 − sin(20)) · σy,0 = 4.6 kPa
For this clay estimate ϕ = 20º, then c ˜ 8 kPa
Eu ≈ 15000·s
50
u
= 15000·10
50
= 3000 kPa
1 1
G = 3 Eu = 3 · 3000 = 1000 kPa
E 0 = 2G(1 + ν) = 83 · G = 2667 kPa
K0
ν = 1+K 0
= 0.5
1.5
= 0.33
Use ‘Undrained A’ as the type of material behaviour.
558 of 607
E5: Exercise on Pile Analysis
Based on an actual project: Pile

Foundations for Flieden Bridge in
Germany
Original exercise made by Dr Yasser El-Mossallamy

(ARCADIS Consult, Germany)
Presented By
RF Shen
25 November 2011
Exercise SG-E5-Shen
559 of 607
y
X = ‐35m
to 35m
X
(‐4 ‐25 0)
(‐.4 ‐25 ‐.8) z y = ‐25m
(.4 ‐25 ‐.8) X to 25m
(4 ‐25 0)
z = 0 to
‐30m
3
The subsoil consists mainly of tertiary
formations of highly plastic clay with lenses of
lignite coal (clay with brown coal). In this
analysis, a uniform clay layer was idealized with
OCR = 1.3
4
Exercise SG-E5-Shen
560 of 607
Soil parameters
Simulation in Plaxis 3D
Step 1: General setting
Step 2: Add in a borehole
Step 3: Define soil properties
Step 4: Create 6 piles
Step 5: Create 1 pile cap
Step 6: Clone another pile group
Step 7: Create the trench
Step 8: Assign vertical loads
Step 9: Generate mesh with refinement
Step 10: Define stages and view results 6
Exercise SG-E5-Shen
561 of 607
Flieden bridge piled-raft foundation
FLIEDEN BRIDGE
PILED-RAFT FOUNDATION
Original excercise made by

Dr. Yasser El-Mossallamy
ARCADIS Consult. Germany
Computational geotechnics 1
562 of 607
INTRODUCTION
The foundation of the 4-span railway bridge of Flieden in Germany (figure 1) was the first
railway bridge in Germany founded on piled rafts.
Figure 1: Geological conditions of the Flieden railway bridge
The subsoil consists mainly of tertiary formations of highly plastic clay with lenses of lignite
coal (clay with brown coal). To ascertain the adequacy of the piles and determine appropri-
ate design values, pile load tests were first conducted on large diameter bored piles with and
without post shaft grouting (El-Mossallamy et al. 2003). These results conform to the me-
chanical sensitivity of the organic silty clay and lignite coal lenses. It was decided to install all
foundation piles applying post shaft grouting.
INPUT
The bridge piers are consisted of two pillars, each founded on a separate group of 6 piles
underneath a raft. The pile arrangements are shown in Figure 2. The rafts are 1.5 meters
thick and are embedded in the soil with the raft base at a depth of 2.3 meters below the soil
surface. The piles where designed with a diameter of 1.2 m and a length of 18 m. The pillars
transfer two working loads of 20 MN and 22 MN respectively from the superstructure to the
piled raft foundation.
Work flow
In this excercise the model is created in a specific order that has proven to be a rather efficient
way to create the model. Please note that many parts of the model can be created in any
other order as well and the work flow presented here is not the only correct method to create
the project. The work flow to create the project presented here is:
1. Enter dimensions of the project and some general visualisation options
2. Define the underground model using 1 borehole and the appropriate soil material sets
563 of 607
Figure 2: Geometry of the piled raft foundation
3. Insert 1 pile in the model
4. Copy this 1 pile 5 times to create the 6 piles needed for 1 piled raft
5. Insert the raft, the lower column and the top load
6. Copy the complete piled-raft 1 time to create the second piled raft
7. Create an extra zone for mesh refinement around the piled-rafts
8. Generate mesh
564 of 607
Geometry
General settings
Start the PLAXIS 3D input program. A Quick select project dialog box will appear in which
you can select an existing project or create a new one; choose Start a new project so that the
Project properties window appears.
1. In the Project properties window on the Model tabsheet the size of the model contour
has to be set. In the Contour box fill in xmin = −35, xmax = 35, y min = −25 and y max = 25.
2. Close the Project properties window, the drawing area will now appear.
3. From the Options menu choose Visualization settings. A new window will open, contain-
ing 2 tabsheets: View and Visibility.
4. On the View tabsheet the grid point distance (Spacing) and number of snap intervals per
grid distance can be set. By default the Spacing is set to 1 m with only 1 snap interval
per grid distance. As can be seen from figure 2 many dimensions of this project have an
accuracy of 0.1 m and therefore just 1 snap interval per 1 m is not sufficient. Therefore,
set the Intervals to 10, this will results in having a snap distance of 0.1m (Spacing /
Intervals).
5. Close the Visualization settings window.
Subsoil
The first step in creating a model in PLAXIS 3D is the definition of the subsoil, which is done
using boreholes.
1. Select the Create borehole button ( ) and move the mouse to the origin of the system
of axis. Click at (x,y,z) = (0 0 0), this will open the Modify soil layers window.
2. In the Modify soil layers window click the Add button in order to define a new soil layer
in this borehole. Set the top of the borehole to 0.0 m and the bottom to -30.0 m.
3. In order to assign a material set to the newly defined model it is necessary to first define
a material set. To do so, press the Materials button ( ) to open the material
sets database.
4. Though the model only has one soil layer (clay) we will have to define two material sets:
the second material set will be used to represent the concrete needed for both raft and
piles. Therefore, create two material sets according to the material parameters specified
in table 1.
5. After defining the two material sets close the window by clicking OK in order to return to
the Material sets window.
565 of 607
Table 1: Parameters for the clay layer and concrete slab

Parameter Name Clay Concrete Unit
Material model Material model Hardening Soil Linear Elastic -
Type of behaviour Drainage type Drained Non porous -
Unsaturated soil weight γunsat 20.0 24.0 kN/m3
Saturated soil weight γsat 20.0 - kN/m3
Young’s modulus Eref - 30000 M N/m2
ref
Drained triaxial test stiffness E50 45.0 - M N/m2
ref
Primary oedometer stiffness Eoed 45.0 - M N/m2
Unloading/reloading stiffness ref
Eur 135.0 - M N/m2
Power for stress-dependent stiffness m 0.9 -
Poisson’s ratio ν - 0.3 -
Unloading-reloading Poisson’s ratio νur 0.2 -
Cohesion c’ 10 - kN/m2
Friction angle ϕ 30 - o
Dilatancy angle ψ 0 - o
Permeability kx , k y , k z 0 - m/day
Interface strength Rinter 0.6 (Manual) Rigid -
Coefficient for initial lateral stress K0 Automatic Automatic -
Overconsolidation ratio OCR 1.3 - -
6. Drag and drop the clay material set from the Material sets window onto the borehole. The
mouse cursor changes shape when the material set can be dropped. After dropping the
borehole should get the colour of the material set. Now close the Material sets window
in order to return to the Modify soil layer window.
7. In the Modify soil layer window directly above the graphical representation of the bore-
hole it is possible to specifty a general phreatic level for this borehole by changing the
Head value. In this project the water level is 0.5 meters below ground level, therefore
change the Head to -0.5 m.
8. Press OK to close the Modify soil layers window and return to the drawing area. In the
drawing area there is now a block of soil with the horizontal dimensions specified in the
Project properties window and a depth according to the borehole.
We have now finished defining the subsoil and we will continue defining the foundation. Press
the Structures option ( ) on the mode toolbar to move to Structures mode.
Create foundation structures

Create piles
The two bridge foundations are equal with exception of the load from the bridge acting on
the foundation. Therefore it’s sufficient to define 1 foundation and then make a copy of the
566 of 607
foundation to get the second. Similarly, each foundation is supported by six equal piles, hence
it is sufficient to define 1 pile and make 5 copies to model all piles to model 1 foundation.
In the current version of PLAXIS 3D the only possibility to insert a pile is by inserting a cylin-
derical volume using the command line. The syntax for inserting a cylinder is:
cylinder <R> <L> <num planes> (<start_x> <start_y> <start_z>) (<dir_x> <dir_y> <dir_z>)
In short, one specifies the radius (R) and length (L) of the cylinder, a set of 3 coordinates to
indicate the starting point of the cylinder and a vector to indicate the direction of the cylinder.
Special attention should be given to <num planes>. In PLAXIS 3D a cylinder is modelled with
a polygon cross section, hence <num planes> gives the number of sides of the polygon. The
higher the number the more accurate the polygon will represent the circular cross section.
9. Insert the first pile at (x,y) = (-8.4, -1.8). Note that the piles have a 1.2m diameter (hence
a radius of 0.6m), are 18 meters long, start at z = -2.3m and go down vertically, that is
in the negative z-direction. The number of planes is set to 15 to accurately model the
cylinderical shape. This results in the following cylinder command:
cylinder 0.6 18 15 (-8.4 -1.8 -2.3) (0 0 -1)
Type this command on the command line and press <Enter>. The cylinder is now in-
serted in the model as a volume.
10. In order to assign interfaces around the pile, the pile has to be split into its separate
surfaces. To do so, right click on the pile and from the popup menu choose Decompose
into surfaces.
11. Now select the outer surfaces of the pile, right-click and select Create negative interface.
This will create a negative interface along the outside of the pile.
12. In order to create an interface below the foot of the pile, select the bottom circular surface
of the pile. It is probably necessary to rotate the model in order to see the foot of the pile
from below. Right-click again and select Create negative interface to create the interface
below the foot as well.
Hint: Interfaces are drawn as planes at a certain distance from the surface they
belong to. Therefore, if a project requires a lot of interfaces it may become
difficult to see the underlying structure as the interfaces are surrounding it.
This can be solved by either reducing the distance between interface and
structure or by making the interfaces invisible.
The distance between interface and surface can be reduced in the
Visualization settings that can be found under the Options menu. On the
View tabsheet the field Interface size controls the distance. By default this
value is set to 1. Reducing this value will reduce the distance between
interface and surface.
Alternatively, in the Object explorer it is possible to make the interfaces
invisible by clicking on the small eye in front of the branch Interfaces (to
make them all invisible) or in front of individual interfaces (to make only a
selection of interfaces invisible).
567 of 607
We have now finished creating the first pile. The next step is to make 5 copies of the pile to
create the group of 6 piles of the first foundation slab.
13. Click the button Select rectangle ( ), ignore the suboptions that become available.
Now draw a rectangle that fits the whole pile so that all parts of the pile are selected.
14. Now click the Create array button ( ) to specify the locations of the copies of the pile.
The Create array window appears, see figure 3.
Figure 3: Copy the pile by creating an 2-dimensional array of piles
In x-direction we need 3 piles with an intermediate distance of 3.4 meters and in the y-direction
we only need 2 piles with a distance of 3.6 meters in between.
15. Set the Shape of the array to 2D, in xy plane as we want to copy the piles in both x and
y direction, keeping the z coordinate constant
16. Fill in 3 columns with a distance of x = 3.4m in between and 2 rows with a distance of y
= 3.6m in between.
17. Press OK to copy the pile to the specified locations.
We have now created the 6 piles for one of the bridge foundations.
Create first raft
After creating the 6 piles now the raft has to be modelled on top of the piles, including the
lower part of the column supporting the bridge:
568 of 607
1. From the horizontal button bar with general options, click the Top view button ( ). This
will show the model seen along the z-axis.
2. In the Movement limitation window that appears, fix the z-coordinate to z = -0.8m by
filling in -0.8 in the z-value field and clicking the Set button.
3. Select the Create surface button ( ) and draw the surface representing the top side of
the raft from (x y) = (-9.6 -3.0) to (-9.6 3.0), (-0.4 3.0) and (-0.4 -3.0).
4. Select the surface that has just been created and click the Extrude button ( ). In the
window that opens fill in an extrusion vector of (x,y,z) = (0 0 -1.5) in order to create the
1.5m thick raft and click OK.
Now the raft has been created as volume, in order to assign interfaces to all sides of the raft,
the raft volume has to be decomposed into its surfaces.
5. From the button bar with general options, click the Perspective view button ( ). .
6. Right-click on one of the vertical sides of the raft and select the option Decompose into
surfaces. This will created surfaces for all sides of the volume.
7. For all 6 sides, right-click on the side and add an interface. Note that all sides need
a negative interface with exception of the vertical side at y = 3.0m; this side needs a
positive interface. Check if all created interfaces are on the outside of the raft!
8. In order to make the lower part of the supporting column, click again the Top view button
and fix the z-coordinate to ground level.
9. Create a surface from (x y) = (-6.0 -1.0) to (-6.0 1.0), (-4.0 1.0) and (-4.0 -1.0).
10. Extrude the surface 0.8 meters downwards, hence in the negative z-direction. This cre-
ates the lower part of the column from groundlevel down to the raft.
11. Decompose the column into surfaces.
12. For all 4 vertical surfaces created, create an interface on the outside. That is, negative
interfaces for all vertical sides but the vertical side at y = 1.0m. The latter side needs a
positive interface.
The only part missing now is the load representing both the weight of the bridge and a passing
train
13. Right-click on the top plane of the column, that is the plane at ground level.
14. From the popup menu that opens, select the option Create surface load to add the load.
The first raft is now complete.
569 of 607
Create second raft
The second raft is equal to the first raft, hence creating the second raft is simply making a
copy of the first raft:
1. Click the button Select rectangle ( ), ignore the suboptions that become available.
Now draw a rectangle that fits the whole structure of piles, raft and column so that all
parts are selected.
2. Now click the Create array button ( ) to specify the location of the copy of the founda-
tion structure in the Create array window.
3. Set the Shape of the array to 1D, in x direction as we want to copy the foundation just
one time in x direction, keeping the y and z coordinates constant
4. Fill in 2 columns with a distance of x = 10m in between and press OK. Now the second
raft is created as copy of the first raft.
Both rafts have now been defined, see figure 4.
Figure 4: Geometry containing the two rafts
Create mesh refinement area
In order to be able to refine the mesh in the area around the rafts it is needed to define a
volume of soil around the rafts where a mesh refinement can be applied. To do so, follow
these steps:
1. Select the Top view and fix the z-coordinate to -25.0 m
570 of 607
2. Draw a rectangular surface from (x y) = (-10.0 -4.0) to (-10.0 4.0), (10.0 4.0) and (10.0
-4.0).
3. Select the Perspective view, select the newly created surface and extrude it 25m up,
hence in the positive z-direction.
We have now created a volume around the foundation structure that we can use for local mesh
refinement.
Mesh generation
In the Mesh mode we will specify global and local refinements and generate the mesh. In order
to generate more accurate results a refinement of the mesh around the foundation structures
will be applied.
1. In the geometry click somewhere close to the origin. This will select the body of soil that
encloses the foundation structures.
2. In the Selection explorer on the left the selected soil body appears, showing a mesh
refinement factor of 1.0. Change this mesh refinement factor to 0.30.
3. Select the Generate mesh button ( ) in order to generate the mesh. The Mesh options
window appears.
4. In the Mesh options window choose a Very coarse element distribution and click OK to
start the mesh generator.
5. After mesh generation has finished one can already see an indication of the amount of
elements and nodes generated in the command line box below the draw area. For this
project about 22,000 elements should be generated.
6. Click the View mesh button ( ) to inspect the generated mesh.
After inspecting the mesh the output window can be closed. Mesh generation has now been
finished and so creating all necessary input for defining the calculation phases has been
finished.
571 of 607
Figure 5: Generated mesh with local refinement
572 of 607
CALCULATION
The calculation consists of the initial phase and three additional phases. Since water levels
will remain constant the Water levels mode can be skipped. Therefore, click on the Staged
construction mode button to move to the defintion of the calculation phases.
Initial phase
By default the Initial phase is set to the K0 procedure, which is fine for this example. No further
changes have to be made.
First phase - construction of the foundations
1. Click on the Add phase button ( ) to add the first calculation phase.
2. As the foundations are surrounded by soil they cannot be accessed directly. In order to
change their properties the surrounding soil has to be made invisible. To do so, right-
click on the soil somewhere far away from the origin and from the menu that pops up
choose Hide to hide the outer soil. Now only the foundations and the refinement zone is
left. Make sure the soil is hidden, not deactivated!
3. Right-click on the refinement zone volume and again choose the Hide option from the
popup menu. With the refinement zone hidden, only the foundations structures remain
visible.
4. Open the material sets database by clicking the Show materials button ( ). Drag and
drop the material set representing the concrete on all piles, the rafts and the two parts
of the column. When assigning the material set, the colour changes from the colour of
the material set representing the clay to the colour of the material set representing the
concrete.
5. In the Model explorer, activate all interfaces by clicking on the checkbox in front of the
interfaces branch so that a checkmark appears.
Second phase - working load
1. Click on the Add phase button ( ) to add the second calculation phase.
2. In the Model explorer open the Surface loads branch and change the value for the two
surface loads. Set the first surface load to a vertical stress of σz = −5000 kN/m2 (20 MN
dived by 4 m2 cross sectional area of the column) and set the second surface load to a
vertical stress of σz = −5500 kN/m2 .
3. Make sure the surface loads are activated, that is that they have a checkmark in the
Model explorer.
573 of 607
Third phase - ultimate limit load
1. Click on the Add phase button ( ) to add the third calculation phase.
2. In the Model explorer change the values of the Surface loads to σz = −10000 kN/m2 for
the first surface load and σz = −11000 kN/m2 for the second surface load.
Press the Calculate button ( ) to start the calculation. Ignore the message "No nodes or
stress points selected for curves" as we will not draw any load-displacement curves in this
example, and continue the calculation.
574 of 607
OUTPUT RESULTS
Figure 6 demonstrates the calculated load settlement behaviour of the piled raft applying the
GAPR (Geotechnical Analysis of Piled Raft, El-Mossallamy 1996). Due to the non-linear re-
sponse of the foundation system the loads have been incrementally applied till the ultimate
limit state. Another aim of the analysis under working loads was to determine the pile/soil
stiffness and subgrade reaction distribution beneath the raft, which are necessary for the de-
sign of the foundation. However, within the framework of this exercise the subgrade reaction
distribution will not be checked. Figure 7 shows the measured settlements in comparison to
the calculated values
Figure 6: Load-settlement behaviour of the piled raft foundation (calculated by program GAPR,
El-Mossallamy)
Figure 7: Measured settlements
575 of 607
3D Excavation Supported By Struts
Three dimensional finite element modelling using
Plaxis 3D

William Cheang, Shen Rui Fu & Tan Siew Ann
11/20/2011
Three dimensional finite element modelling of a 50 x 10 excavation supported by two levels of struts.
Excavation depth is 8.5m and the toe of the retaining structure is located at 12 m below ground level. Struts
spaced at 10m centre‐to‐centre spacing.
576 of 607
INTRO
ODUCT
TION
Figurre 1: The model

m
In tthis exercisse a long excavation
n of 50m le
ength, 10m
m width annd 8.5m de
eep will be
e
modelled. The analyysis investigates the difference between a 2D and 33D model. The result
of from the
e 2D analyysis is included in th
he result se
ection of th
his exercisse for comparison. In
n
this finite e
element model
m the fo
ocus is on creating th
he model in
i 3-dimennsional spa
ace, setting
g
the variou
us stages to simulate the wh
hole construction pro
ocess andd at each stage the
e
excavated
d zone is dewatere
ed. The a
analysis will
w look into the sshort-term condition.
Dewatering is done
e with the nterpolate within walls
e “cluster dry” optiion and in w below
w
n level to general
excavation g ph
hreatic leve
el. Time in
nterval of 5 days forr each stag
ge to allow
w
plotting of curves forr compariso
on
A. Ge
eometry
y
1. The
e geometryy of the exe
ercise is sh
hown in Ta
able 1.
2. The
e excavatio
on is 50m in 10m in width and 8.5m deep.
i length, 1
3. The
e toe of the
e diaphragm
m wall is lo
ocated at 12m
1 below
w the grounnd.
4. The
e excavatio
on is done in three sttages and supported
d by two leevels of struts spaced
d
at 1
10m centre
e-to-centre distance a
along the le
ength.
each level there are 5 numberss of struts.
5. At e
6. The
e ground co
ondition is simple an
nd it is a sin
ngle layere
ed problem
m.
577 of 607
B. Dime
ensions
s
1. The
e size of th
he models and
a dimen
nsions of th
he excavation is indiccated in Ta
able 2.
2. The
e size of th
he finite ele
ement mod
del in Plax
xis 3D is se
et at 100 x 60 x 22m
m in length,
width and dep
pth respecttively.
Figure
F 2: M
Model size in
i Plaxis 3D
578 of 607
C. Flow
w of worrk
The
e flow of wo
ork consistt the follow
wing steps. In genera
al the stepss will be:
1. Settting up of the
t projectt informatio
on and model size in “Project S
Settings”.
2. In M
Mode-Soil the depth of the mod
del is set and
a the initial positionn of the wa
ater table iss
set..
3. In M
Mode-Structures the dimension
ns of the excavation
e n and excaavation lev
vels will be
e
con
nstructed. This
T will be
b followed
d by the creation of the retain ing wall and supporrt
sysstem.
4. In M
Mode-Mesh
h the mode
el is discre
etised. Refinement will
w be madee on imporrtant zoness
to e
enhance th
he analysis
s and resul ts.
Mode-Watter Levels the wate r condition
5. In M n at each stages oof excavatiion will be
e
seq
quentially lo
owered using the ‘Clu
uster Dry’ option.
6. In Mode-Stag
ged Cons
struction th
he excava
ation proc
cess and constructiion of the
e
reta
aining syste
em will be made. Fin
nally the ca
alculation will
w be madde.
7. Insp
pect the ca
alculated re
esults in O
Output and construct displacemeent chart versus
v time
e
usin
ng the Cha
art option.
ANAL
LYSIS
A. PR
ROJECT
T SETTINGS
Fiigure 3: Qu
uick Selectt option (Se
elect “Starrt a new project” for thhis exercis
se)
579 of 607
Figure 4: S
Setting the size
s of a Pla
axis 3D mo
odel via “Mo
odel Tab”. The
T point off origin is automaticallyy
set at the ccentre of the
e model (x-o
origin = 0, yy-origin = 0) when using
g the abovee stated valu
ues.
1. The
e size in x-y space is set using the “Conto
our” input box.
b
2. The
e dimensio
ons in x-y space
s are 100 x 60 m in length
h and widtth. The min
nimum and
d
maxximum vallues are chosen as such that the point of origin w
will be loca
ated at the
e
cen
ntre in x-p space.
s
3. The
e units and
d general quantities fo
or the anallysis are given in the above figu
ure.
4. The
e dimensio
ons in ‘z-direction is sset using the
t ‘Borehole’ input ooption in ‘M
Modify Soiil
Layyers’ box.
5. Oncce the inpu
ut is complete click ‘O
OK’.
580 of 607
B.MODE: SOIL
L
Plaxis 3D has 5 main mo
odes to deffine the prroject, that is Soil, Sttructures, Mesh,
M
Wate
er Levels and
a Stage
ed Constru
uction. Eve
ery mode has a deedicated ve
ertical
toolbar located between the explorrer boxes and
a drawin
ng area. Inn Soil mod
de we
define the soil layers for projects. In
n this curre
ent case th
here is onlly a single layer
of so
oil. Informa
ation of this soil laye
er and loc
cation of th
he water ttable is en
ntered
using
g a single borehole. To define
e use the borehole
b option
o to deefine the single
s
soil la
ayer, follow
w these ste
eps:
1. Selecct the Bore
ehole tool ffrom the ge
eometry toolbar.
at represents the soil.. It is suggested
2. Clickk at a location in the ccluster tha
to cliick on (x; y) = (-50; -30) but any
a other position
p woould be fin
ne as
well. This place
es a boreh
hole at loc
cation (x; y)
y = (-50; --30) and opens
o
the Modify
M soil layers win
ndow (Figure 5). The Modify soiil layers wiindow
provides a grap
phical reprresentation
n of the cu
urrent boreehole on th
he left
and a spreads
sheet for d
data inputt. However when crreating the
e first
hole in an
boreh a analyssis both the
t graph
hical repreesentation and
sprea
adsheet is empty.
Figure
e 5: Modify soil
s layers window
w
581 of 607
3. Presss Add button to defin
ne a new la
ayer in the borehole. By default both
the top
t and bo
ottom bou
undaries off the bore
ehole are set to z = 0.0.
Chan
nge the bo
ottom boun
ndary to -2
22m. This action in effect is setting
s
the depth
d of the
e finite elem
ment mode
el (see Figure 6)
4. To define
d the material properties
s for this layer of soil, press the
Mate
erial button. The Mate
erial sets window
w will appear
Fig
gure 6: Defin
ning a single
e layered soil using a single
s borehhole
Table
e 1: Material Properties
s for soil
Param
meter Symbol Soil nit
Un
Soil M
Model Name Mohr-Coulomb -
Drain
nage type Undrained (B) -
Unsaturated soill weight unsat 18.0 kN//m3
Saturrated soil we
eight sat 19.0 kN//m3
Young’s moduluss E-ref 15000 kN//m3
Poissson’s ratio Nu 0.25 -
Cohe
esion Su 30 kN//m2
Frictio
on angle  0 -
Interfa
faces Rigid -
Coeffficient of earth pressure
e at rest Ko Man
nual = 0.55 -
582 of 607
6. Creatte a new material set ffor the soil material ac
ccording to the data given in
Table
e 1. Any parrameters no
ot mentioned should be
e left at thei r default va
alue.
7. After entering the material p
properties for
f the soil close
c the w
window by clicking
OK in
n order to re
eturn to the Material se
ets window.
8. The material properties a d into the model byy ‘drag-and
are entered d-drop’
oach onto the boreho
appro ole layer (s
see Figure 7). Once this is don
ne the
desig
gnated soil layer in th
he borehole
e should be
e indicatedd with the colour
chose
en for this material
m set..
Figure
re 7: Boreho
ole with matterial properrties properrly in place.
9. By de
efault the water-table
w is set at 0 m. In this example thhe water ta
able is
ed at ground level and therefore no
locate n alteration
n is neededd. Therefore
e keep
the ground level and there
efore no alteration is needed.
n Th erefore kee
ep the
d at 0 m.
Head
We havve now fin
nished de
efining the subsoil and
a we w
will continu
ue to
Structure
es Mode to
o define the
e geometrry of the ex
xcavation.
C. MO
ODE: ST
TRUCTU
URES
Press tthe Structu
ures option
n on the m
mode toolb
bar to mov
ve to Strucctures mod
de. We willl
create the excavvation size by constru
ucting a vo
olume. This volume w
will be loca
ated at the
e
point o
of origin of the
t model..
583 of 607
To construct the diaphragm wall we will decompose this volume into surfaces. We
will change the surfaces located at the four sides of the volume to plate elements. Also
interface elements will be introduced by using the same approach. Finally we will
construct 3 excavation levels by creating 3 horizontal surfaces at 3 locations along the
z-axis.
The 5 individual struts located at a given level are constructed using Beam elements.
One single strut is drawn and set with the corresponding material properties. This will be
copied and replicated using the Array tool along the x-direction at 10 m centre-to-centre
distance. The waling system located along the perimeter of the retaining wall will be
constructed using Beam elements. The struts and walers located at the first level will be
grouped together to form ‘support system level 1’. This is to allow easy identification and
activation during the staged construction process.
Again by using the Array tool all the struts and waler system (support system level 1)
located at the first level will be copied to the second level and this therefore forms
‘support system level 2’. Create the following material sets and properties listed in Table
2, 3 and 4.
Table 2: Diaphragm Wall

Element type Plate Unit
Identification Name D-wall -
Thickness d 0.6 M
Weight  24 kN/m3
Behaviour Linear -
Isotropic Yes -
Modulus of elasticity E1=E2 30e+6 kN/m2
Table 3: Strut
Element type Beam Unit
Identification Name Strut -
Area A 1.225E-1 m2
Weight  78.5 kN/m3
Behaviour  Linear -
Modulus of Elasticity E 210.0E6 kN/m2
584 of 607
Second moment
m are
ea I1=
=I2 1.800E-2
1 m4
Table 4: Waling
Element type Beam Unit
Identifica
ation Na
ame Waling -
Area A 8.682E-3
8 m2
Weight  78.5 kN/m33
Behaviou
ur  Linear -
Modulus of Elasticity
y E 210.0E6
2 kN/m22
Second moment
m are
ea I1=
=I2 1.045E-2
1 m4
eate the ex
C1.Cre xcavation volume
Figure
e 8: Switch
h to Top Viiew (Plan vview) and movement
m limitation w
window ap
ppears
To expedite the
t constru
uction of th
he foot prin
nt of the excavation we switch
h to the top
p
view
w mode byy clicking the
t tool loccated on the menu. The viewppoint will be switched
d
to p
plan view. A ‘Movem
ment limitatiion’ window
w appears
s (see Figuure 8). Usin
ng this too
ol
it iss possible to
t fix the drawing
d pla
ane at a sp
pecific elev
vation. In tthis case th
he drawing
g
plan
ne is fixed at z = 0 m).
m This is to
o facilitate drawing of
o the excavvation foottprint
585 of 607
e volume is created by firstly d
The drawing a surface in x-y planee at z= 0 m and then
n
extrruding dow
wnwards by 12 m ussing the ex
xtrusion too
ol. From thhe vertical tool bar in
n
Stru
uctures Mo
ode select the Create
e surface option.
o
1. Construct a surface of 50 (L) a
and 10 (W)) m in size by locatinng the four points of a
rectangula
ar at the fo
ollowing c oordinates
s. A clockw
wise or annti-clockwis
se scheme
e
can be cho
osen (see Figure 9 a
and 10)
Table 5: Location
L of the surfa ce points at
a z=0m
Poiint 1 Poiint 2 Poin
nt 3 Poin
nt 4
x y x y x y x y
-25 -5 -25 5 25 5 25 -5
Figure 9
9: Construcct surface sttarting from
m point 1 to 3 ( and finallly 4) in clocckwise direc
ction
586 of 607
Figure 10: Constru
uction of surrface at z = 0 m (comp
pleted). The position off the points can
c still be
fine tun
ne using the
e table inputt seen here at the top right
r corner..
After inserrting the surface rep
presenting the cross section off the excavvation in plan
p we willl
extrude this surface to form the
e size (dep
pth) of the retaining structure
s (D
Diaphragm wall).
2. From the horizontal button ba r, click the

e Perspecttive view bbutton. This
s will show
w
the model view to iso
ometric.
3. The create
ed surface is marked
d in dark grrey. Selectt the surfacce so that it becomess
red.
4. Click the Extrude
E button and th
he Extrude
e window appears (seee Figure 11)
1
Figure
re 11: Extrussion along z-axis
z using
g a vector le
ength of 12 (vector z = -1)
587 of 607
5. Create a volume
v by filling an e
extrusion le
ength of 12
2m. To extrrude down
nwards that
is beneath
h the groun
nd surface the extrusion vector is z = -12 (see Figurre 11).
6. Click Applly to confirm
m the extru
usion process.
C2.Cre
eate the re
etaining sttructure w
with interfa
aces
1. Ensure tha
at the view
w is set as P
Perspectiv
ve
2. Select Vollume 1 by clicking th
he volume
e in the dra
awing areaa. Once se
elected the
e
volume is marked in red and in
n the Selec
ction explorer Volumee 1 is indic
cated.
3. Using the ‘Right Mo
ouse Butto
on (RMB)’’ gesture turn the vvolume into surfacess
using ‘Deccompose in
nto surface
es’ tool (se
ee Figure 12).
Figure 12: Decomp

posing a vo
olume into ssurfaces usiing the RMB
B gesture
4. Six surface
es will be formed
f aro
ound the original volu
ume (in thiss case Volume 1) but
we will ne
eed on the
e 4 vertica
al surfaces
s to form the
t diaphrragm wall by turning
g
these 4 surfaces
s in
nto plates. One may
y choose to delete the top and
a bottom
m
surfaces to
o get a cle
ean model or mainta
ain these knowing thaat we can modify the
e
elevation of these 2 horizonttal surface
es into the
e excavatioon levels. The latter
option is straight
s forw
ward for exxperienced
d user.
5. In this exe
ercise we will
w use the
e first apprroach that is to deletee the top and
a bottom
m
surface. Select
S the to
op and botttom surfac
ce and delete them.
588 of 607
6. Select the
e 4 vertical surfaces located along the perimeter
p oof the volume. To do
o
this, click each of the
t vertica
al surfaces
s (while prressing thee ‘ctrl’ buttton on the
e
keyboard).
7. Perform the RMB gesture a
and select ‘Create plate’.
p Youu should obtain the
e
following schematic
c as show
wn in Fig
gure 13 showing
s thhe selecte
ed vertica
al
surfaces.
Figure 13: Turning

g the vertica
al surfaces located alo
ong the perrimeter into plates usin
ng the RMB
B
e.
gesture
8. On a succcessful con
nversion th
he 4 surfaces will be
e become plates. Th
hese platess
odel the diiaphragm walls. Pla
will be ussed to mo aced the m
material se
ets for the
e
diaphragm
m wall using
g the drag--and-drop procedure
e (Figure 144).
589 of 607
Figure 14: The plate elements
e rep
presenting the
t dimensiions of the D
D-wall.
9. To create the interfaces at th

he outer and inner side
s of thee diaphrag
gm wall we
e
again use the same
e procedurre of selec erimeter of
cting the surfaces aloong the pe
the volum
me. (Again
n you can
n use the
e select-w
while-pressiing-the-ctrl
rl-button to
o
accumulatte all the 4 surfaces)..
10. Using the RMB gestture selectt ‘create po
ositive inte
erfaces’. Innterface ele
ements willl
be created
d on the po
ositive side
e of a ‘plate
e’( see figu
ure 15)
11. A positive on of side face whicch is on th
e interface refers to the locatio he positive
e
direction when
w referrred to the llocal axis (see
( figure 15a)
12. Positive or negative interfacess do not re
efer to reta
ained or exxcavation side of the
e
retaining wall
w (plate)) but is refe
erred to th
he direction
n of z-direcction in the
e local axiss
of the plate. To cons
struct interffaces along the plate
e both optioons are ne
eeded. The
e
d interface model is sshown in Figure 16.
completed
590 of 607
Figure 15a Figure
e 15b Figure 15c
Figure
re 15: Locall axis on a plate
p (z axiss which is in
ndicated in blue, Fig.1 5a). A positive value iss
the ooutward direection. A neegative valuue is the innward direction. By seelecting ‘cre
eate positive e
interfa
face’ interface elementts will be crreated on thhe positive side (Fig.155b). Similarrly to create
e
an inwward interfa
ace the ‘crea
ate negative e interface’ option is ch
hosen (see Fig.15.c)
Figure
re 16: Com mplete interfface elemen nts along the
t perimetter of the pplate using the ‘create
e
positiive interfacces’ and ‘c create neg gative interffaces’ apprroach. Thee diagram shows the e
directtion of ‘local blue axis’ along the p
plate.
eate the ex
C3.Cre xcavation levels
To constru
uct the stag
ged excava
ation proce
ess 3 horiz
zontal surfa
aces locateed at (z = -3m),
- (z = -
6m) and (zz = -8.5m)) will be co
onstructed.. The first surface is drawn butt the subse
equent two
o
surfaces a t ‘Array’ tool.
are formed by using the
1. Construct a surface
e using the
e same po
oints as in
ndicated inn Table 5 but with z
et to -3 m.
position se
2. Change to
o ‘Top view
w’ and the ‘Movemen
nt limitation tool’ apppear again. Set the z
value to -3
3.0 m.
591 of 607
ate surface
3. Use ‘Crea e’ tool loca
ated in the
e vertical toolbar annd draw a horizonta
al
surface ussing the co
oordinates given in Ta
able 5. This is the firsst excavation level to
o
3 m below
w ground le
evel (see F igure 17)
Figure 17: Location of

o a horizon
ntal surface at z = -3m
m. This is thhe extent of
o first stage
e
excavation. The soil volume
v with
hin the exccavation an
nd plates arre switchedd to 30% to o
facilitate vie
ewing
4. To constru
uct the exc
cavation sta
age 2 the position off the seconnd surface will be at z
= -6m. The
e firs surfa
ace will be ccopied and
d place at z = -6m. T
The verticall difference
e
between th
he first and
d second ssurface is 3m.
3
5. Select the first surfac
ce (after se
election it will
w be marrked red).
6. Select the
e ‘Array tool’ from th
he vertical toolbar. The
T ‘Creatte array window’
w willl
appear as shown in Figure 18. By default the array
y pattern is ‘rectangullar’.
7. In the ‘Co
onfiguration’ section
n choose ‘1D,
‘ in z direction to copy th
he chosen
n
surface an
nd place th
he second horizontal surface at
a z = -6m
m. To perfo
orm this we
e
set the ‘Diistance between colu
umns’ as -3.
- The ‘ne
egative’ vallue is need
ded to shifft
the copied
d surface in the dirrection opp
posite to the
t global z-directio
on (upward
d
positive). A value off 3 is the d
difference in distance
e between the first and
a second
d
surface
8. The create
ed second surface re
epresenting
g the second excavaation level is shown in
n
Figure 19.
592 of 607
Figure 18: The ‘Create
e array’ tooll.
Figure 19
9: Horizonta
al surface 1 and 2 for sstaged exca
avation
9. For the fin

nal surface which rep
present the avation leveel (z= -8.5m) the first
e final exca
or second surface can
c be cop
pied and th
he third su
urface is cconstructed
d using the
e
rray’ tool ag
‘Create arr gain.
10. In this exxample we
e will use the first horizontal
h surface aagain. Aga
ain set the
e
‘shape’ to 1D, in z direction’ an
nd the ‘Dis
stance betw
ween the ccolumns’ which
w is the
e
spacing will
w be (8.5
5-3.0 = 5.5
5m). Again
n a ‘negative’ value is require
ed to set a
downward
d direction. The comp
peted mode
el at this sttage is shoown in Figu
ure 20.
593 of 607
Figure 20: The excava
ation levels represented by 3 horiz
zontal surfaaces located
d at
(z1= -3m, z2
z = -6m annd z3 = 8.5mm)
eate struts
C4.Cre s and wale
ers for sup
pport systtem at level 1(z = -1 .5 m)
The suppo m for the firrst excavattion stage is made of
ort system o a line oof ‘waling’ transecting
t g
along the perimeter of the wall at z1 = -1 .5m and 5 numbers of struts loocated at a centre-to-
centre spa
acing of 10
0m along x--direction.
1. The
e walers will be mode
elled using
g ‘beam’ ele
ements.
2. . Tw
wo approacches can be
b used to
o construct the walers
s using thee ‘beam’ ellements. In
n
the first, the beams
b can
n be drawn
n using the ‘beam’ too
ol found inn the vertic
cal tool barr.
The
e coordinattes in x-y plane
p is ag
gain the same in Table 5 but thee elevation
n (z) will be
e
-1.5
5m below ground.
g Th
his approa ch is named ‘Cad in
nput’ approoach. We will
w use the
e
seccond appro
oach whe
ere the ‘b
beam’ elem
ments rep
presenting the wale
ers will be
e
con
nstructed using the ‘C
Command line’ appro
oach.
3. The
e ‘Comman
nd line’ box n in Figure 21 and the
x is shown e coordinaates for the
e beam line
e
is sshown. In
n this exa
ample bra
ackets are
e used to
o indicate the five blocks of
coo
ordinates. The
T bracke
ets are nott necessary
y but used here for cclarity reaso
ons.
Figu
ure 21: A be
eam line is drawn using
g 5 points. The last point is similarr to the firstt point
4. The
e constructed waling
g system iss shown in
n Figure 22.
2 The maaterial for the waling
g
can
n be inserrted using the same
e ‘drag-and
d-drop’ ap
pproach orr by desig
gnating the
e
matterial set th
hrough the
e ‘Model exxplorer’ box
x (see Figu
ure 22).
594 of 607
Figu
ure 22: Inpu
ut of materia
al set ‘Walin
ng’ to ‘Beam
ms’ represen
nting first leevel of walin
ng.
Figure
e 23: First le
evel waling system
s with
h material set placed in
n.
5. The
e 5 nos. of strut will be
b construccted. The first
f strut will
w be placeed using tw
wo points.
6. The
e centre-to
o-centre sp
pacing is 1
10m and elevation
e of
o the struut is z= -1.5m below
w
ground level. The first point is lo
ocated at (x1,
( y1, z1) = (-20, --5, -1.5) and
a second
d
poin
nt located at (x2, y2, z3) = (-20 , 5, -1.5).
7. The
ere are two
o possible options. In
n the first option
o the strut can bbe constru
ucted using
g
the Cad inputt approach using the coordinate
es given above. Thiss is done by choosing
g
595 of 607
the ‘beam’ ele
ement option from th
he vertical toolbar and drawing the two points using
g
the coordinate
e set.
8. The second option
o gain using the ‘Comm
is ag mand line’ approach..
9. Forr the secon
nd approac
ch type “be
eam -20 -5 -1.5 -20 5 -1.5”.
10. Pla
aced the material
m set for the strut eith
her using the
t ‘Drag--and-drop’ or ‘Mode
el
exp
plorer inputt’ approach
h.
11. The
e complete
ed model at
a this with one single
e strut with
h the walingg system is shown in
n
Figu
ure 24
Figu
ure 24: Loca
ation of firstt level ‘walin
ng’ with one
e strut. Firstt excavationn level and interface
i are
e
swittched off.
12. The
e remaining 4 nos. of
o struts a
are replicatted using the ‘Array tool’. The
e important
parrameters when
w using
g this too l are settiing the sh
hape = 1D
D, x directtion, no of
o
umns = 5 and
colu a the dis
stance bettween the columns = 10 (note ‘positive’). The strutss
will be replica
ated along x direction
n. The outtcome of this processs is shown
n in Figure
e
25.
596 of 607
Figu
ure 25: The
e complete support
s sys tem for leve
el 1 (z=-1.5m
m)
C5.Cre
eat struts and walerrs for supp
port syste
em at level 2 (z = -4. 5 m)
Figure 26: Using the selection

s box
x for beamss and selec
cting the stru
uts and waliling.
1. The
e second level supp
port system
m is creating by rep
plicating thhe first lev
vel supporrt
sysstems to a position off z= -4.5m.
2. Using the ‘sellect beams
s’ in the ve
ertical toolb
bar the enttire system
m consisting
g of waling
g
d struts are
and e selected.
3. Using the ‘Arrray tool’ ag
gain the wh
hole suppo
ort system is copied bby setting the
t ‘shape
e’
to 1
1D, z direction, numb
ber of colum
mns = 2 an
nd distance
e between columns is (-4.5m).
4. The
e Outcome
e of this pro
ocess is sh
hown in Fig
gure 27.
5. The
e first levell support system
s is ‘G
Grouped’ together
t by
y using thee ‘Select beams’
b too
ol
from
m the verttical toolba
ar. Using tthe RMB gesture
g an
nd ‘Group’ option all the beam
m
597 of 607
ments loccated here
elem ein are grrouped tog
gether. We
W can renname this
s group to
o
‘Gro
oup_1_Sup
pport_Leve
el_1’ unde
er the ‘Mod
del explorerr’ box (seee Figure 28
8)
6. The
e second level su
upport sysstem is also
a ‘Grou
uped’ andd then re
ename ass
‘Gro
oup_2_Sup
pport_Leve
el_2’ unde
er ‘Model ex
xplorer’ bo
ox (see Figgure 28)
Figure 27: Second levvel support system

s repllicate from the
t first usin
ng the ‘Arraay tool’
Figure 28: Using the Model

M explo
orer box to rrename the Group to Group_1_Su
G upport_Leve
el_1
598 of 607
C. MO
ODE: MESH
Figure 29: M
Mesh option
n using the default ‘Ele
ement Distri
ribution’ optiion of ‘Veryy Course’ mesh
Figure 30: Generated mesh, num

mber of elem
ments and nodes
n (Valu
ues will varyy according
g to the leve
el
ent)
of refineme
oceed to mesh discre

1. Pro etization byy moving to
o ‘Mode: Mesh’.
M In thhis example
e we use a
veryy coarse mesh.
m Click
k ‘OK’.
2. Oncce meshin
ng is comp
pleted a m
message is given in the
t ‘Sessioon Box’ as
s shown in
n
Fig.30. For example this curre
ent model has app
proximatelyy 12K of 10-noded
d
tetrrahedral ele
ements.
3. Pro
oceed to check
c the constructe
ed mesh by clicking on the ‘View Me
esh’ button
n
loca
ated on the ‘Vertical Toolbar’. A window
w will appea
ar as show
wn in Fig.3
31. and we
e
che
eck the qua
ality of the mesh :
Mesh
h> Mesh Quality
Q
599 of 607
Figure 31:: Mesh quallity check
Click the ‘U
Update’ bu
utton locate
ed at the to
op of the window
w and
d return to ‘Mode: Me
esh’.
4. We
e will now proceed
p to ‘Mode: Wa
ater Levels
s” to design the channge of the water-table
w e
with
hin the exccavated zone.
D. MO
ODE: WATER
W LEVELS
L S
1. In ‘Mode: Wa
ater Levels
s’ the conffiguration of
o the initial groundw
water cond
dition is ass
succh that the ‘General Phreatic
P Su
urface’ is set
s right at ground levvel as seen
n in Fig.32.
Thiss is the inittial groundwater leve
el that was set at ‘Borrehole Inpuut via Head
d’.
Figure 32: M
Mode: Wate
er Levels. Visual
V in deffault Fig
gure 33: Distribution of hydrostatic
c porewater
settings pressuure
600 of 607
2. Clicck ‘Preview
w’ button lo
ocated at th
he ‘Vertica
al Toolbar’.. The initiaal porewate
er pressure
e
disttribution fo
or this sta
age (Initia l phase) is shown in Fig 333. This is the input
porrewater pre
essure thatt we want tto adopt fo
or this phas
se.
ercise therre will be additional

In this exe a sstages. Th
he water le
evel within the excav
vation zone
e
will be low
wered at each
e succ
cessive sta
age. To prroceed furrther it is useful to switch the
e
visibility off the surrou
unding soill and gene
eral phreatiic level to 30%
3 (see F
Fig 34).
Figure 34: Visibility fo

or the Gene
eral Phreatiic Surface and
a surroun
nding soil bbeing switch
hed to 30%
%
visibility.
Phase 1: E
Excavation Stage 1(M
Mode Waterr Levels)
1. Add
d a new ph
hase and th
his phase rrepresent excavation
e n to level 1 (-3mBGL)).
2. In the currentt mode ourr concern i s to design
n the lowering of thee water leve
el by direcct
inpu s not switcched off at this stage
ut. The soil cluster is e (Mode: W
Water Leve
els) but the
e
soil zone representing
g excavatiion zone level 1 is selectedd using the ‘Selecct
Recctangular-S
Select Soils’ option ((Fig.35a). This zone
e is set to ‘Dry’ via the
t optionss
ava
ailable in ‘S
Selection Explorer’
E wiindow Fig 35b).
3. In this same phase
p the porewaterr pressure within zone2 (beneaath zone 1)) will be set
to ‘IInterpolate
e’ as shown
n in Fig 36 a and 36b.
In summa
ary the su
urrounding soils and
d soil bene
eath zone 2 is refe rred to the ‘Genera
al
Phreatic S
Surface’. Zone 1 is set
s to ‘Dry’ and Zone
e 2 is set to
t ‘Interpollate’ betwe
een zone 1
eral Phreattic Surface’.
and ‘Gene
601 of 607
Phase 2: S
Support System
S 1 (Mode
( Wa ter Levels
s)
1. Thiss phase iss added and labelle

ed as ‘Con
nstruct Sup
pport Systeem 1’. This stage iss
ded to the sequence as we willl use this Phase to activate thhe ‘waling and strutss’
add
thatt representt support system
s 1.
2. The
e water con
ndition at this stage i s similar to
o Phase 1.
Phase 3: E
Excavatio
on Stage 2 (Mode W
Water Levels)
1. Thiss phase re
epresent ex
xcavation o
of zone 2 to
t 6m below
w ground llevel.
2. The
e ground water
w is low
wered by selecting zone
z 2 and setting tthe condition to ‘Dryy.
The
e procedurre is similarr to the me
ethod state
ed in Phase
e 1.
3. The
e porewate
er pressure
e within zon
ne3 that is beneath zone
z 2 is s et to ‘Interpolate’
Phase 4: S
Support System
S 2 (Mode
( Wa ter Levels
s)
1. Add
d phase 4 and name
ed this sta
age as ‘Support System 2’ Thiss phase will
w be used
d
late
er.
2. The
e porewate
er pressure
e is kept the same as
s in Phase 3.
Phase 5: E
Excavatio
on to Finall Level (Mo
odel Wate
er Levels)
el which is -8.5m beloow ground level. Thiss

1. Thiss is the final excavated state off the mode
zon
ne 3.
2. Sim
milarly the water
w cond
dition for zo
one 3 is ‘Dry’
3. The
e zone 4 be
eneath zon
ne 3 is set to ‘Interpo
olate’
Figure 35a:: Select exccavation zon

ne 1 using ‘‘Window Re
ectangular-S
Select Soil’’ Option
602 of 607
Figure 35b:: Changing the water condition
c to ‘Dry’ for ex
xcvation zon
ne 1
Figure 36a:: Select ‘Zo

one 2’ that is
s beneath Z
Zone 1
Figure 36
6b: Select ‘In
nterpolate’ option
o for zzone 2 and a preview of the input h ydrostatic porewater
cond
dition for Phase 1
603 of 607
E. MODE: STAGED CONSTRUCTION
Initial Phase
By default ‘Initial Phase’ is already included. All the structural elements are switched off.
1. The retaining structure is constructed by ‘wish-in-place’ (WIP) method. No

installation effects are included.
2. The retaining structure is represented by ‘Plate’ elements. Activate the ‘Plates’ by
using the ‘RMB’ gesture or through the ‘Selection explorer’ box.
Phase 1: Excavation Stage 1 (Mode: Stage Construction)
In this stage:
1. Zone 1 is ‘deactivated’ to simulate excavation to level -3mbgl

2. Ensure that the ‘Plate’ elements are activated for the retaining wall
Phase 2: Support System 1 (Mode: Stage Construction)
In this stage:
1. Support system level 1 is activated.

2. Ensure that Zone 1is deactivated and the ‘Plates are activated.
Phase 3: Excavation Stage 2 (Mode: Stage Construction)
In this stage:
1. Zone 1 and Zone 2 is deactivated to simulate excavation to level -6mbgl

2. Ensure that Support System Level 1 and the ‘Plates’ are still activated.
Phase 4: Support System 2(Mode: Stage Construction)
1. In this stage ‘Support System Level 2’ is activated.

2. Ensure that structural members from previous phases are activated.
Phase 5: Excavation to Final Stage (Mode: Stage Construction)
1. The final construction stage is modelled by deactivating zone 4.

2. The final excavation level is -8.5mbgl
604 of 607
Proceed to
o ‘Calculatte’. Prior to
o calculatio
on we nee
ed to selec
ct two poinnts. These two pointss
are used to plot the ‘Displacem
‘ ment vs. tim
me’ curves.
1. Clicck ‘Select Points for Curves’ in

n the vertic
cal toolbar. A window
w (Select Nodes
N and
d
Stre
ess Points) will appear. Two po
oints are selected using the ‘x-yy-z points--of-interestt’
optiion (see Fiig 37).
2. Loccate Point ‘A’ using the coord
dinates giv
ven in Tab
ble 6. Clicck the nea
arest node
e
loca
ated on at this coordiinate and sselect the correspond
c ding node..
3. Rep
peat for Po
oint ‘B’.
4. Clicck ‘Calcula
ate’ located
d in the ‘Ve
ertical Toolbar’.
5. Mon
nitor the ru
un and wait for the co
ompletion of
o the calculations.
Table 6: Co
oordinates
s for ‘A’ and
d ‘B’
Point x y z
A 0 5 0
B -24.5
5 5 0
Figure 37: S
Selection of point ‘A’ and ‘B’ for pllotting
605 of 607
F. OU
UTPUT OF
O RES
SULTS
Figure 38: P
Plot: Displa
acement usiing iso-surfa
aces
Figure 39: A
Axial forcess for struts at
a final exca
avation leve
el.
Figure 40: D
Deformation
n of retainin
ng structure
e
606 of 607
Figure 41: Wall displaccements at point A and
d B with tim
me.
607 of 607

Plaxis Advanced Computational Geotechnics-Singapore 2011 (Compiled) PDF

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Plaxis Advanced Computational Geotechnics-Singapore 2011 (Compiled) PDF

Transféré par

Droits d'auteur :

Formats disponibles

Plaxis AsiaPac Pte Ltd

Time Module On Hard and Soft Soils

10:45 11:00 Tea Break

04:30 05:30 E2 Simulation of Laboratory Tests 2 Shen

Time Module Application of Plaxis 2D

Time Module Application of Plaxis 3D

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Singapore 2011 5 of 607

Stresses and strains

 Hydrostatic (constant head)

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Singapore 2011 6 of 607

Stress invariants (p and q):

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Stresses and strains

Normal strains Shear strains

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Singapore 2011 7 of 607

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Stresses and strains

Singapore 2011 8 of 607

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Stress paths in standard soil tests

Loose sand Dense sand

Singapore 2011 9 of 607

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Stress paths in standard soil tests

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Singapore 2011 10 of 607

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Stress paths in standard soil tests

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Singapore 2011 11 of 607

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Stress paths in standard soil tests

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Singapore 2011 12 of 607

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Stress paths in standard soil tests

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Singapore 2011 13 of 607

Strains (or increments) are divided into elastic strains

A soil model relates increments of stress to

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Basic concepts of the M-C model

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Singapore 2011 14 of 607

Classical plasticity: g = f (associated plasticity)

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Basic concepts of the M-C model

Determination on the basis of a yield function f =

• If f < 0 Pure elastic behaviour

Singapore 2011 15 of 607

Can be represented as a contour in (principal) stress

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Basic concepts of the M-C model

The condition   c’ - ’ tan’ must hold for arbitrary angles 

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Singapore 2011 16 of 607

Basic concepts of the M-C model

• Mohr-Coulomb plastic potential:

PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES

Singapore 2011 17 of 607