You are on page 1of 449

Advanced Computational Geotechnics

ADV
VANCED CO
OMPUT
TATIO
ONAL GEOT
G TECHN
NICS
2012
HON
NG KOONG

Page 1
Advanced Computational Geotechnics

ADVA
A ANCED COU
C RSE
E
ON
O COMPUTTATION
NAL G
GEOTE
ECHNICS
HONG
H KONG
G

Venue HKMA,Sh nChai,Hong Kong
hop2,G/F,PicoTower,666GloucesterRoad,Wan
Dates 68Noveember2012:AdvanceddCourse6
9Novem
mber2012:3DApplicattionCourse
Leccturers

Pro
ofessorHelm
mutSchweiger(Course
eleader) GrazUniveersityofTech
hnology,Grazz,Austria

Pro
ofessorAntonioGens TechnicalU
UniversityofCatalonia(U ona,Spain
UPC),Barcelo

DrLeeSiewW
Wei GolderAsssociates(HK))Ltd.

DrJohnnyCheeuk AecomAsia g
a,HongKong

DrWilliamCheang PlaxisAsia
aPac,Singap
pore

Organisedby
So
olutions Reesearch Centre Ltd
1709-11, Leig
ghton Centre
e
77 Leighton R
Road
Ca
auseway Ba ay, Hong Ko
ong

Pla
axis AsiaPaac Pte Ltd
16 Jalan Kilan
ng Timor
05-07 Redhill Forum
Sin
ngapore

Page 2
CONTENTS

CG MODULES PAGE
1 GeotechnicalFiniteElementAnalysis 7
2 ElastoplasticityandMohrCoulomb 23
3 Exercise1:ElastoplasticAnalysisofShallowFoundations 43
4 CriticalStatesSoilMechanicsandSoftSoilModel 81
5 HardeningSoilandHSsmallModel 101
6 Exercise2:SimulationofTriaxialandOedometerTests 131
7 ModellingofDeepExcavations 179
8 StructuralElementsinPlaxis 203
9 Exercise3:ExcavationofaBuildingPit 217
10 DrainedandUndrainedAnalysis 253
11 ConsolidationAnalysis 273
12 ModellingofGroundwaterinPlaxis 289
13 Exercise4:ExcavationandDewatering 317
14 UnsaturatedSoilMechanicsandBarcelonaBasicModel 329
15 InitialStresses,PhiCReductionandSlopeStabilityAnalysis 361
16 Exercise5:StabilityofaSlopeStabilisedbySoilNails 389
17 HoekBrownandRockJointedModel 403
18 Tunnellingin2D 427
19 Exercise6:ExcavationofaTunnelinRock 439

Special3DModellingModules
20 Plaxis3D 450
21 3DModellingofTunnels 468
22 Exercise7:StabilityofaNATMTunnel 496
23 3DModellingofDeepFoundations 518
24 3DModellingofDeepExcavations 552
25 Exercise8:ModellingofExcavation 578

Advanced Computational Geotechnics

COURSE STRU
UCTU
URE AND
A LECT
TURE
ERS
The course is divided intto nineteen lectures (CG ) which incluudes six 2-D modelling exxercises for the t 3-day
advanced co ourse. For th
he add-on sp pecial coursee on 3-D moddelling, there is an additioon of six lectu
ures and it
includes two
o 3-D modelling exercises. The modu ules and exercises are grrouped into 4 themes to form f the
Advanced C Computationa al Geotechnics course in Hong Kong using Plaxis s finite elemeent programs s
Theme1 Advancced Computa echnics and Soil Behavio
ational Geote our
Theme2 Applica ations I: Exca avations andd Modelling of
o Groundwater
Theme3 Applica ations II: Initial Stresses, Unsaturated
d Soils and Modelling
M of P
Problems in Rock
Theme4 Applica ations III: 3D Analysis of Deep Found dations, Excaavations & Tuunnels

HHelmut obtainned his Ph.D frrom the Univeersity College of o Swansea,


UU.Kingdom an nd teaches courses on Advaanced Soil Me echanics and
Professor Helmut CComputationa al Geomechan nics at the Graaz University of
o Technology,,
Schweigger AAustria. He haas over 15 yea ars of experiennce in developpment and
aapplication of the finite elem
ment method. A As a member of several
ccommittees Helmut is involv ved in formulaating guidelinees and
rrecommendattions for the us se of finite eleements in prac
ctical
ggeotechnical engineering.
e
C
COURSE LE
EADER

AAntonio is a fa
aculty member at the Technnical University y of Catalonia
ssince 1983 aftter a Ph.D. at Imperial Colleege, London. He
H has been
Pro
ofessor Antonio Gens iinvolved in ge
eotechnical ressearch, educaation and practtice for more
tthan 25 years with special reference
r to thhe application of numerical
Teechnical Univ
versity of aanalysis to engineering probblems. He hass consulted in a variety of
Catalonio (U
UPC) pprojects involv
ving deep excavations, tunnnels, ground im mprovement
ttechniques, daams, power sttations, foundaations and sloopes. He
ddelivered the the
t 47th Rank kine lecture in 2007

WWilliam obtainned his Ph.D frrom the Nationnal University of Singapore.


Dr William Cheang
D C HHis interest is in Computatioonal Geotechnnics. He has worked
w as a
Plaxis Asia
a Pac GGeotechnical Engineer in Malaysia,
M Singaapore and Tha ailand. He is
iinvolved with many seminarrs and workshhops around Asia A for the
ppromotion of good
g ctive usage off Plaxis Finite Element
and effec
CCodes.
COORDINA
ATOR

SSW obtained his PhD in Ge eotechnical Enngineering fromm Cambridge


Dr Lee Siew
w Wei UUniversity in 2001
2 and has been working for Geotechn nical
CConsulting Grroup (Asia) Ltd d (now Golderr Associates (HHK) Ltd)
Technical Director ssince. He is a qualified civil//geotechnical engineer, and
d has applied
Go
older Associa
ates (HK) nnumerical mod delling to the design,
d assesssment and invvestigation of
ggeotechnical problems
p in Hong Kong andd other Asian countries
c

JJohnny obtainned his PhD degree from Cambridge University.


U Hee
hhad lectured at
a City Univerrsity of Hong Kong and thee University off
HHong Kong before
b joining AECOM in 20009. Johnny has extensive e
eexperience in research and practicee in offshore e and slope e
eengineering. He is a member of the Intternational Soociety for Soil
Dr Johnny Cheuk
C MMechanics and
a Geotechnical Engineeering (ISSMGE) TC103 3
Senior Engiineer
((Numerical Methods
M in Geomechanic
G cs) and TC1 104 (Physical
AECOM Asia Co.
A C Ltd.
MModelling in Geotechnics)). Johnny is a member and a Assistantt
SSecretary of Geotechnical Division Com e Hong Kong.
mmittee of the
JJohnny has been an Honorrary Assistant Professor at the Universityy
oof Hong Kong since 2009.

Page 4
Advanced Computational Geotechnics

Time Day1AdvancedComputationalGeotechnics
9:00 10:00 CG1 GeotechnicalFiniteElementAnalysis Gens
10:00 11:00 CG2 ElastoPlasticityandMohrCoulomb Gens
11:00 11:15 Break
11:15 1:00 CG3 Exercise1:FoundationonElastoPlasticSoils Cheang
1:00 2:00 Lunch
2:00 3:00 CG4 CriticalStateSoilMechanicsandSoftSoilModel Schweiger
3:00 4:00 CG5 HardeningSoilandHSsmallModel Schweiger
4:00 4:15 Break
4:15 5:30 CG6 Exercise2:TriaxialandOedometer Cheang

Time Day2:CGApplications1:Excavations
9:00 9:45 CG7 ModellingofDeepExcavations Schweiger
9:45 10:30 CG8 StructuralElementsinPlaxis Cheuk
10:30 10:45 Break
10:45 12:00 CG9 Exercise3:TiedBackExcavation Lee
12:00 1:00 Lunch
1:00 1:45 CG10 DrainedandUndrainedAnalysis Gens
1:45 2:30 CG11 ConsolidationAnalysis Gens
2:30 3:15 Break
3:15 3:30 CG12 ModellingofGroundwaterinPlaxis Lee
3:30 5:00 CG13 Exercise4:DewateringinExcavation Cheuk

Time Day3:CGApplications2:UnsaturatedSoilsandRock
9:00 10:30 CG14 UnsaturatedSoilsandBarcelonaBasicModel Gens
10:30 11:30 CG15 InitialStressesandSlopeStabilityAnalysis Schweiger
11:30 11:45 Break
11:45 1:30 CG16 Exercise5:SlopeStabilityExercise Cheuk
1:30 2:30 Lunch
2:30 3:30 CG17 HoekBrownandRockJointedModels Schweiger
3:30 4:00 CG18 ModellingofTunnelsin2D Schweiger
4:00 4:15 Break
4:15 5:30 CG19 Exercise6:TunnellinginRock Cheang

Time Day4:CGApplications3:3DAnalysis(Optional)
9:00 10:00 CG20 IntroductiontoPlaxis3D Cheang
10:00 11:00 CG21 3DModellingofTunnels Schweiger
11:00 11:15 Break
11:15 1:00 CG22 Exercise7:TunnelStability Cheang
1:00 2:00 Lunch
2:00 3:00 CG23 3DModellingofDeepFoundations Schweiger
3:00 4:00 CG24 3DModellingofDeepExcavations Schweiger
4:00 4:15 Break
4:15 5:30 CG25 Exercise8:ModellingofExcavations Cheang

Page 5
Advanced Computational Geotechnics

H
HKCG 1

GEOTEC
CHNICA
AL FIN
NITE ELEME
E ENT A NALYSIS

Professo
P or Anton
nio Gens
s

Page 6
Advanced Computational Geotechnics

CG1: GEOTECHNICAL FINITE ELEMENT ANALYSIS

Antonio Gens

Technical University of Catalunya, Barcelona

some of the slides were originally created by:


Andrew Abbo (University of Newcastle)
Cino Viggiani (Laboratoire 3S, Grenoble, France)
Dennis Waterman (Plaxis)

Outline

Introduction
Finite Elements displacement analysis
Elements for two-dimensional analysis
Displacement interpolation
Strains
Constitutive equation
Element stiffness matrix
Global stiffness matrix
Solution of the global stiffness equations
Elasticity as applied to soils
Fundamentals, and elastic parameters
Two-dimensional elastic analysis

Page 7
Advanced Computational Geotechnics

design requirements in geotechnical engineering

Stability (local and general)

Admissible deformation and displacements

design requirements in geotechnical engineering

Flow problems

Sometimes flow and stability/deformation problems are solved together


See tomorrows lecture on consolidation (CG11)

Page 8
Advanced Computational Geotechnics

geotechnical analysis: basic solution requirements

Equilibrium (3 equations)
Unknowns: 15
Compatibility (6 equations)
(6 stresses, 6 strains, 3 displacements)
Constitutive equation (6 equations)

Potts & Zdravkovic


(1999)

geotechnical numerical analysis

methods for numerical analysis


Finite difference method
Boundary element method (BEM)
Discrete element method (DEM)
Finite element method (FEM)
Others (meshless methods, particle methods)

While the FEM has been used in many fields of engineering practice for over 40 years, it
is only recently that it has begun to be widely used for analyzing geotechnical problems.
This is probably because there are many complex issues which are specific to
geotechnical engineering and which have been resolved relatively recently.

when properly used, this method can produce realistic results which are of value to
practical soil engineering problems

A good analysis, which simulates real behaviour, allows the engineer to understand
problems better. While an important part of the design process, analysis only provides
the engineer with a tool to quantify effects once material properties and loading
conditions have been set

Page 9
Advanced Computational Geotechnics

geotechnical finite element analysis

Objectives of the numerical (finite element) analysis


Selection of design alternatives
Quantitative predictions
Backcalculations
Understanding!
Identification of critical mechanisms
Identification of key parameters

geotechnical finite element analysis

Advantages of numerical (finite element) analysis


Simulation of complete construction history
Interaction with water can be considered rigorously
Complex geometries (2D-3D) can be modeled
Structural elements can be introduced
No failure mechanism needs to be postulated (it is an outcome of the analysis)

(Nearly) unavoidable uncertainties


Ground profile
Initial conditions (initial stresses, pore water pressure)
Boundary conditions (mechanical, hydraulic)
Appropriate model for soil behaviour
Model parameters

Page 10
Advanced Computational Geotechnics

geotechnical finite element analysis

Some requirements for successful numerical modelling


Construction of an adequate conceptual model that includes the basic features of the
model. The model should be as simple as possible but not simpler
Selection of an appropriate constitutive model. It depends on:
type of soil or rock
goal of the analysis
quality and quantity of available information
Pay attention to patterns of behaviour and mechanisms rather than just to
quantitative predictions
Perform sensitivity analyses. Check robustness of solution
Model calibration (using field results) should be a priority, especially of quantitative
predictions are sought
Check against alternative computations if available (even if simplified)

three final remarks

1. geotechnical engineering is complex. It is not because youre


using the FEM that it becomes simpler

2. the quality of a tool is important, yet the quality of a result


also (mainly) depends on the users understanding of both
the problem and the tool

3. the design process involves considerably more than analysis

Borrowed from C. Viggiani, with thanks

Page 11
Advanced Computational Geotechnics

introduction: the Finite Element Method

the FEM is a computational procedure that may be used to obtain an approximate solution
to a boundary value problem

the governing mathematical equations are approximated by a series of algebraic equations


involving quantities that are evaluated at discrete points within the region of interest. The
FE equations are formulated and solved in such a way as to minimize the error in the
approximate solution

this lecture presents only a basic outline of the method


attention is focused on the "displacement based" FE approach

introduction: the Finite Element Method

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


solution to a boundary value problem

What kind of problem?

Apply load obtain displacements


stiffness matrix

Apply head obtain flow


permeability matrix

Though we would like to know our solution at any coordinates in our project, we will only
calculate them in a certain amount of discrete points (nodes) and estimate our solution
anywhere else

Page 12
Advanced Computational Geotechnics

introduction: the Finite Element Method

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

The first stage in any FE analysis Footing


width = B
is to generate a FE mesh

A mesh consists of elements


connected together at nodes
Node
We will calculate our solution in the
nodes, and use some sort of
mathematical equation to estimate the
solution inside the elements. Gauss point

examples: embankment

Page 13
Advanced Computational Geotechnics

examples: multi-anchored diaphragm wall

introduction: the Finite Elements Method

the nodes are the points where values Footing


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

Node

the values of nodal displacements are


interpolated within the elements to give
algebraic expressions for displacement Gauss point

and strain throughout the complete mesh

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


leads to the calculation of forces acting at the element nodes

the nodal forces are related to the nodal displacements by equations


which are set up and solved to find values of the nodal displacements

Page 14
Advanced Computational Geotechnics

introduction: the Finite Elements Method

Apply load obtain displacements


stiffness matrix

Ku F
For soil we dont have a direct relation between load and displacement,
we have a relation between stress and strain.

Displacements Strains Stresses Loads

Differentiate Material Integrate


model
Bu D F dV

Combine these steps: K B T DB dV

introduction: the Finite Elements Method

The FEM involves the following steps (1/2)

Elements discretization
This is the process of modeling the geometry of the problem under
investigation by an assemblage of small regions, termed finite elements. These
elements have nodes defined on the element boundaries, or within the elements

Primary variable approximation


A primary variable must be selected (e.g., displacements) and rules as how it
should vary over a finite element established. This variation is expressed in terms of
nodal values

A polynomial form is assumed, where the order of the polynomial


depends on the number of nodes in the element
The higher the number of nodes (the order of the polynomial), the
more accurate are the results (the longer takes the computation!)

Page 15
Advanced Computational Geotechnics

introduction: the Finite Elements Method

The FEM involves the following steps (2/2)


Element equations
Derive element equations:

where is the element stiffness matrix, is the vector of nodal


displacements and is the vector of nodal forces

Global equations
Combine element equations to form global equations

Boundary conditions
Formulate boundary conditions and modify global equations. Loads affect P,
while displacements affect U

Solve the global equations


to obtain the displacements at the nodes

Compute additional (secondary) variables


From nodal displacements secondary quantities (stresses, strain) are evaluated

displacement interpolation

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


either triangular or quadrilateral elements

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

Page 16
Advanced Computational Geotechnics

introduction: the Finite Elements Method

Displacement interpolation
primary unknowns: values of the nodal displacements
displacement within the element: expressed in terms of the nodal values using
polynomial interpolation
n
u ( ) N i ( ) ui , N i shape function of node i
i 1

Shape function of node i


Is a function that has value 1 in node i
and value 0 in all other n-1 nodes of the element

Shape functions for 3-node line element


1 1
N 1 (1 ) , N 2 (1 )(1 ) , N3 (1 )
2 2

introduction: the Finite Elements Method

Illustration for the six-noded triangular element

3 y quadratic interpolation

6
v
5
x u ( x , y ) a0 a1 x a2 y a3 x 2 a4 xy a5 y 2
u v ( x , y ) b0 b1 x b2 y b3 x 2 b4 xy b5 y 2
1 2
4

12 coefficients, depending on the values of the


12 nodal displacements

Page 17
Advanced Computational Geotechnics

introduction: the Finite Elements Method

Illustration for the six-noded triangular element

Strains may be derived within the element using the standard definitions

u
xx a1 2 a3 x a4 y
x
Lu
v
yy b2 b4 x 2b5 y
y
u v
xy (b1 a2 ) ( a4 2b3 ) x (2a5 b4 ) y
y x

Lu LNU e BU e BU e

introduction: the Finite Elements Method

Constitutive relation (elasticity)


Elasticity: one-to-one relationship between stress and strain
in a FE context, stresses and strains are written in vector form
the stress-strain relationship is then expressed as: = D

linear isotropic elasticity in plane strain


material stiffness matrix


1 v v 0
D
E v 1 v 0
(1 2v )(1 v ) 1 2v
0 0
2

in this case the coefficients of the matrix are constants, which means
that (for linear kinematics) the resulting F.E. equations are linear

Page 18
Advanced Computational Geotechnics

introduction: the Finite Elements Method

What happens with inelastic constitutive relations?

Advantage with elasticity: the coefficients of the matrix are constants,


the resulting F.E. equations are linear, hence the problem may be solved
by applying all of the external loads in a single calculation step

soils usually do not behave elastically

D
with D depending on the current and past stress history

It is necessary to apply the external load in separate increments and to adopt


a suitable non-linear solution scheme

Element stiffness matrix

body forces and surface tractions applied to the element may be


generalized into a set of forces acting at the nodes (vector of nodal forces)

nodal forces may be related P1x


to the nodal displacements by:
3 P1 y

P2 x
K eUe Pe 6 5 P2 y
Ke element stiffness matrix P e
P1x 1
2


4

Ke B T DBdv P1y
P6 x

P6 y
D material stiffness matrix
recall B matrix relating nodal displacements to strains

Page 19
Advanced Computational Geotechnics

introduction: the Finite Elements Method

Gauss points

Ke
B T DBdv
To evaluate Ke, integration must be performed for each element
A numerical integration scheme must be employed (Gaussian integration)

Essentially, the integral of a function is replaced by a weighted sum


of the function evaluated at a number of integration points

introduction: the Finite Elements Method

Global stiffness matrix (1)

The stiffness matrix for the complete mesh is evaluated by combining the individual
element stiffness matrixes (assembly)

This produces a square matrix K of dimension equal to the number of degrees-of-


freedom in the mesh

The global vector of nodal forces P is obtained in a similar way by assembling the
element nodal force vectors

The assembled stiffness matrix and force vector are related by:

KU P
where vector U contains the displacements at all the nodes in the mesh

Page 20
Advanced Computational Geotechnics

introduction: the Finite Elements Method

Global stiffness matrix (2)

if D is symmetric (elasticity), then Ke and hence K will be symmetric


The global stiffness matrix generally contains many terms that are zero
if the node numbering scheme is efficient then all of the non-zero
terms are clustered in a band along the leading diagonal

assembly
schemes for storage
solution

take into account its sym and


banded structure

number of dofs

introduction: the Finite Elements Method

Solution of the global stiffness equations

Once the global stiffness equations have been established


(and the boundary conditions added), they mathematically
form a large system of symultaneous (algebraic) equations

KU P
These have to be solved to give values for the nodal displacements

It is advantageous to adopt special techniques to reduce


computation time (e.g. bandwidth and frontal techniques)

Detailed discussion of such techniques is beyond the scope of


this lecture

Page 21
Advanced Computational Geotechnics

introduction: the Finite Elements Method

Compute additional (secondary) values


once the nodal displacements have been obtained from the inversion
of the matrix K
KU Pe

The complete displacement field can be obtained:


n
u ( x, y ) N i ( x, y ) ui , N i shape function of node i
i 1

Strains and stresses are computed at the Gauss points:

BU e
= D

introduction: the Finite Elements Method

some practical issues

1. A good finite element mesh is important. A poor mesh will give


a poor (inaccurate) solution.

2. Post processing Stress are computed at Guass points only.


Contour plots of stresses involve further processing of the results.

3. Do the results make sense?

4. FEA can be very time consuming!

Page 22
Advanced Computational Geotechnics

H
HKCG 2

EL
LASTO
O-PLAS TY AND MOH
STICIT HR-CO
OULOM
MB
FA
AILUR
RE CRITERIO
ON
Professo
P or Anton
nio Gens
s

Page 23
Advanced Computational Geotechnics

CG2: ELASTO-PLASTICTY AND MOHR COULOMB

Antonio Gens

Technical University of Catalunya, Barcelona

some of the slides were originally created by:


Cino Viggiani (Laboratoire 3S, Grenoble, France)
S.W. Lee (GCG Asia Golder Associates)
Helmut Schweiger (Technical University of Graz, Austria)

Contents

A quick reminder of (linear isotropic) Elasticity


Motivations for plasticity (elasticity vs. plasticity)
Basic ingredients of any elastoplastic model
elastic properties (how much recoverable deformation?)
yield surface (is plastic deformation occurring?)
plastic potential (direction of plastic strain increment?)
consistency condition (magnitude of plastic strain increment?)
hardening rule (changes of yield surface?)
Element tests: (drained) simple shear & triaxial tests
Tips and tricks
Advantages and limitations

Page 24
Advanced Computational Geotechnics

Constitutive models

Constitutive models provide us with a relationship with stresses and


strains expressed as:
= D

Elasticity
Linear-elastic Non-linear elastic

= D

Hookes law

C
xx 1 0 0 0 xx
1 0 0 0
yy yy
zz 1 1 0 0 0 zz
E 0 0 0 2 2 0 0
xy xy
yz 0 0 0 0 2 2 0 yz

zx 0 0 0 0 0 2 2 zx

Page 25
Advanced Computational Geotechnics

Model parameters in Hookes law:

d1
Two parameters:
- d1
- Youngs modulus E
- Poissons ratio
d3
- 1

Meaning (axial compr.):

d1 E
E
d1 1
- 1
d
3
d1 1

E 0 ; -1 0 .5 3

Alternative parameters in Hookes law:


In spherical and deviatoric stress / strain components:
v 1/ K 0 p
p 1
3 1 2 3
0 1/ 3G q 1
s q ( 1 2 ) 2 ( 2 3 ) 2 ( 3 1 ) 2
2
dxy
Shear modulus:
d xy E dxy
G
d xy 21

Bulk modulus: dp
dp E
K dv
d v 31 2

9KG 3 K 2G
E v
G 3K 6 K 2G

Page 26
Advanced Computational Geotechnics

Hookes law

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

zx 0 0 0 0 0 2 zx

4G 2G 2G
K K K 0 0 0
xx
3 3 3
xx
K
0
2G 4G 2G
yy K K 0 0 yy
3 3 3
zz zz
K
2G
K
2G
K
4G
0

xy 3 3 3
0 0
xy
yz 0 0 0 G 0 0 yz

zx 0 0 0 0 G 0 zx
G
0 0 0 0 0

Hookes law

Plane strain = D
4G 2G
K K 0
1 0 3 3


0
E 2G 4G
D 1 0 K K
(1 )(1 2 ) 3 3
1 2
0 0 0 0 G
2

Axisymmetry
4G 2G 2G
K K K 0
1 0 3 3 3

1 0
0
2G 4G 2G
E K K K
D 1 0 3 3 3
(1 )(1 2 )
0 1 2 K
2G
K
2G
K
4G
0
0 0
2 3 3 3
G
0 0 0

Page 27
Advanced Computational Geotechnics

Elasticity vs. Plasticity (1)

In elasticity, there is a one-to-one relationship between stress and


strain. Such a relationship may be linear or non-linear. An essential
feature is that the application and removal of a stress leaves the
material in its original condition

Elasticity vs. Plasticity (2)

for elastic materials, the mechanism of deformation depends on the


stress increment

for plastic materials which are yielding, the mechanism of (plastic)


deformation depends on the stress

reversible = elastic irreversible = plastic

Page 28
Advanced Computational Geotechnics

Plasticity: some definitions (1)

LINEAR ELASTIC - PERFECTLY PLASTIC


One-dimensional
Y0 = yield stress

IMPORTANT: yield stress = failure stress for perfect plasticity

e p
General three-dimensional stress state e p

Plasticity: some definitions (2)

LINEAR ELASTIC PLASTIC HARDENING


One-dimensional

Y0 = yield stress
YF = failure stress

IMPORTANT: yield stress failure stress

e p
General three-dimensional stress state e p

Page 29
Advanced Computational Geotechnics

Plasticity: some definitions (3)

LINEAR ELASTIC - PLASTIC WITH SOFTENING

One-dimensional

Y0 = yield stress
YF = failure stress

yield function (1)

when building up an elastic-plastic model,


the first ingredient that we need is a yield surface
(is plastic deformation occurring?)

Page 30
Advanced Computational Geotechnics

yield function (2)

F = 0 represents surface in stress space

f f 1 , 2 , 3

f 0 stress state is elastic

f 0 stress state is plastic

f 0 stress state not admissible

The yield surface bounds all elastically attainable states


(a generalized preconsolidation pressure)

yield function (5)

Basically:
changes of stress which remain inside the yield surface are
associated with stiff response and recoverable deformations,
whereas on the yield surface a less stiff response is obtained and
irrecoverable deformations are developed

Where do we get this function f ?

The dominant effect leading to irrecoverable changes in particle


arrangement is the stress ratio, or mobilized friction

The mean normal effective stress p is of primary importance.


The range of values of q for stiff elastic response is markedly
dependent on p
Tresca & Von Mises yield functions are not appropriate

Page 31
Advanced Computational Geotechnics

Mohr-Coulomb Model, yield function

To most engineers the phrase strength of soils conjures


up images of Mohr-Coulomb failure criteria

frictional
resistance

independent of
normal stress

Classical notions of Mohr-Coulomb failure can be


reconciled with the patterns of response that we are
modeling here as elasto-plastic behavior

Mohr-Coulomb Model, yield function

1 and 3 : major and minor principal stresses

Page 32
Advanced Computational Geotechnics

The Mohr-Coulomb failure criterion

MC criterion: t*c cos - s* sin

t* = (3 - 1)
s* = (3+1)

1
2 '3 '1 c' cos ' 12 '3 '1 sin '

2c' cos ' 1 sin '


'1 '3
1 sin ' 1 sin '

Note: Compression is negative, and 1: major,


2: intermediate, 3: minor principal stress
19

Mohr-Coulomb Model, yield function

MOHR COULOMB IN 3D STRESS SPACE

f 1
2
'1 '3 12 '1 '3 sin ' c ' cos '
-1

f > 0 Not acceptable


f = 0 Plasticity

f < 0 Elasticity

-3
-2

Page 33
Advanced Computational Geotechnics

plastic potential (1)

Summing up:

Plastic strain increment arises if:


1) the stress state is located on the yield surface (f = 0)
AND
2) the stress state remains on the yield surface after a stress increment

knowledge of function f tells us whether plastic strain is occurring or not

But, this is only one part of the story:


We would also like to know direction and magnitude of plastic strain
will we get plastic volume changes?
and plastic distortion?
for that, we need another concept (another function: g)

plastic potential (2)

flow rule
Recall: plastic deformations depend
on the stress state at which yielding
is occurring, rather than on the route
by which that stress is reached

we have now two functions, f and g


the question is: where do we get g ?

Page 34
Advanced Computational Geotechnics

associated and non associated flow rules

it would be clearly a great advantage if, for a given material, yield locus
and plastic potential could be assumed to be the same
f = g only 1 function has to be generated to describe plastic response
also advantageous for FE computations:
the solution of the equations that emerge in the analyses is faster
the validity of the numerical predictions can be more easily guaranteed

is f = g a reasonable assumption?

for metals, it turns out that YES, it is


for geomaterials, NOT
Where is the problem? The assumption of normality of plastic strain
vectors to the yield locus would result in much greater plastic volumetric
dilation than actually observed

Mohr-Coulomb Model, plastic potential

dilatancy angle

Page 35
Advanced Computational Geotechnics

plastic dilatancy

how to understand dilatancy


i.e., why do we get volume changes when applying shear stresses?

= + i
the apparent externally mobilized angle of friction on horizontal planes () is
larger than the angle of friction resisting sliding on the inclined planes (i)

strength = friction + dilatancy

consistency condition

Page 36
Advanced Computational Geotechnics

Parameters of MC model

E Youngs modulus [kN/m2]


Poissons ratio [-]
c (effective) cohesion [kN/m2]
(effective) friction angle []
Dilatancy angle []

MC model for element tests

yy
tan
xy

Page 37
Advanced Computational Geotechnics

MC model for element tests

MC model for element tests

Page 38
Advanced Computational Geotechnics

limitations of MC model (1)

limitations of MC model (2)

Page 39
Advanced Computational Geotechnics

warning for dense sands

Possibilities and limitations of the Linear Elastic- Perfectly


Plastic (LEPP) Mohr-Coulomb model

Possibilities and advantages

Simple and clear model


First order approach of soil behaviour in
general 1
Suitable for a good number of practical
applications (not for deep excavations and
tunnels)
Limited number and clear parameters
Good representation of failure behaviour 3
2
(drained)
Dilatancy can be included
34

Page 40
Advanced Computational Geotechnics

Possibilities and limitations of the Linear Elastic- Perfectly Plastic (LEPP) Mohr-
Coulomb model

Limitations and disadvantages

Isotropic and homogeneous behaviour


Until failure linear elastic behaviour
No stress/stress-path/strain-dependent 1
stiffness
No distinction between primary loading and
unloading or reloading
Dilatancy continues for ever (no critical state)
3
Be careful with undrained behaviour 2
No time-dependency (creep)
35

Page 41
Advanced Computational Geotechnics

H
HKCG 3

EXERCIS
SE 1
F
FOUND
DATIO
ON ON ELAS
STO-PL
LASTIC SOIIL

Dr Wiilliam Ch
heang

Page 42
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

ELASTOPLASTIC ANALYSIS OF A
FOOTING

Computational Geotechnics 1

Page 43
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

INTRODUCTION
One of the simplest forms of a foundation is the shallow foundation. In this exercise we will
model such a shallow foundation with a width of 2 meters and a length that is sufficiently long
in order to assume the model to be a plane strain model. The foundation is put on top of a 4m
thick clay layer. The clay layer has a saturated weight of 18 kN/m3 and an angle of internal
friction of 20.

Figure 1: Geometry of the shallow foundation.

The foundation carries a small building that is being modelled with a vertical point force.
Additionally a horizontal point force is introduced in order to simulate any horizontal loads
acting on the building, for instance wind loads. Taking into account that in future additional
floors may be added to the building the maximum vertical load (failure load) is assessed. For
the determination of the failure load of a strip footing analytical solutions are available from for
instance Vesic, Brinch Hansen and Meyerhof:

Qf
B
= c Nc + 12 0 B N
0
Nq = e tan tan2 (45 + 12 0 )
0
q 1) cot
Nc = (N
0
2(Nq + 1) tan
(V esic)
N = 1.5(Nq 1) tan 0 (Brinch Hansen)

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

This leads to a failure load of 117 kN/ m2 (Vesic), 98 kN/m2 (Brinch Hansen) or 97 kN/m2
(Meyerhof) respectively.

2 Computational Geotechnics

Page 44
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

SCHEME OF OPERATIONS
This exercise illustrates the basic idea of a finite element deformation analysis. In order to
keep the problem as simple as possible, only elastic perfectly-plastic behaviour is considered.
Besides the procedure to generate the finite element mesh, attention is paid to the input of
boundary conditions, material properties, the actual calculation and inspection of some output
results.

Aims
Geometry input

Initial stresses and parameters

Calculation of vertical load representing the building weight

Calculation of vertical and horizontal load representing building weight and wind force

Calculation of vertical failure load.

A) Geometry input

General settings

Input of geometry lines

Input of boundary conditions

Input of material properties

Mesh generation

B) Calculations

Initial pore pressures and stresses

Construct footing

Apply vertical force

Apply horizontal force

Increase vertical force until failure occurs

C) Inspect output

Computational Geotechnics 3

Page 45
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

GEOMETRY INPUT
Start PLAXIS by double-clicking the icon of the PLAXIS Input program. The Quick select
dialog box will appear in which you can select to start an new project or open an existing
one. Choose Start a new project (see Figure 2). Now the Project properties window appears,
consisting of the two tabsheets Project and Model (see Figure 3 and Figure 4).

Figure 2: Quick select dialog

Project properties
The first step in every analysis is to set the basic parameters of the finite element model.
This is done in the Project properties window. These settings include the description of the
problem, the type of analysis, the basic type of elements, the basic units and the size of the
drawing area.

The Project tabsheet

Figure 3: Project tabsheet of the Project Properties window

In order to enter the proper settings for the footing project, follow these steps:

4 Computational Geotechnics

Page 46
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

In the Project tabsheet, enter Exercise 1 in the Title box and type Elasto-plastic
analysis of drained footing or any other text in the Comments box.

In the General options box the type of the analysis (Model) and the basic element type
(Elements) are specified. As this exercise concerns a strip footing, choose Plane strain
from the Model combo box. Select 15-node from the Elements combo box.

The Acceleration box indicates a fixed gravity angle of -90, which is in the vertical
direction (downward). Independent acceleration components may be entered for pseudo-
dynamic analyses. Leave these values zero and click on the Next button below the
tabsheets or click on the Model tabsheet.

The Model tabsheet

Figure 4: Model tabsheet of the Project properties window

In the Model tabsheet, keep the default units in the Units box (Length = m; Force = kN;
Time = day).

In the Geometry dimensions box the size of the considered geometry must be entered.
The values entered here determine the size of the draw area in the Input window.
PLAXIS will automatically add a small margin so that the geometry will fit well within
the draw area. Enter Xmin =0.00, Xmax =14.00, Ymin =0.00 and Ymax =4.25.

The Grid box contains values to set the grid spacing. The grid provides a matrix of dots
on the screen that can be used as reference points. It may also be used for snapping to
regularly spaced points during the creation of the geometry. The distance of the dots is
determined by the Spacing value. The spacing of snapping points can further be divided
into smaller intervals by the Number of snap intervals value. Enter 1.0 for the spacing
and 4 for the intervals.

Click on the Ok button to confirm the settings. Now the draw area appears in which the
geometry model can be drawn.

Computational Geotechnics 5

Page 47
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Hint: In the case of a mistake or for any other reason that the project properties
should be changed, you can access the Project properties window by
selecting the Project properties option from the File menu.

Creating the geometry


Once setting the project properties have been completed, the draw area appears with an
indication of the origin and direction of the system of axes.
The cursor is automatically switched in the Geometry line drawing mode. If not, the user can
change the drawing mode to Geometry line by clicking the geometry line button .
In order to construct the contour of the proposed geometry as shown in Figure 5, follow these
steps. (Use Figure 5 for orientation, it represents the completed geometry).

Figure 5: Geometry model

Create sub-soil

Position the cursor (now appearing as a pen) at the origin (point 0) of the axes (0.0; 0.0).
Click the left mouse button once to start the geometry contour.
Move along the x-axis to (14.0; 0.0). Click the left mouse button to generate the second
point (number 1). At the same time the first geometry line is created from point 0 to point
1.
Move upward to point 2 (14.0; 4.0) and click again.
Move to the left to point 3 (0.0; 4.0) and click again.
Finally, move back to the origin (0.0; 0.0) and click the left mouse button again. Since
the latter point already exists, no new point is created, but only an additional geometry
line is created from point 3 to point 0. PLAXIS will also automatically detect a cluster
(area that is fully enclosed by geometry lines) and will give it a light colour.

6 Computational Geotechnics

Page 48
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Click the right mouse button to stop drawing.


This action created the sub-soil cluster. The next step is to introduce the footing.

Create footing

Position the cursor at point 4, (6.0, 4.0) and click the left mouse button once.
Move vertical to point 5, (6.0; 4.25). Click the left mouse button to generate a vertical
line.
Move horizontal to point 6, (8.0; 4.25). Click the left mouse button to generate a horizontal
line.
Generate a second cluster by clicking the left mouse button on coordinate (8.0; 4.0).
Click the right mouse button to stop drawing.
This action created the footing.
The proposed geometry does not include plates, hinges, geogrids, interfaces, anchors or
tunnels. Hence, you can skip the corresponding buttons in the second toolbar.
Hints: Mispositioned points and lines can be modified or deleted by first choosing the

Selection button from the toolbar. To move a point of line, select the point or
the line and drag it to the desired position. To delete a point or a line, select the
point or the line and press the Delete key on the keyboard.
> Undesired drawing operations can be restored by pressing the Undo button

from the toolbar or by selecting the Undo option from the Edit menu or by
pressing <Ctrl><Z> on the keyboard.
Hint: The full geometry model has to be completed before a finite element mesh can be
generated. This means that boundary conditions and model parameters must be
entered and applied to the geometry model first.
Hint: During the input of geometry lines by mouse, holding down the Shift key will
assist the user to create perfect horizontal and vertical lines.

Input of boundary conditions


Boundary conditions can be found in the second block of the toolbar and in the Loads menu.
For deformation problems two types of boundary conditions exist: Prescribed displacements
and prescribed forces (loads). In principle, all boundaries must have one boundary condition
in each direction. That is to say, when no explicit boundary condition is given to a certain
boundary (a free boundary), the so-called natural condition applies, which is a prescribed
force equal to zero and a free displacement. In order to avoid the situation where the displacements
of the geometry are undetermined, some points of the geometry must have prescribed
displacements. The simplest form of a prescribed displacement is a fixity (zero displacement),
but non-zero prescribed displacements may also be given.
To create the boundary conditions for this exercise, follow the steps below.

Computational Geotechnics 7

Page 49
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Prescribed displacements

Click on the Standard fixities button on the toolbar or choose the Standard fixities option
from the Loads menu to set the standard boundary conditions. As a result PLAXIS will
automatically generate a full fixity at the base of the geometry and roller conditions at the
vertical sides (ux =0; uy =free). A fixity in a certain direction is presented as two parallel lines
perpendicular to the fixed direction. Hence, the rollers appear as two vertical parallel lines and
the full fixity appears as cross-hatched lines.
Hint: The Standard fixities option is suitable for most geotechnical applications. It is
a fast and convenient way to input standard boundary conditions.

Vertical load

Click on the Point load - load system A button on the toolbar or choose the Point load
- static load system A option from the Loads menu to enter another point force. Click on the
coordinate (7.0, 4.25) to enter a point force. As a result PLAXIS will automatically generate a
vertical point force on the indicated point with a unity force (f = 1).

Horizontal load (see also next step "Changing direction .....")

Click on the Point load - load system B button on the toolbar or choose the Point load -
static load system B option from the Loads menu to enter a point force. Click on the coordinate
(7.0, 4.25) to enter a point force. As a result PLAXIS will automatically generate a vertical point
force on the indicated point. As a horizontal force is needed, the direction of load B needs to
be changed.

Changing direction and magnitude of loads

Choose the Selection button from the toolbar. Double click on the geometry point 8 with
coordinate (7.0, 4.25) which will display a box as indicated in Figure 6. Select Point Load -
load system B, click OK and enter 1.0 as x-value and 0.0 as y-value. These values are the
input load of point force B. Click OK to close the window.

Input of material properties


In order to simulate the behaviour of the soil, a proper soil model and corresponding parameters
must be applied to the geometry. In PLAXIS, soil properties are collected in material data sets
and the various data sets are stored in a material database. From the database, a data set
can be assigned to one or more clusters. For structures (like walls, plates, anchors, geogrids,
etc.) the system is similar, but obviously different types of structures have different parameters
and thus different types of data sets. PLAXIS distinguishes between material data sets for Soil

8 Computational Geotechnics

Page 50
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Figure 6: Select window and Point load window

& Interfaces, Plates, Anchors and Geogrids. The creation of material data sets is generally
done after the input of boundary conditions. Before the mesh is generated, all material data
sets should have been defined and all clusters and structures must have their appropriate data
set.

Table 1: Material properties of the clay layer and the concrete footing.
Parameter Symbol Clay Concrete Unit
Material model Model Mohr-Coulomb Linear elastic
Type of behaviour Type Drained Non-porous
Weight above phreatic level unsat 16.0 24.0 kN/m3
Weight below phreatic level sat 18.0 kN/m3
Youngs modulus Eref 5.0103 2.0107 kN/m2
Poissons ratio 0.35 0.15
Cohesion c 5.0 kN/m2
Friction angle 20
Dilatancy angle 0

The input of material data sets can be selected by means of the Material Sets button on
the toolbar or from the options available in the Materials menu.

Create material data sets

To create a material set for the clay layer, follow these steps:

Select the Material Sets button on the toolbar.

Click on the <New> button at the lower side of the Material Sets window. A new dialog
box will appear with five tabsheets: General, Parameters, Flow parameters, Interfaces
and Initial (see figure 7).

In the Material Set box of the General tabsheet, write Clay in the Identification box.

Select Mohr-Coulomb from the Material model combo box and Drained from the Material
type combo box.

Computational Geotechnics 9

Page 51
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Enter the proper values for the weights in the General properties box according to the
material properties listed in table 1

See also figure 8 and figure 9. In these figures the Advanced parameters part has been
collapsed.

Figure 7: General tabsheet of the soil and interface data set window for Clay

Click on the Next button or click on the Parameters tabsheet to proceed with the input of
model parameters. The parameters appearing on the Parameters tabsheet depend on
the selected material model (in this case the Mohr-Coulomb model).

Enter the model parameters of table 1 in the corresponding edit boxes of the Parameters
tabsheet. The parameters in the Alternatives and Velocities group are automatically
calculated from the parameters entered earlier.

Since the geometry model does not include groundwater flow or interfaces, the third and
fourth tabsheet can be skipped. Click on the OK button to confirm the input of the current
material data set.

Now the created data set will appear in the tree view of the Material Sets window.

For the concrete of the footing repeat the former procedure, but choose a Linear Elastic
material behaviour and enter the properties for concrete as shown in table 1 (see also
figures 9 and 10).

10 Computational Geotechnics

Page 52
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Figure 8: Parameters tabsheet of the soil and interface data set window for Clay

Figure 9: General tabsheet of the soil and interface data set window for Concrete

Figure 10: Parameters tabsheet of the soil and interface data set window for Concrete

Computational Geotechnics 11

Page 53
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Assigning material data sets to soil clusters

Drag the data set Clay from the Material Sets window (select it and keep the left mouse
button down while moving) to the soil cluster in the draw area and drop it there (release
the left mouse button). Notice that the cursor changes shape to indicate whether or not
it is possible to drop the data set. When a data set is properly assigned to a cluster, the
cluster gets the corresponding colour. Drag the concrete material set to the footing and
drop it there.
Click on the OK button in the Material Sets window to close the database.
Hint: PLAXIS distinguishes between a project database and a global database of
material sets. Data sets may be exchanged from one project to another using
the global database. In order to copy such an existing data set, click on the
Show global button of the Material Sets window. Drag the appropriate data set
(in this case Clay) from the tree view of the global database to the project
database and drop it there. Now the global data set is available for the current
project. Similarly, data sets created in the project database may be dragged
and dropped in the global database.
Hints: Existing data sets may be changed by opening the material sets window,
selecting the data set to be changed from the tree view and clicking on the Edit
button. As an alternative, the material sets window can be opened by double
clicking a cluster and clicking on the Change button behind the Material set box
in the properties window. A data set can now be assigned to the corresponding
cluster by selecting it from the project database tree view and clicking on the
OK button.
> The program performs a consistency check on the material parameters and will
give a warning message in the case of a detected inconsistency in the data

Mesh generation

When the geometry model is complete, the finite element model (mesh) can be generated.
PLAXIS includes a fully automatic mesh generation procedure, in which the geometry is
automatically divided into elements of the basic element type and compatible structural elements,
if applicable. The mesh generation takes full account of the position of points and lines in the
geometry model, so that the exact position of layers, loads and structures is reflected by
the finite element mesh. The generation process is based on a robust triangulation principle
that searches for optimised triangles, which results in an unstructured mesh. This may look
disorderly, but the numerical performance of such a mesh is usually better than for regular
(structured) meshes. In addition to the mesh generation itself, a transformation of input data
(properties, boundary conditions, material sets, etc.) from the geometry model (points, lines
and clusters) to the finite element mesh (elements, nodes and stress points) is made.
In order to generate the mesh, follow these steps:

Click on the Generate mesh button in the toolbar or select the Generate option from
the Mesh menu. After the generation of the mesh a new window is opened (PLAXIS
Output window) in which the generated mesh is presented (see Figure 11).

12 Computational Geotechnics

Page 54
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Click on the Close button to return to the geometry input mode.

Figure 11: Generated finite element mesh of the geometry around the footing

If necessary, the mesh can be optimised by performing global or local refinements. Mesh
refinements are considered in some of the other exercises. Here it is suggested to accept the
current finite element mesh.
Hints: By default, the Global coarseness of the mesh is set to M edium, which is
adequate as a first approach in most cases. The Global coarseness setting
can be changed in the M esh menu. In addition, there are options available to
refine the mesh globally or locally.

> At this stage of input it is still possible to modify parts of the geometry or to add
geometry objects. In that case, obviously, the finite element mesh has to be
regenerated.

Press the close button to close the output program and return to PLAXIS input.
Creating the input for this project now finished. Press the green Calculation button on the
toolbar to continue with the definition of the calculation phases.

Computational Geotechnics 13

Page 55
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

CALCULATION
After the finite element model has been created, the calculation phases need to be defined.
This analysis consists of four phases. In the initial phase the initial pore pressures and
stresses are generated, in the first phase the footing is constructed, during the second phase
the vertical load is applied and in the third phase the horizontal load is applied.
When starting the PLAXIS Calculation program the Calculation mode window appears. In
this window the user can choose how he wants PLAXIS to handle pore pressures during the
calculation. This is important when calculating with undrained behaviour and/or groundwater
flow. In this first exercise this is not important and so the default setting of Classical mode is
chosen. Press <OK> to close the Calculation mode window. PLAXIS now shows the General
tabsheet of the initial phase (see Figure 12).

Figure 12: General tabsheet of the initial calculation phase

Initial phase (generation of initial conditions)


Before starting the construction of the footing the initial conditions must be generated. In
general, the initial conditions comprise the initial groundwater conditions, the initial geometry
configuration and the initial effective stress state. The clay layer in the current footing project is
fully saturated with water, so groundwater conditions must be specified. On the other hand, the
situation requires the generation of initial effective stresses. As we want to include the footing
construction in the simulation process, the footing should not be present in the initial situation
(prior to construction). In PLAXIS it is possible to switch off clusters in order to calculate
correct initial effective stresses. The initial stresses in this example case are generated using
the K0 -procedure. The initial conditions are entered in separate modes of the Input program.
In order to generate the initial conditions properly, follow these steps:

In the phase list select the initial phase

14 Computational Geotechnics

Page 56
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Make sure the Calculation type is set to K0 -procedure on the General tabsheet. This is
the default setting.

Go to the Parameters tabsheet by clicking the Parameters button or by directly selecting


the tabsheet.

On the Parameters tabsheet press the Define button located in the Loading input box.
This will start a window presenting the problem in Staged construction mode. In Staged
construction mode it is possible to switch on and off various parts of the geometry,
change loads, apply strains etc.

In the initial condition of this exercise, that is the situation before we start constructing
our project, the footing is not present. Therefore the footing has to be deactivated. In
order to do so, click on the area that represents the footing so that it will change color
from the material set color to white. The footing is now disabled.

Click on Water conditions in the button bar in order to move to the Water conditions
mode of the program.

Select the Phreatic level button .

Position the cursor (appearing as a pen) at coordinate (0.0, 4.0) and click the left mouse
button to start the phreatic level.

Move along the x-axis to position (14.0, 4.0). Click the left mouse button to enter the
second point of the phreatic level.

Click the right mouse button to stop drawing.

Press the Water pressures button to view the pore pressures.

The pore pressures are generated from the specified phreatic level and the water weight.
Directly after the generation, a PLAXIS Output window is opened, showing the pore pressure
as presented in Figure 13. The colors indicate the magnitude of pore pressure. The pore
pressures vary hydrostatically, ranging from 0 kN/m2 at the top to -40 kN/m2 at the bottom.

Close the output program in order to return to the input program.

Click on Update in order to save the changes made and return to the PLAXIS Calculations
program. This completes the definition of the initial conditions.

Hints: For the generation of initial stresses based on the K0 procedure it is necessary
to specify the coefficient of lateral earth pressure, K0 . This K0 value is defined
per material set and therefore has to be set when entering material set data. If
the K0 value is not explicitly set PLAXIS uses a value according to Jakys
formula (K0 = 1-sin()).
> The K0 procedure may only be used for horizontally layered geometries with a
horizontal ground surface and, if applicable, a horizontal phreatic level.

Computational Geotechnics 15

Page 57
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Figure 13: Initial pore pressures

First calculation phase (construction of footing)

Click on the Next button . This will introduce a new calculation phase and
present the corresponding tabsheets for the first calculation stage. Enter a suitable name
in the Number/ ID box (e.g. Construction of footing).

Select the second tabsheet called Parameters. On this sheet Staged construction is
selected by default in the Loading input combo box. Click the Define button. This will
open the window presenting the problem in Staged construction mode.

Click on the cluster that represents the strip footing, in order to switch on the footing
(original colour should reappear).

Click on Update to conclude the definition of the first calculation phase. Updating will
automatically present the calculation window.

Second calculation phase (apply vertical load)


Click on the Next button . This will introduce a new calculation phase and present the
corresponding tabsheets for the second calculation stage. Enter a suitable name in the
Number/ ID box (e.g. apply vertical load).

Select the Parameters tabsheet. On this tabsheet accept the selection Staged construction
in the Loading input combo box. Click on the Define button. This will open the window
presenting the problem in Staged construction mode.

Click on the point forces in the middle of the footing, a Select items window comes up.
Select the Point load - Load System A to activate point load A and press the Change
button to change the load value. Change the y-value to -50 kN/m and press the Ok
button.

16 Computational Geotechnics

Page 58
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Figure 14: Parameters tabsheet of the first calculation phase

The point load A is now active (blue) and has a load value of 50 kN/m.
Press Update.

Figure 15: Select items window

Third calculation phase (add horizontal load)


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

Computational Geotechnics 17

Page 59
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Fourth calculation phase (vertical load to failure)


Click on the Next button . This will present the tabsheets for the fourth calculation stage.
Enter a suitable name in the Number/ID box (e.g. vertical load failure).

Directly below the Number/ID box select from the Start from phase dropdown list the
second calculation phase. By selecting this the 4th phase will be a continuation of the
2nd phase, hence we will continue to apply the vertical load without having the additional
horizontal load that was applied in phase 3.

Select the second tabsheet called Parameters. On this sheet choose the selection Total
multipliers in the Loading input group box. Select the third tabsheet called Multipliers by
either clicking on the Define button or directly selecting the tabsheet.

Enter a MloadA of 10. In this way the working force is increased to a maximum load of
10 x 50 = 500 kN/m.

In PLAXIS two methods exist to increase an active load. The magnitude of the
activated load is the input load multiplied by the total load multiplier. Hence, in
this excersise MloadA x (input load of point load A) = Active load A
The value of the input load A can be changed using Staged construction as
Loading input while using Total multipliers as Loading input may be used to
change the load multiplier.

Define load displacement points


After the calculation it is possible to create load-displacement curves. These can be used
to inspect the behaviour in a node during the calculation steps. In order to create load-
displacement curves it is first necessary to indicate for which node(s) the displacements
should be traced.

Click on the Select points for curves button in the toolbar. This will result in a plot of
the mesh, showing all generated nodes. Click on the node, located in the centre directly
underneath the footing. For a correct selection of this node it may be necessary to use
the zoom option . After selection of the node it will be indicated as point A. Press
the Update button to proceed to calculations.

Start the calculation


After definition of the last calculation phase, the calculation process is started by clicking the
Calculation button . This will start the calculation. During the calculation a calculation
window appears showing the status and some parameters of the current calculation phase.

18 Computational Geotechnics

Page 60
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

INSPECT OUTPUT
After each successful execution of a calculation phase PLAXIS will indicate the phase with
a green check mark ( ). This indicates a successful calculation phase. If during execution
either failure or an error occurs, PLAXIS marks the stage with a red cross ( ).

Figure 16: Calculation window with all phases calculated

While phase 3 is highlighted, press the View calculation results button that will start
the output program, showing the deformed mesh for the situation with both horizontal
and vertical load applied, as presented in figure 17.

Figure 17: Deformed mesh at the end of phase 3

Computational Geotechnics 19

Page 61
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Check the various types of output, such as the deformed mesh, displacement contours,
effective (principal) stresses etc. These can be found from the Deformations and
Stresses menus.

Still in the Output program, select from the dropdown list at the right of the toolbar the
output step belonging to phase 4.

From the Displacements menu in the Output program now select Incremental
displacements and then the option |u|. Display the incremental displacements as
contours or shadings. The plot clearly shows a failure mechanism (see Figure 18).

Figure 18: Shadings of displacement increments after phase 4

Load displacement curves


In the Output program, select the Curves manager from the Tools menu. The Curves
manager has 2 tabsheets, one for the curves defined in this project (currently none) and
one for the points selected to make load-displacement curves (currently 1 node that was
pre-selected, that is before the calculation).

In the Curves manager select the button New to define a new curve. Now the Curve
generation window opens.

On the x-axis we want to plot the settlement of our chosen point in the middle of the
footing. In the x-axis box choose point A from the dropdown list and then below in
Deformations and then Total displacements choose |u|.

On the y-axis we want to plot the force applied on the footing, which is a global value
not connected to a specific node or stress point. In y-axis box choose Project from the
dropdown list to indicate we want to plot a global value, and then in Multipliers choose
MLoadA.

20 Computational Geotechnics

Page 62
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Figure 19 shows the Curve generation window after applying the steps mentioned.

Press OK to show the resulting curve. See also figure 20.

Figure 19: Curves generation window

Figure 20: Load displacement curve for the footing

The input value of point load A is 50 kN/m and the load multiplier MloadA reaches approximately
4.6. Therefore the failure load is equal to 50 kN/m x 4.6 = 230 kN/m. You can inspect the load
multiplier by moving the mouse cursor over the plotted line. A tooltip box will show up with the
data of the current location.

Computational Geotechnics 21

Page 63
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

RESULTS DRAINED BEHAVIOUR


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

Table 2: Results for the maximum load reached on a strip footing on the drained sub-soil for
different 2D and 3D meshes

Mesh size Element Nr. of Max. Failure


type elements load load
[kN/m] [kN/m2 ]
very coarse mesh with local refinements 6-noded 79 281 146
under footing
coarse mesh 6-noded 121 270 141
very fine mesh 6-noded 1090 229 121
very coarse mesh with local refinements 15-noded 79 236 124
under footing
coarse mesh 15-noded 121 248 130
very fine mesh 15-noded 1090 220 116
Analytical solutions of:
- Vesic 117
- Brinch Hansen 98
- Meyerhof 97

In this table the failure load has been calculated as:

Qu M aximum f orce M aximum f orce


B
= B
+ concrete d = 2
+6

From the above results it is clear that fine FE meshes give more accurate results. On the other
hand the performance of the 15-noded elements is superior over the performance of the lower
order 6-noded elements. Needless to say that computation times are also influenced by the
number and type of elements.

22 Computational Geotechnics

Page 64
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

ADDITIONAL EXERCISE:

UNDRAINED FOOTING

Computational Geotechnics 23

Page 65
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

24 Computational Geotechnics

Page 66
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

INTRODUCTION
When saturated soils are loaded rapidly, the soil body will behave in an undrained manner, i.e.
excess pore pressures are being generated. In this exercise the special PLAXIS feature for
the treatment of undrained soils is demonstrated.

SCHEME OF OPERATIONS
In PLAXIS, one generally enters effective soil properties and this is retained in an undrained
analysis. In order to make the behaviour undrained one has to select undrained as the Type
of drainage. Please note that this is a special PLAXIS option as most other FE-codes require
the input of undrained parameters e.g. Eu and u .

Aims
The understanding and application of undrained soil behaviour

How to deal with excess pore pressures.

A) Geometry input

Use previous input file

Save as new data file

Change material properties, undrained behaviour for clay

Mesh generation, global mesh refinement

B) Calculations

Re-run existing calculation phases

Construct footing

Apply vertical force

Apply horizontal force

C) Inspect output

Inspect excess pore pressures

Computational Geotechnics 25

Page 67
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

GEOMETRY INPUT

Use previous input file


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

Change material properties


Change material properties by selecting the item Soils & Interfaces from the Materials menu
or click on the Material sets button . Select the clay from the Material sets tree view and click
on the Edit button. On the first tab sheet, General, change the Drainage type to "Undrained
A" and close the data set.

Figure 21: Set drainage type to "Undrained A"

Mesh generation
The mesh generator in PLAXIS allows for several degrees of refinement. In this example
we use the Refine global option from the Mesh menu, which will re-generate the mesh,
resulting in an increased number of finite elements to be distributed along the geometry lines.
Notice the message that appears about staged being reconstructed: the program will take into
account the newly generated mesh for the previously generated initial conditions and staged
construction phases. From the output window, in which the mesh is shown, press the continue
button to return to the Input program.
Hint: After generation of a finer mesh, the geometry may be refined until a
satisfactory result appears. Besides the option Refine global several other
methods of refinement can be used.

26 Computational Geotechnics

Page 68
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Hint: After re-generation of the finite element mesh new nodes and stress points
exists. Therefore PLAXIS has to regenerate pore water pressures and initial
stresses. This is done automatically in the background when regenerating the
mesh. Also, the new mesh is taken into account for any change to calculation
phases with the exception of ground water flow analysis.
After generating the mesh one can now continue to the calculation program. Click on the
Caculations button to proceed to the calculations program. Click yes to save the data.

CALCULATIONS

Re-run existing calculation list


The calculation list from example 1 appears, as indicated below. All phases are indicated
by (blue arrows). After mesh (re)generation, staged construction settings remain and phase
information has been rewritten automatically for the newly generated mesh. However, this is
not the case for points for load displacement curves due to the new numbering of the mesh
nodes.
Click on the Select points for curves button in the toolbar. Reselect the node located
in the centre directly underneath

Click on the Calculate button to recalculate the analysis. Due to undrained behaviour
of the soil there will be failure in the 3rd and 4th calculation phase.

INSPECT OUTPUT
As mentioned in the introduction of this example, the compressibility of water is taken into
account by assigning undrained behaviour to the clay layer. This results normally, after
loading, in excess pore pressures. The excess pore pressures may be viewed in the output
window by selecting:
Select in the calculation program the phase for which you would like to see output results.

Start the output program from the calculation program by clicking the View output button .
Select from the Stresses menu the option Pore pressures and then pexcess , this results in
Figure 22 .

The excess pore pressures may be viewed as stress crosses ( ), contour lines ( ),
shadings ( ) or as tabulated output ( ). If, in general, stresses are tensile stresses
the principal directions are drawn with arrow points. It can be seen that after phase 3 on the
left side of the footing there are excess pore tensions due to the horizontal movement of the
footing. The total pore pressures are visualised using the option of active pore pressures.
These are the sum of the steady state pore pressures as generated from the phreatic level
and the excess pore pressures as generated from undrained loading.

Computational Geotechnics 27

Page 69
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Figure 22: Excess pore pressures at the end of the 3rd phase

Select from the Stresses menu the option Pore pressures and then pactive . The results
are given in Figure 23.

From the load displacement curve it can be seen that the failure load in the last phase is
considerably lower for this undrained case compared to the drained situation, as expected.
For the undrained case the failure load is approx. 70 kPa.

28 Computational Geotechnics

Page 70
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

Figure 23: Active pore pressures at the end of phase 3

Computational Geotechnics 29

Page 71
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

30 Computational Geotechnics

Page 72
Advanced Computational Geotechnics

Elastoplastic analysis of a footing

APPENDIX A: BEARING CAPACITY CALCULATION


Given the formula for bearing capacity of a strip footing:

Qf
B
= c Nc + 12 0 B N
0
Nq = e tan tan2 (45 + 12 0 )
0
q 1) cot
Nc = (N
0
2(Nq + 1) tan
(V esic)
N = 1.5(Nq 1) tan 0 (Brinch Hansen)

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

Filling in given soil data:

Nq = e tan(20) tan2 (55) = 6.4


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

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

The effective weight of the soil:

0 = w 10 kN/m3 = 18 10 = 8 kN/m3
For a strip foundation this gives:

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

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

Qf

L=
B

III
I

II

Computational Geotechnics 31

Page 73
Advanced Computational Geotechnics

Shallow foundation in elasto-plastic soil under


drained and undrained condition

EXERCISE 1

Introduction
1. Terzaghi (1943) derived bearing capacity equations:
a. Based on Prandtl (1920) failure mechanism and the limit equilibrium
method for shallow foundation
b. Shallow foundations embedment depth (Df)< (B)
c. Various bearing capacity equations based on limit equilibrium (Terzaghi,
1943; Meyerhof, 1963; Hansen, 1970; Vesic, 1973 and Davis & Booker,
1971)
2. Modelling of a shallow foundation
a. Strip footing is modelled using solid elements
b. Size is 2 (B) x 1(L) x 0.250 (D)
c. Foundation is located at the surface
3. Soil is elasto-perfect-plastic
a. Deformation
b. Limit to strength based on Mohr-Coulomb failure criterion
4. Model is half-space
5. Loading
a. Vertically loaded (Working Load)
b. Vertically and horizontally loaded (Working Load)
c. Vertically Loaded to ultimate limit state

Page 74
Advanced Computational Geotechnics

2.Geometry

50kN/m
(7,4.25)
(6,4.25)20kN/m (8,4.25) Foundation : Linear Elastic
0.25m
(0,4) (6,4) (8,4) (14,4)
Soil model : Mohr-Coulomb
4.00m

(0,0) (14,0)
14.00m

Geometry: Plaxis 2D Model of Footing in Plane


Strain

Completed geometry with material sets

Page 75
Advanced Computational Geotechnics

Geometry: Meshing

Medium coarseness The generation process is based on:


264 nos of 15-node tria-elements
1. Robust triangulation principle that searches for
2229 nos of nodes optimised triangles
2. Unstructured mesh. This may look disorderly, but the
numerical performance of such a mesh is usually better
than for regular (structured) meshes.

Calculation Phases
1. Initial phase (Default)
2. Phase 1: Construction of foundation
3. Phase 2: Apply vertical working load (50kN/m)
4. Phase 3: Apply vertical and horizontal working load (50 & 20
kN/m)
5. Phase 4: Apply vertical load until ultimate state (from Phase 2 x
sigmaMultiplier)

Sequence of stages and child phase (4)


Phase 3
Initial Phase Phase 1 Phase 2
Phase 4

Page 76
Calculations: Phases
Advanced Computational Geotechnics

Initial Phase [default]


1. Deactivate zone that represent foundation
2. This phase generate initial greenfield soil stresses

Phase 1-Construct Foundation [new]


1. Zone representing foundation is activated.
2. The foundation is wish-in-place

Phase 2-Apply Vertical Load [new]


1. Change vertical load A to 50kN/m

Phase 3-Apply Vertical and Horizontal Load [ne


1. Maintain vertical load of 50kN/m
2. Change horizontal load to 20kN/m

Phase 4- Increase vertical load to failure


1. Enter a SigmaMload A of 10.
2. In this way the working force is increased to a maximum
load of 10 x 50 = 500 kN/m

Displacements : Shadings & Vectors

Phase 2

Phase 3

Phase 4

Page 77
Advanced Computational Geotechnics

Phase 4: Mechanism and Load Factor


4.632

Page 78
Influence of mesh discretization
Advanced Computational Geotechnics and no of
elements
Very Coarse
Elements = 73
Nodes = 647
Load Factor = 4.900

Coarse
Elements = 132
Nodes = 1141
Load Factor = 4.776

Fine
Elements = 523
Nodes = 4347
Load Factor = 4.455

Very Fine
Elements = 1066
Nodes = 8757
Load Factor = 4.416

Reference
1. Terzaghi, K., Theoretical Soil Mechanics, John Wiley, New York, 1943.
2. Vesic, A.S., Analysis of ultimate loads of shallow foundations, J. Soil
Mech. Foundation. Div., ASCE, 99(1), 45, 1973.
3. Meyerhof, G.G., The ultimate bearing capacity of foundations,
Geotechnique, 2,301, 1951.
4. Meyerhof, G.G., Some recent research on the bearing capacity
foundations, Canadian Geotech.J., 1(1), 16, 1963.
5. Hansen, J.B., A revised and extended formula for bearing capacity,
Bulletin No.28, Danish Geotechnical Institute, Copenhagen, 1970.

Page 79
Advanced Computational Geotechnics

H
HKCG 4

CRIT
TICAL
L STAT
TE SO
OIL ME
ECHAN
NICS
AND
SOFT
S S
SOIL MODE
M L
Pro
ofessor Helmut Schweiger

Page 80
Advanced Computational Geotechnics

S C I E N C E P A S S I O N T E C H N O L O G Y

CG 04
CRITICAL STATE SOIL MECHANICS
SOFT SOIL MODEL

Helmut F. Schweiger
Computational Geotechnics Group
Institute for Soil Mechanics and Foundation Engineering
Graz University of Technology

PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


2
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

CONTENTS

Direct shear test


Triaxial tests on saturated clay
Critical state line
Modified Cam Clay model (MCC)
Drained and undrained triaxial stress paths (NC / OC)
Plaxis Soft Soil model
Possible enhancements of Critical State Models

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 81
1
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


3
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

Mayne, 2006
DIRECT
Slow DirectSHEAR
Shear Tests TEST
on Triassic Clay,NC Slow Direct Shear Tests on Triassic Clay, Raleigh, NC
140 140
n ' Strength Parameters:

(kPa)
Shear Stress, (kPa)

120
Peak
(kPa)=
120
c' = 0; ' = 26.1 o
214.5
100 100


80

Shear Stress,
80 Peak

60 135.0 60
0.491 = tan '
40 40
Peak
20 45.1 20

0 0
0 1 2 3 4 5 6 7 8 9 10 0 50 100 150 200 250

Displacement, (mm) Effective Normal Stress, n' (kPa)

v v

Direct Shear Box (DSB) Direct Simple Shear (DSS)


PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


4
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

Mayne, 2006
CC
Void Ratio, e

NC NC

CSL CSL

Log v' CSL


tan'
Shear stress

CSSM Premise:
All stress paths fail on the
critical state line (CSL)

c=0
Effective stress v'
PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 82
2
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


5
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

Mayne, 2006
CC
Void Ratio, e

e0
e NC
NC
ef

CSL CSL

Log v' vo
CSL
STRESS PATH No.1 max = c + tan tan'

Shear stress
NC Drained Soil

Given: e0, vo, NC (OCR=1)

Drained Path: u = 0
Volume Change is Contractive:
c=0
vol = e/(1+e0) < 0 Effective stress v' vo
PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


6
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

Mayne, 2006
CC
Void Ratio, e

e0
NC
NC
CSL
CSL
vf vo Log v' CSL
tan'
STRESS PATH No.2
Shear stress

NC Undrained Soil
max = cu = su u
Given: e0, vo, NC (OCR=1)
Undrained Path: V/V0 = 0
+u = Positive Excess Porewater vf
Pressures vo
Effective stress v'
PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 83
3
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


7
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

Mayne, 2006
CC
Void Ratio, e

CS OC

NC
NC
CSL
CSL

Log v'
p' CSL

Shear stress
tan'
Overconsolidated States:
e0, vo, and OCR = p/vo
where p = vmax = Pc =
preconsolidation stress;
OCR = overconsolidation ratio
Effective stress v' p'
PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


8
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

Mayne, 2006
CC
Void Ratio, e

e0
OC
CS NC NC

CSL CSL

vo' vf' Log v'


CSL
Stress Path No. 3
Shear stress

tan'
Undrained OC Soil: u
e0, vo, and OCR

Stress Path: V/V0 = 0


Negative Excess u
vo' Effective stress v'
PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 84
4
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


9
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

Mayne, 2006
CC
Void Ratio, e

e0
OC
CS NC
NC

CSL CSL

vo'
Log v'
CSL
Stress Path No. 4 tan'
Drained OC Soil:
e0, vo, and OCR

Stress Path: u = 0

Dilatancy: V/V0 > 0


vo'
Effective stress v'
PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


10
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

a) b)

Typical results from drained (a) and undrained (b) triaxial tests on normally consolidated soils
(from Atkinson & Bransby, 1978)
PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 85
5
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


11
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

a)
b)
Typical results from drained (a) and undrained (b) triaxial tests on overconsolidated soils
(from Atkinson & Bransby, 1978)
PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


12
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

DRAINED TRIAXIAL TEST (NC)

1+e

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 86
6
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


13
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

UNDRAINED TRIAXIAL TEST (NC)

PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


14
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 87
7
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


15
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

NCL AND CSL IN p-q-v - SPACE

PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


16
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

UNDRAINED PLANES

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 88
8
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


17
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

DRAINED PLANES

PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


18
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

STATE BOUNDARY SURFACE

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 89
9
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


19
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

OVERCONSOLIDATION

OCR is very important for


soil behaviour

PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


20
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

In invariant formulation: F
M pc 2 3 2
1 0
pc
2
Mcs pc 2

3 sin '
and MC-failure condition: M cs
3 cos sin sin '

d vol P 1 e0
pc pco

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 90
10
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


21
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

Right from the M-line (wet side): q < M p

d vp 0, d sp 0 (contraction, hardening)

Left from the M-line (dry side): q > M p

d vp 0, d sp 0 (dilatancy, softening)

On the ellipse top: q = M p M


1
d vp 0, d sp Failure!

CSL f=0
PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


22
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

MODIFIED CAM CLAY WITH MOHR COULOMB

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 91
11
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


23
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

Stiffness: primary loading

B-C-D-E-F: slope of yield locus becomes flatter


ratio distortional/volumetric strain becomes larger
normally consolidated
drained compression
D. Muir Wood, 1990
PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


24
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

Stiffness:
- unloading / reloading
- primary loading

lightly overconsolidated
drained compression D. Muir Wood, 1990
PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 92
12
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


25
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

heavily overconsolidated
drained compression

D. Muir Wood, 1990


PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


26
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

due to plastic soil behaviour

due to change of stress

normally consolidated
undrained compression

D. Muir Wood, 1990


PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 93
13
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


27
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

due
duetotoplastic
plasticsoil
soilbehaviour
behaviour

due
duetotochange
changeofofstress
stress

lightly overconsolidated
undrained compression

D. Muir Wood, 1990


PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


28
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

due to change of stress

due to plastic soil behaviour

heavily overconsolidated
undrained compression

D. Muir Wood, 1990


PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 94
14
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


29
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

Elastic deformation is generated according to:


p'
e e 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

PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


30
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

Generally we prefer notation in strains:

p'
ve ve 0 * ln 0 , *
p 1 e
p'
vp vp 0 ( * * ) ln 0
, *

p 1 e

v = volumetric strain
* = modified swelling index
* = modified compression index

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 95
15
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


31
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

PLAXIS SOFT SOIL MODEL

M
q 1 MC-line
K0NC


p
pp

Soft Soil model:


Mohr-Coulomb failure surface for strength
M-line for determining K0NC
(no longer acts as CSL, only determines shape of cap)

PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


32
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

p
Input Parameters:
*
* = / 1+e .. Modified compression index
* = / 1+e .. Modified swelling index *

c Cohesion
Friction angle
Dilatancy angle
ur .. Poisson's ratio for unloading

K0nc . Coefficient of lateral earth pressure in normal consolidation


M . K0nc parameter

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 96
16
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


33
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

SS vs MC MODEL - OEDOMETER TEST


Chart 1
vertical strain
0
SS

MC

-0.1

-0.2

-0.3

-0.4

-0.5
0 -100 -200 -300 -400 -500 -600
vertical stress [kN/m2]

vertical stress vs vertical strain


PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


34
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

SS vs MC MODEL - OEDOMETER TEST Chart 1


vertical stress [kN/m2]
-600
SS

MC

-500

-400

-300

-200

-100

0
0 -50 -100 -150 -200 -250 -300 -350
horizontal stress [kN/m2]

horizontal stress vs vertical stress


PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 97
17
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


35
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

PRIMARY LOADING - UNLOADING / RELOADING

elastic
Stiffness: primary loading
region

Stiffness: unloading / reloading

current yield surface

Stiffness: unloading / reloading

PLAXIS Hong Kong Course, 6 - 9 November 2012

Critical State / PLAXIS Soft Soil model


36
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

MCC-MODEL - FURTHER DEVELOPMENTS

"Bubble models" with


kinematic hardening
e.g. 3-SKH Model
(Baudet & Stallebrass, 2004)

Anisotropic models based on


Modified Cam Clay (rotated yield
surfaces)
e.g. Wheeler, Ntnen, Karstunen
& Lojander (2003)

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 98
18
Advanced Computational Geotechnics

Critical State / PLAXIS Soft Soil model


37
Direct Shear Test | Triaxial Test | Critical State Line | Modified Cam Clay | Stress Paths | Plaxis Soft Soil | Possible Enhancements

MCC-MODEL - FURTHER DEVELOPMENTS

Atkinson & Sallfors (1991)


e.g. Leroueil & Vaughan (1990)

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 99
19
Advanced Computational Geotechnics

H
HKCG 5

H
HARDENING
G SOIL
L AND
D HS-small M
MODEL
L

ofessor Helmut Schweiger


Pro

Page 100
Advanced Computational Geotechnics

CG5
HARDENING SOIL SMALL MODEL

Helmut F. Schweiger
Computational Geotechnics Group
Institute for Soil Mechanics and Foundation Engineering
Graz University of Technology

PLAXIS HONG KONG COURSE / 6 - 9 November 2012

HARDENING SOIL (SMALL) MODEL


CONTENTS

Introduction (why advanced model?)


Short description of Hardening Soil Model
Parameters of Hardening Soil Model
Comparison with experimental data
Influence of important parameters
Extension to account for small strain stiffness
(HS-Small)
Summary

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 2

Page 101
1
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


INTRODUCTION
Soil behaviour includes:
difference in behaviour for primary loading
reloading/unloading
nonlinear behaviour well below failure conditions
stress dependent stiffness
plastic deformation for isotropic or K0-stress paths
dilatancy is not constant
small strain stiffness
(at very low strains and upon stress reversal)
influence of density on strength and stiffness

cannot be accounted for with simple


elastic-perfectly plastic constitutive models
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 3

HARDENING SOIL (SMALL) MODEL


INTRODUCTION
oedometer test

1-

HS MC

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 4

Page 102
2
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


INTRODUCTION

q [kN/m2]
250

Mohr Coulomb Model


200

HS-Model
150

100

50

0
0 0.01 0.02 0.03 0.04 0.05
eps_axial

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 5

HARDENING SOIL (SMALL) MODEL


INTRODUCTION

smax
Model
[mm]

LE 33

MC 36

HS 60

distance [m ]

0 3 6 9 12 15
-0,2

0,2
s / smax [-]

0,4
Linear Elastic
0,6
Mohr Coulomb All models calculate settlements
0,8 Hardening Soil Differences in shape of trough
and maximum values
1

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 6

Page 103
3
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


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

Hard. Soil
60

40

20

0
0 5 10 15 20
-20

-40

distance from wall [m]

> Hardening Soil Model calculates Settlements


> Mohr-Coulomb Model calculates Heave

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 7

HARDENING SOIL (SMALL) MODEL


TRIAXIAL TEST

Applied stress path and results for standard drained triaxial test
1
1
3= constant
3

1
isotropic loading
q 1 3 1

dense soil
3

loose dense
qf qr

1 1
qf = failure value vol loose

qr = residual value vol volumetric 1 2 3

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 8

Page 104
4
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


HYPERBOLIC APPROXIMATION OF CURVE FROM STANDARD DRAINED TEST

q 1 3
E50
1

50%
Hyperbola

50%

1
qa q
3 a
m 1
E50 Eref
50
2 E50 qa q
pref a

Eref
50 = reference modulus for primary loading at 50% of strength

msand 0.5 ; mclay 1

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 9

HARDENING SOIL (SMALL) MODEL


HYPERBOLIC APPROXIMATION OF CURVE FROM STANDARD DRAINED TEST

q 1 3

Asymptote

qf
qa R f 0.9
Hyperbola qf
Rf

hyperbolic for q < qf otherwise q = qf

2 sin
qf 3 a a c cot MC failure criterion
1 sin

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 10

Page 105
5
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


HYPERBOLIC APPROXIMATION OF CURVE FROM STANDARD DRAINED TEST

q 1 3

Asymptote

qa
Hyperbola

qa q
1
2 E 50 qa q

3 3 qa q
shear strain 1 3 1
2 4 E 50 qa q

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 11

HARDENING SOIL (SMALL) MODEL


SHEAR STRAIN CONTOURS IN P-Q-PLANE

c 0

q q
0.05

0.01

p p

sands : m 0.5 clays : m 1


curved lines straight lines

m
3 qa q 3 a 2 sin a
E50 Eref qa (3 a)
50 1 sin a
4E50 q qa pref a
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 12

Page 106
6
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


LINES OF EQUAL SHEAR STRAINS IN TRIAXIAL TEST

Ref. : Ishihara, Tatsuoka and Yasuda (1975). Undrained deformation and liquefaction of sand under
cyclic stresses. Soils and Foundations, Vol. 15, No. 1.

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 13

HARDENING SOIL (SMALL) MODEL


HYPERBOLIC APPROXIMATION OF CURVE FROM STANDARD DRAINED TEST

2c ' cos ' 2 '3 sin '


R f qa Failure according to MC criterion
1 sin '

m
c ' cos ' '3 sin '
E50 E ref
50 c ' cos ' p ref sin '

m
c ' cos ' '3 sin '
Eur E ref
ur c 'cos ' p ref sin '

Note: Stress-dependent stiffness based on 3

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 14

Page 107
7
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


DEFINITION OF E50

E50, reference pressure = 200 kPa

E50, reference pressure = 100 kPa


1 3 kPa

3 600 kPa

1500

300

500
100

0 0.1 0.2 0.3

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 15

HARDENING SOIL (SMALL) MODEL


STIFFNESS IN UNLOADING-RELOADING

Triaxial tests:
Unloading is
purely elastic in
HS model

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 16

Page 108
8
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


LINES OF EQUAL VOLUMETRIC STRAINS IN TRIAXIAL TEST

Biarez, J. & Hicher, P.-Y. (1994), Elementary Mechanics of Soil Behaviour, Balkema - Publishers.
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 17

HARDENING SOIL (SMALL) MODEL


LINES OF EQUAL VOLUMETRIC STRAINS IN TRIAXIAL TEST

q [MN/m]

vol 0.25

vol 0.20
0.14
0.07

p [MN/m]

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 18

Page 109
9
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


DENSITY HARDENING IN THE HS MODEL

q
MC failure line

pp Cap

pp p

1 m
c ' cot ' p p is determined by K0nc
p

1 m c ' cot ' p ref


v
is determined by Eoedref

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 19

HARDENING SOIL (SMALL) MODEL


DEFINITION OF Eoed

m
c cot 1
Eoed E ref
oed c cot p holds strictly for K0-stress paths only

PLAXIS HONG KONG COURSE / 6 - 9 November 2012
ref
20

Page 110
10
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


ELASTIC REGION

f < 0 + fc < 0
q

fc
f
1
pc
p
q = 1 - 3
REGION 1
p = (1 + 2 + 3) / 3 no yield surface active > elastic

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 21

HARDENING SOIL (SMALL) MODEL


SHEAR HARDENING

f > 0 + fc < 0
q

2
fc
f

pc
p

REGION 2
shear hardening surface active
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 22

Page 111
11
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


SHEAR AND VOLUMETRIC HARDENING

q f > 0 + fc > 0

fc
f

pc
p

REGION 3
shear hardening and volumetric hardening surfaces active
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 23

HARDENING SOIL (SMALL) MODEL


VOLUMETRIC HARDENING

f < 0 + fc > 0
q

f
4
pc p

REGION 4
volumetric hardening surface active
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 24

Page 112
12
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


PLASTICITY IN HS MODEL

q 1. Elasticity
(unloading reloading)
2. Plasticity
5 4 (compression)
3 3. Plasticity
(shear)
1 1 2 4. Plasticity
(compression + shear)
1 1 5. Plasticity
(failure criterion)
p

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 25

HARDENING SOIL (SMALL) MODEL


HARDENING SOIL MODEL IN PRINCIPAL STRESS SPACE

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 26

Page 113
13
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


FLOW RULE

Volumetric behaviour
"stress dilatancy theory" (Rowe, 1962)

dilatancy angle > non-associated flow rule

vp sin m p
sin m sin cv
sin m
1 sin m sin cv

1' 3'
sin m
1' 3' 2c cot

sin sin
sin cv
1 sin sin

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 27

HARDENING SOIL (SMALL) MODEL


FLOW RULE
mobilized dilatancy angle for = 35
35
30 =0
25 =5 dilation
mobilized dilatancy angle []

20 = 20
15 = 35
10
5
0
-5
-10
-15
-20
-25 contraction
-30 cv cv
-35
0 5 10 15 20 25 30 35

mobilized friction angle []


q

negative values of
are cut-off in Plaxis
cv
mob

p'
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 28

Page 114
14
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


FLOW RULE

FLOW RULE sin m sin cv


sin m
"stress dilatancy theory" (Rowe, 1962) 1 sin m sin cv

plastic potential Q
m > cv = positive > dilation
q

x cv
x m = cv = 0

x
m < cv = negative > contraction
p'

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 29

HARDENING SOIL (SMALL) MODEL


PARAMETERS OF HARDENING SOIL MODEL

friction angle
c cohesion
dilatancy angle

ref
E 50 secant modulus from triaxial test (controls deviatoric hardening)

ref
E oed tangential modulus from oedometer test
(controls volumetric hardening)

ref
E ur unloading / reloading modulus

ref ref
default: E ur = 3 E 50

m power for stress dependency of stiffness

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 30

Page 115
15
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


PARAMETERS OF HARDENING SOIL MODEL

ur Poisson ratio for unloading / reloading (default ur = 0.2)

pref reference stress


(default pref = 100 stress units)

K0nc K0-value for normal consolidation (default = 1-sin)


(controls volumetric hardening)

Rf = qf / qa (default Rf = 0.9)

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 31

HARDENING SOIL (SMALL) MODEL


PARAMETERS OF HARDENING SOIL MODEL

SANDS

ref
E oed I D 60 MPa Correllation for pref=100 kPa (Lengkeek)
ID = relative density
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 32

Page 116
16
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


PARAMETERS OF HARDENING SOIL MODEL

Stiffness of sand

ref
E50 Eoed
ref
How can this be true?
p ref

-1= p ref

-3 = p ref

-3 -1

Note: always plot Eoed, Eur distribution for initial stress state when
using HS-model
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 33

HARDENING SOIL (SMALL) MODEL


PARAMETERS OF HARDENING SOIL MODEL

Loose sands: E50ref 15 MPa


Dense sands: E50ref 50 MPa

m for clays:
approx. 0.9 1.0
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 34

Page 117
17
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


PARAMETERS OF HARDENING SOIL MODEL

For normally consolidated clays (m=1):

ref
Eoed 12 E50
ref
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 * Relation with Soft Soil model

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 35

HARDENING SOIL (SMALL) MODEL


PARAMETERS OF HARDENING SOIL MODEL

drained

Ohde / Janbu:
m
1
Eoed E ref
oed ref
p

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 36

Page 118
18
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


PARAMETERS OF HARDENING SOIL MODEL

Parameter limitations

HS model has internal parameters that are computed from our


engineering input parameters > not all combinations of input
parameters can be used. For very soft soils this could be a
problem in certain cases.

E50
pc
Eur E oed
e.g. E50 / Eoed > 2 difficult to input
p
double hardening

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 37

HARDENING SOIL (SMALL) MODEL


COMPARISON WITH EXPERIMENTAL DATA

Triaxial tests on loose sand

|1-3| [kPa]
1000

800

600
Hardening soil model
400
test data
200

0
0 5 10 15
v [%] -1 [%]
0 5 10 15
1

-1

-2

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 38

Page 119
19
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


COMPARISON WITH EXPERIMENTAL DATA

Triaxial tests on dense sand


|1-3| [kPa]
1400

1200

1000

800

600 Hardening soil model


test data
400

200

0
0 5 10 15
-1 [%]

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 39

HARDENING SOIL (SMALL) MODEL


COMPARISON WITH EXPERIMENTAL DATA

Triaxial tests on dense sand


v [%]
8

3
Hardening soil model
2
test data

-1

-2
0 5 10 15
-1 [%]

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 40

Page 120
20
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


COMPARISON WITH EXPERIMENTAL DATA

Oedometer tests on loose sand


-yy [kPa]
400

Hardening soil model

test data

300

200

100

0
0 0.005 0.01 0.015 0.02 0.025
-yy

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 41

HARDENING SOIL (SMALL) MODEL


COMPARISON WITH EXPERIMENTAL DATA

Oedometer tests on dense sand


-yy [kPa]
400

Hardening soil model


300
test data

200

100

0
0 0.005 0.01 0.015 0.02 0.025
-yy

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 42

Page 121
21
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


INFLUENCE E50 / Eoed

deviatoric
Stress [kN/m2] stress Triaxial Compression (drained)
250

Reference values
200

150
Eoed increased

100

50

E50 reduced

0
0 -5.00E-03 -0.010 -0.015 -0.020 -0.025 -0.030
Displacement [m]
axial strain

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 43

HARDENING SOIL (SMALL) MODEL


INFLUENCE E50 / Eoed

volumetric
Strain strain Triaxial Compression (drained)
0.000

-2.00E-03

-4.00E-03
Eoed increased

-6.00E-03

-8.00E-03
Reference values

-0.010
0 -5.00E-03 -0.010 -0.015 -0.020 -0.025 -0.030 -0.035

Strain
axial strain

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 44

Page 122
22
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


OEDOMETER TEST - COMPARISON MC / SS / HS

0.0

-0.1
vertical strain

-0.2
Hardening Soil Model
Soft Soil Model
Mohr Coulomb Model

-0.3

-0.4

-0.5
0 -100 -200 -300 -400 -500 -600
vertikal stress [kN/m2]

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 45

HARDENING SOIL (SMALL) MODEL


OEDOMETER TEST - COMPARISON MC / SS / HS

-600

-550

-500

-450 Mohr-Coulomb model:


vertical stress [kN/m2]

ratio 3/1 determined by


-400

-350

-300

-250
Hardening (Soft) Soil model:
Hardening Soil Model ratio 3/1 determined
Soft Soil Model
-200 by K0nc
Mohr Coulomb Model

-150
Unloading: ur
-100

-50

0
0 -50 -100 -150 -200 -250 -300 -350
horizontal stress [kN/m2]

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 46

Page 123
23
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


OVERCONSOLIDATION IN HS-MODEL Cap position based on
previous stress history
Initial pre-consolidation pressure
(p*, q*)
pp0 relates to initial p0 q
Calculation of p0 based on: pp
OCR (Over-Consolidation Ratio)
POP (Pre-Overburden Pressure)
(p*,q*)

pP = Isotropic pre-cons. pressure


p = Vertical pre-cons. pressure

'0yy , p p
'0xx , p '0zz , p K0nc p pp p

p * = 13 '0xx , p '0yy , p '0zz , p current initial stress state


Stress point due to

q* | '0xx , p '0yy , p |
2
q*
p p ( p*)2

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 47

HARDENING SOIL (SMALL) MODEL


OVERCONSOLIDATION

Calculation of p0 based on OCR: Calculation of p0 based on POP:

p 0 OCR ' yy 0 p 0 ' yy 0 POP

p0
OCR POP
'yy0

'yy0 p0 'yy0 p0

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 48

Page 124
24
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


PLASTIC POINTS

-1

Cap point

Mohr-Coulomb point
f<0
Cap & Hardening point

Hardening point

-3
-1

Tension point
-3
Tension cut-off: Principal tensile stress is set to zero

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 49

HARDENING SOIL (SMALL) MODEL


PLASTIC POINTS

elastic-plastic

elastic

double hardening

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 50

Page 125
25
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


SUMMARY

Mohr- Hardening
Coulomb Soil Model
Model

stress dependent stiffness*


NO YES

distinction in stiffness for primary


loading and unloading / reloading
NO YES

plastic strains for stress states


below MC - failure line
(deviatoric and volumetric hardening) NO YES

failure according to
Mohr-Coulomb YES YES

* (not only dependent on 0, this is possible also with MC-Model)

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 51

HARDENING SOIL (SMALL) MODEL


SMALL STRAIN STIFFNESS

Typical curve of degradation of stiffness with strain

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 52

Page 126
26
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


THE 1-DIMENSIONAL MODEL BY HARDIN & DRNEVICH

1.0

0.8
In its modified form, the reduction curve is specified
by the two parameters
0.6 G0 (or E0) and 0.7
G/G0[]

Experimental results after Santos & Correia


0.4 G 1

G0 1 a( / 0.7 )
0.2
a = 3/7

0.0
1e3 1e2 1e1 1e0 1e1 1e2 1e3
normalized shear strain / 0.7 [-] /0.7[]
NormalisierteScherdehnung

J.A. Santos, A.G. Correia. Reference threshold shear strain of soils. Its application to obtain a unique strain-dependent
shear modulus curve for soils. Proc. 15th ICSMGE, Istanbul (2001), Vol.1, pp. 267-270.

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 53

HARDENING SOIL (SMALL) MODEL


SMALL STRAIN STIFFNESS BEHAVIOUR IN THE HS-SMALL MODEL
If the small strain stiffness model after Hardin & Drnevich predicts a stiffness lower
than Gurref (Eurref respectively) the model switches to hardening plasticity of the
standard Hardening Soil model.
IMPORTANT NOTE: flow rule for deviatoric yield surface (volumetric behaviour) is
different in HSS model.
additional input parameters for HS-small
G0ref
40000 40000
TangentmodulusG[kN/m]
SecantmodulusG[kN/m]

30000 0.7 30000

20000 20000

Gurref
10000 10000
HSSmall
Hardin&Drnevich
0 0
1E5 0.0001 0.001 0.01 1E5 0.0001 0.001 0.01
Shearstrain[] Shearstrain[]

see also:
Thomas Benz, Small-Strain Stiffness of Soils and its Numerical Consequences,
Mitteilung 55 des Instituts fr Geotechnik, Universitt Stuttgart, 2007.

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 54

Page 127
27
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


SMALL STRAIN STIFFNESS BEHAVIOUR IN THE HS-SMALL MODEL

Drained triaxial test, HS vs. HS-small model

160

140

120
HSsmall E0

HS
q [kN/m]

100

80

60
E0ref
40 G ref

2(1 ur )
0

20
Et Eur
0
0 -0.002 -0.004 -0.006 -0.008 -0.01 -0.012 -0.014

yy

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 55

HARDENING SOIL (SMALL) MODEL


DATA FOR SMALL STRAIN STIFFNESS

Hsu & Vucetic Vucetic & Dobry

Shear strain Shear strain

C-C. Hsu, M. Vucetic. Dynamic and cyclic behavior of soils over a wide range of shear strains in
NGI-type simple shear testing device, UCLA Report ENG-02-228, 2002.

M. Vucetic, R. Dobry. Effect of soil plasticity on cyclic response, Journal of Geotechnical


Engineering, ASCE 117 (1991), No. 1, 89-107.

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 56

Page 128
28
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


DATA FOR SMALL STRAIN STIFFNESS

Typical values for G0 (in MPa) for reference stress of 100 kPa

Jamiolkowski et al. 1991

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 57

HARDENING SOIL (SMALL) MODEL


DATA FOR SMALL STRAIN STIFFNESS

(2.97 e)2
G0ref 33 [MPa] Hardin & Black (1969)
1 e
G0ref RD 70MPa 60MPa Lengkeek
0.385
0.7 2c(1 cos(2 )) 1(1 K0 )sin(2 ) Benz (2007)
4G0

Order of magnitude:

G0ref (2.5 to 10)Gurref

0.7 (1 to 2) 104 E urref


where G ref

2(1 ur )
ur

PLAXIS HONG KONG COURSE / 6 - 9 November 2012 58

Page 129
29
Advanced Computational Geotechnics

HARDENING SOIL (SMALL) MODEL


SMALL STRAIN STIFFNESS BEHAVIOUR IN THE HS-SMALL MODEL

Example Deep Excavation

Region of larger Region of small


strains: G ~ Gur strains: G > Gur
0
3.0
20

40
Gm=Gref /Gurref

Gm []
2.0
60

80
1.0
100

20 0 20 40 60 80 100 120
see also:
Thomas Benz, Small-Strain Stiffness of Soils and its Numerical Consequences,
Mitteilung 55 des Instituts fr Geotechnik, Universitt Stuttgart, 2007.
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 59

Page 130
30
Advanced Computational Geotechnics

H
HKCG 6

EXERCISSE 2
TRIAX AND OEDOM
XIAL A METER
R

Dr Wiilliam Ch
heang

Page 131
Advanced Computational Geotechnics

Simulation of laboratory tests

SIMULATION OF LABORATORY TESTS

Computational Geotechnics 1

Page 132
Advanced Computational Geotechnics

Simulation of laboratory tests

2 Computational Geotechnics

Page 133
Advanced Computational Geotechnics

Simulation of laboratory tests

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.

Computational Geotechnics 3

Page 134
Advanced Computational Geotechnics

Simulation of laboratory tests

Figure 1: Oedometer test results on sand

Figure 2: Development of horizontal and vertical stress in oedometer test

4 Computational Geotechnics

Page 135
Advanced Computational Geotechnics

Simulation of laboratory tests

Figure 3: Triaxial test unloading-reloading (cell pressure = 100 kPa)

Figure 4: Axial vs. volume strain in drained triaxial test

Computational Geotechnics 5

Page 136
Advanced Computational Geotechnics

Simulation of laboratory tests

Collect the soil parameters in table 1:

Table 1: Hardening Soil Parameters of the sand


Parameter Unit Value
ref
E50 [kPa]
ref
Eoed [kPa]
ref
Eur [kPa]
pref [kPa]
ur [-]
c [kPa]
0 [o ]
[o ]
m [-]
K0N C [-]

With these data perform a triaxial test in the SoilTest program.

6 Computational Geotechnics

Page 137
Advanced Computational Geotechnics

Simulation of laboratory tests

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

Computational Geotechnics 7

Page 138
Advanced Computational Geotechnics

Simulation of laboratory tests

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


Parameter Unit Value
ref
E50 [kPa]
ref
Eoed [kPa]
ref
Eur [kPa]
pref [kPa]
ur [-]
c [kPa]
0 [o ]
[o ]
m [-]
K0N C [-]

With these data perform an oedeometer test in the SoilTest program.

8 Computational Geotechnics

Page 139
Advanced Computational Geotechnics

Simulation of laboratory tests

APPENDIX A: PARAMETER DETERMINATION


SAND
First we determine parameters from the triaxial test data.

Figure 8: Determine stiffness parameters from drained triaxial test

Cohesion and friction angle

For a cell pressure 30 = 100 kPa a maximum value of approximately |10 30 | = 400 kPa is reached at failure.
The Mohr-Coulomb failure criterium is:

1 0
2 |1 30 | + 12 (10 + 30 ) sin c cos = 0

Considering it is sand we assume that the cohesion is zero and so the Mohr-Coulomb failure criterium reduces
to:

|10 30 |
(10 +30 ) = sin

Filling in 30 = 100 kPa and 10 = 500 kPa as obtained from the test we find for the
friction angle0 = 420

Reference stiffness from triaxial test

The triaxial test stiffness E50 is the secant stiffness over the first 50% of the failure value for | 10 30 |. This is
indicated in red in the triaxial test graph of figure 8.

0 =100 kP a 400
E503 = 0.013 = 30800 kP a

Computational Geotechnics 9

Page 140
Advanced Computational Geotechnics

Simulation of laboratory tests

The triaxial test stiffness ,E 50 , is within the Hardening Soil model defined as:

m m
c cos30 sin 30
 
ref ref
E50 = E50 c cos+pref sin , c = 0 E50 = E50 pref

The reference stress pref is chosen equal to the cell pressure of this triaxial test then

ref 0 =100 kP a
E50 = E503 30000 kPa

Reference unloading-reloading stiffness

Similar to the determination of the reference stiffness for triaxial testing the reference unloading-reloading stiffness
can be determined. In the triaxial test results an unloading-reloading cycle is done for this. The Hardening Soil
model does not have unloading-reloading behaviour with hysteresis but simple non-linear elastic unloading-
reloading behaviour. Therefore a secant value is taken for the unloading-reloading behaviour, as given with the
green line in the triaxial test results.

0 =100 kP a 400
Eur3 = 0.0260.021 = 80000 kPa

Under the same assumptions as for the stiffness in triaxial testing counts:

ref 0 =100 kP a
Eur = Eur3

But this is a bit low value for the unloading reloading stiffness and so

ref
Eur = 90000 kPa

is chosen

Dilatancy angle

From the plot of axial strain versus volume strain the dilatancy angle can be determined according to

v
sin = 21 +v

See figure 9 for details.


With v = 0.048-0.004 = 0.044 and 1 = -0.09-(-0.03) = -0.06 the dilatancy can be calculated as =16o
Note: The Poissons ratio needed for the Hardening Soil model cannot be determined from this graph as this
graph represents an oedometer test in primary loading and the Poissons ratio needed is an unloading-reloading
Poissons ratio.
An acceptable value for the unloading-reloading Poissons ratio is ur = 0.2.

10 Computational Geotechnics

Page 141
Advanced Computational Geotechnics

Simulation of laboratory tests

Figure 9: Determination of diltancy angle from drained triaxial test

Oedometer stiffness and power of stress dependent stiffness

From the oedeometer test results we determine the stiffness Eoed for vertical stresses y0 = 100 kPa en y0 =
200 kPa, see figure 10. Note that Eoed is a tangent stiffness. Make sure to use the primary loading part of the
oedometer test results.

0 =100 kP a
y 3200
Eoed = 1.4%0.33% = 29900 kPa
y0 =200 kP a 4000
Eoed = 1.4%0.47% = 43000 kPa

Within the Hardening Soil model the stress dependent oedometer stiffness is defined as:

m m
c cosy0 sin y0
 
ref ref
Eoed = Eoed c cos+pref sin , c = 0 Eoed = Eoed pref

Choosing the reference pressure pref = 100 kPa gives

ref 0 =100 kP a
Eoed = Eoed
3
30000 kPa

The power m for stress dependent stiffness can now be determined as:

0 =200 kP a m
y
y0

Eoed 43000 200 m

ref
Eoed
= pref 30000 = 100 m = 0.5

Computational Geotechnics 11

Page 142
Advanced Computational Geotechnics

Simulation of laboratory tests

Figure 10: Determination of oedometer stiffness and power of stress dependency

K0 value for normal consolidation

The K0 value for normal consolidation (K0N C )can only be obtained if measurements for horizontal stresses have
been performed during the oedometer test. If so, results as given in figure 11 may be obtained. From the primary
loading line can be obtained that

0
x 30 100
K0N C = y0 = 10 = 300 = 0.33

Alternatively one can use Jakis formula

K0N C 1 sin = 1 sin(42o ) = 0.33

12 Computational Geotechnics

Page 143
Advanced Computational Geotechnics

Simulation of laboratory tests

Figure 11: Horizontal/vertical stress ratio during oedometer test

Note on unloading-reloading stiffness

If no triaxial test with unloading-reloading is available the unloading-reloading stiffness can also be determined
from an oedometer test with unloading. However, the unloading-reloading stiffness required for the Hardening
Soil model is stress dependent on 3 while the oedometer test results presented in figure 10 give the strain vs
the vertical stress y (= 1 voor oedometer testing).

0 =100 kP a 0 =100/K0N C kP a 0 =300 kP a 400


Eur3 = Eur1 = Eur1 = 1.28%0.91% = 108000 kPa

With pref = 100 kPa (pref refers to 30 !) it follows that

ref 0 =100 kP a
Eur = Eur3 110000 kPa

ref
This is a bit high and so a value of Eur = 90000 kPa is chosen.

Table 3: Summary of Hardening Soil Parameters for the sand


Parameter Unit Value
ref
E50 [kPa] 30,000
ref
Eoed [kPa] 30,000
ref
Eur [kPa] 90,000
pref [kPa] 100
ur [-] 0.2
c [kPa] 0
0 [o ] 42
[o ] 16
m [-] 0.5
K0N C [-] 0.33

Computational Geotechnics 13

Page 144
Advanced Computational Geotechnics

Simulation of laboratory tests

CLAY

Cohesion and friction angle

We start with the determination of the strength parameters based on the CU triaxial tests.

Figure 12: Determination of soil strength parameters for clay

The black dotted lines is the Mohr-Coulomb failure criterium in the p-q plane. In principal stresses the Mohr-
Coulomb failure criterium is defined as:

|1 3 | 1 +3

2 + 2 sin c cos = 0

With p0 = (10 + 230 )/3 and q = 10 30 under triaxial test conditions this can be rewritten as:

2p0 + 13 q
 
q 6sin 0 6c cos
2 = 2 sin c cos = 0 q = 3sin p + 3sin

Hence, the slope M of the Mohr-Coulomb line in p-q plane is defined as:

6sin 195
M= 3sin = 200 = 250

From the intersection between Mohr-Coulomb line and the vertical axis where p=0 the cohesion can be determined:

6c0 cos
q= 3sin = 0 c = 0 kPa

14 Computational Geotechnics

Page 145
Advanced Computational Geotechnics

Simulation of laboratory tests

Reference oedometer and unloading-reloading stiffness

From the results of the oedometer test the oedometer stiffness as well as the unloading-reloading stiffness can
be determined. As the graph is given on logarithmic scale one cannot simply draw a tangent line as was done
for the oedometer test on sand.

Figure 13: Determination of oedometer and unloading/reloading stiffness

Considering that both primary loading and unloading/reloading paths are straight lines in the log(p)-v graph,
hence they have a relation of the form:

y = v = A log(y0 )
2 1 0.3700.270
A= log(2 )log(1 ) = log(120)log(30) =0.166

In order to determine the stiffness we calculate the derivative of the strain over the stress and change to natural
logarithm:

ln(y0 )
y = v = A ln(10)
dy dy0 ln(10)
dy0 =A 1
ln(10) 1
y0 E= dy = A y0

The E modulus found is the oedometer stiffness can be rewrittens as:

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

m
y0

ref
Eoed = Eoed pref

Hence:

ref ln(10)
Eoed = A pref and m=1

If we choose pref = 100 kPa and with the previously determined A = 0.166 we get:

Computational Geotechnics 15

Page 146
Advanced Computational Geotechnics

Simulation of laboratory tests

ref ln(10) 2.3


Eoed = A pref = 0.166 100 = 1.4 MPa.

The determination of the unloading-reloading stiffness follows the same method:

y = v = B log(y0 )
2 1 0.4270418
B= log(2 )log(1 ) = log(120)log(30) =0.0149
dy0 ln(10)
Eur = dy = B y0

However, the Eur in the Hardening Soil model is dependent on the smallest principal stress, which is x0 in an
oedeometer test and not y0 .
During the unloading process there is no linear relation between horizontal and vertical stress, as in the beginning
of unloading y0 > x0 where as after much unloading y0 < x0 . Therefore the assumption is made that during
unloading on average x0 = y0 .

0
 
ln(10) ln(10) ln(10) x
Eur = B y0 = B x0 = B pref pref

With the definition of Eur in the Hardening Soil model of

 0
m
ref x
Eur = Eur pref

ref
Follows, in a similar way as for the Eoed , that

ref ln(10) 2.3


Eur = B pref = 0.0149 100 =15 MPa and m = 1

Stiffness from triaxial test

As only undrained triaxial test data is available it is only possible to determine an undrained E50 and not an
effective E50 . Therefore the only solution is to estimate the E50 with several runs of the SoilTest program using
different input values for the reference E50 until the best fit for the undrained triaxial test data is found. Typically
for normally consolidated clays the effective reference E50 is in the range of 2-5 times the effective reference
ref
Eoed , hence this can be used as a start value for the estimation procedure. By doing so a value E50 3.5 MPa
of is found.

K0 value for normal consolidation

The K0-value for normal consolidation can only be obtained if measurements for horizontal stresses have been
performed during the oedometer test. As this is not the case here we can only use the estimation according to
Jakys rule:

K0N C 1 sin = 1 sin(250 ) =0.58

Poissons ratio

The Poissons ratio for unloading and reloading is again estimated as ur = 0.2

16 Computational Geotechnics

Page 147
Advanced Computational Geotechnics

Simulation of laboratory tests

Table 4: Summary of Hardening Soil Parameters for the clay


Parameter Unit Value
ref
E50 [kPa] 3,500
ref
Eoed [kPa] 1,400
ref
Eur [kPa] 15,000
pref [kPa] 100
ur [-] 0.2
c [kPa] 0
0 [o ] 25
[o ] 0
m [-] 1.0
K0N C [-] 0.58

Computational Geotechnics 17

Page 148
Advanced Computational Geotechnics

Simulation of laboratory tests

18 Computational Geotechnics

Page 149
Advanced Computational Geotechnics

Simulation of laboratory tests

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.

Computational Geotechnics 19

Page 150
Advanced Computational Geotechnics

Simulation of laboratory tests

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.

20 Computational Geotechnics

Page 151
Advanced Computational Geotechnics

Simulation of laboratory tests

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.

Computational Geotechnics 21

Page 152
Advanced Computational Geotechnics

Simulation of laboratory tests

Modelling a triaxial test with unloading/reloading

The standard functionality in SoilTest for simulation of a triaxial test does not allow for an intermediate unloading-
reloading path. However, the SoilTest functionality contains a General option with which soil test can be defined
in terms of boundary stresses or strains on all sides of a soil test cube. Hereafter it will be shown how this can
be used for the simulation of a triaxial test with unloading/reloading path.
After opening the SoilTest option from the material set definition window the tabsheet General should be chosen.
On this tabsheet a list of calculation phases can be defined where stress or strain increments can be applied.

Initial phase

First of all we have to specify whether stresses or strains will be applied on the boundaries during the test. For
this exercise stresses will be applied. Now the values of the initial stresses on the soil sample have to specified.
For a triaxial test the initial stresses are the cell pressures acting on the soil, hence for xx , yy and zz the cell
pressure has to entered. The cell pressure is a water pressure and so there will be no shear stress acting on the
soil: xy = 0. See figure for details.

Figure 19: General option for simulation of laboratory tests used for triaxial test

Phase 1

Apply a stress increment in vertical direction (yy ) until the stress level where the unloading path should start.
Note that the horizontal stresses (xx and zz ) remain the same as they represent the cell pressure. Hence,
the horizontal stress increments are zero in this phase.

Phase 2

Press the Add button to add another phase to the phase list. This phase represents the unloading phase. See
figure for details.

Phase 3

Press the Add button once more in order to add the 3rd phase. This phase represents the reloading of the soil
as well as the continuation of primary loading until either failure or a higher stress level from where for instance

22 Computational Geotechnics

Page 153
Advanced Computational Geotechnics

Simulation of laboratory tests

another unloading/reloading cycle is going to be made.

Figure 20: Unloading/reloading cycle in a triaxial test using the General option

Computational Geotechnics 23

Page 154
Advanced Computational Geotechnics

DERIVATION OF SOIL PARAMETERS FROM LAB TEST RESULTS &


VERIFICATION IN PLAXIS SOILAB TEST

Dr William Cheang Wai Lum

Slides: Dr Shen Rui Fu (NUS)

EXERCISE 2

SAND CLAY
SECTION 1 SECTION 3
1. Typical test results
1. Typical Test results
SECTION 2 SECTION 4
1. Using Soil Lab Test
Using Soil Lab Test

Page 155
Advanced Computational Geotechnics

SAND: OEDOMETER AND TRI-AXIAL TESTING

SECTION 1

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.

Page 156
Advanced Computational Geotechnics

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

For sand, one of the most common lab tests is


Triaxial Isotropically Consolidated Drained (CID)Test

A Triaxial setup in NUS Geotechnical Lab Page 157


For sand, one of the most common lab tests is Advanced Computational Geotechnics

Triaxial Isotropically Consolidated Drained (CID)Test

Fa/A = q (deviatoric
stress)
Typical sample size 38 mm x 76 mm a = q + r

450

400

350
Deviatorstress(kPa)

300

250

200 3 = 100 kPa


150

100 Testdata

50

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07

Axialstrain

Test data: Deviator stress ~ axial strain


curve (Triaxial) Page 158
Advanced Computational Geotechnics
0.06

0.05

0.04
Volumetricstrain

0.03

0.02

0.01
Testdata

0
0 0.02 0.04 0.06 0.08 0.1

0.01

Axialstrain

Test data: Volumetric strain ~ axial


strain curve (Triaxial)

Another common lab test is Oedometer Test

Oedometer setups in NUS


Geotechnical Lab Page 159
Another common lab test is Oedometer Test
Advanced Computational Geotechnics

Settlementdialgauge

OedometerCell
Sample:dia.=75mm
Protruded
leverarm Height=20mm

Heavydeadweights

Another common lab test is Oedometer Test

Typical sample size 75 mm x 20 mm

Boundary
conditions Page 160
Advanced Computational Geotechnics
0
0.1
0.2
0.3
0.4
Verticalstrain(%)

0.5
0.6
0.7 Testdata
0.8
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 400
Verticalpressure(kPa)

Vertical stress ~ vertical strain curve


(Oedometer)

400

300
Verticalpressure(kPa)

200

100

Testdata

0
0 50 100 150 200

Lateralstress(kPa)

Vertical stress ~ lateral stress curve Page 161


(Oedometer)
Advanced Computational Geotechnics

SOIL LAB TEST IN PLAXIS (CALIBRATION)

SECTION 2

Hardening Soil Parameters to be derived based on the


above typical lab test data

Page 162
Part 1: Strength parameters
Advanced Computational Geotechnics

C=0 for sand

Part 1: Strength parameters

450

400
Since c = 0 for sand, it can be
350
simplified to:
Deviatorstress(kPa)

300

250

200 3 = 100 kPa


150

100 Testdata

50

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
500 100
sin ' 0.67
Axialstrain
500 100

' 42

Page 163
Part 1: Strength parameters
Advanced Computational Geotechnics

0.06

0.05
So,
0.048
0.04
Volumetricstrain

1 sin 0.09 0.03


0.03
1-sin 1.36
2 sin 0.048 0.004
0.02
2sin
0.01

0.004
Testdata sin 0.27
0
0 0.02 0.04 0.06 0.08 0.1
0.03 0.09
16
0.01

Axialstrain

BTW, why there is an initial contraction before the


soil sample to dilate prominently ??

Contraction then Dilation

What contributes to the sample contraction? dp '


(1) dp >0 elastic volumetric contraction! e
v
K
(2) Isotropic hardening plastic volumetric contraction! 1 m
pc
p ,cap

1 m p ref
v

What contributes to the sample dilation?


(1) As the stress path cut through series of shear yield line, plastic shear strain
was generated. d p
(2) the plastic shear strain will be accompanied by plastic volumetric strain by
, and it is dilative! d vp , fric d p , fric sin m
q
MC line

Page 164

pc p
Part 2: Stiffness parameters
Advanced Computational Geotechnics

Part 2: Stiffness parameters

450

400 400

350
Deviatorstress(kPa)

300

250
3 = 100 kPa
200

150

100 Testdata

50

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
0.013
Axialstrain

400
E50ref 30800 kPa 30000 kPa
0.013
Page 165
Part 2: Stiffness parameters
Advanced Computational Geotechnics

450

400400

350

Deviatorstress(kPa)
300

3 = 100 kPa
250

200

150

100 Testdata

50

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
0.021 0.026
Axialstrain

400
Eurref 80000 kPa
0.026 0.021

As sand unload-reloading stiffness Eurref is generally


about 3~5 times of E50ref, we may set Eurref = 90000kPa

Part 2: Stiffness parameters

Page 166
Part 2: Stiffness parameters
Advanced Computational Geotechnics

0
0.1
0.2
0.3
0.33
0.4
Verticalstrain(%)

0.5
0.6
0.7 Testdata
0.8
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 320 400
Verticalpressure(kPa)

320
ref
Eoed 29900kPa 30000kPa
1.4% 0.33%

Part 2: Stiffness parameters


m
c cos ' '3 sin '
E50 E50ref
c cos ' p sin '
ref

m
ref c cot ' '1
Eoed Eoed
c cot ' p ref

Page 167
Part 2: Stiffness parameters
Advanced Computational Geotechnics

0
0.1
320
29900kPa 30000kPa
0.2 ref
Eoed
0.3
0.4
1.4% 0.33%
0.47
Verticalstrain(%)

0.5
400
0.6 200 kPa
Eoed 43000kPa
0.7
0.8
Testdata
1.4% 0.47%
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 400
Verticalpressure(kPa) 400

m
c cot ' '1
m
200
200 kPa
Eoed 43000
m = 0.5
c cot ' p 100
ref ref
Eoed 30000

Part 3: Other parameters

Jakys formula:

K 0NC 1 sin ' 1 sin 42 0.33


400

300
Verticalpressure(kPa)

200

100

Testdata

0
0 50 100 150 200

Lateralstress(kPa)

x ' 100
K 0NC 0.33
y ' 300 Page 168
Summary of Hardening Soil Parameters
Advanced Computational Geotechnics

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


Page 169
Advanced Computational Geotechnics

CLAYS: OEDOMETER AND TRIAXIAL TESTS

SECTION 3

For Clay, one of the most common lab tests is Triaxial


Isotropically Consolidated UnDrained (CIU) Test

A Triaxial setup in NUS


Geotechnical Lab Page 170
For Clay, one of the most common lab tests is Triaxial
Advanced Computational Geotechnics

Isotropically Consolidated UnDrained (CIU)Test

Fa/A = q (deviatoric
stress)
Close the valve = Undrained test =
a = q + r
Excess will accumulate with shearing

350

Testdata
300

250
q(kPa)

200
195

150

100

50

0
0 50 100 150 200 250 300 350 400 450
p'(kPa)

Test data: stress path p~q Page 171


CIU stress path
Advanced Computational Geotechnics

Gradient:
350

Testdata
300

250 6 sin ' 195



3 sin ' 200
q(kPa)

200
195

150

100 = 25
50

0 Intercept:
0 50 100 150 200 250 300 350 400 450
p'(kPa)

6c' cos '


0
3 sin '

c = 0

Another common lab test is Oedometer Test

Oedometer setups in NUS


Geotechnical Lab Page 172
Another common lab test is Oedometer Test
Advanced Computational Geotechnics

Typically less test points are available


due to long consolidation period for
each loading stage
Boundary
conditions

Testdata
0.1
Verticalstrain(%)

0.2

0.3

0.4

0.5
1 10 100 1000
Verticalpressure(kPa)

Typically oedometer test results are presented in SI


report as logv ~ yy which is linear (unlike sand) which Page 173
must be dealt with cautions!
Oedometer test for clay
Advanced Computational Geotechnics

d y '
Testdata

Eoed
0.1

d y
Verticalstrain(%)

0.2

d (log y ' )
Gradient _ k
d y
0.3

Obviously, Eoed Gradient _ k


0.4

0.5
1 10 100 1000
Verticalpressure(kPa)

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

Oedometer test for clay


0

Testdata
0.1
ref
Eoed 2.3 100 6.02 1350 kPa
Verticalstrain(%)

0.2
Eoed y '
0.27 ref

0.3
Eoed pref

0.37 m
0.4
ref c cot ' '1

Eoed Eoed c cot ' p ref
0.5

1 10 30 100 1000
120 m
Eoed '1
Verticalpressure(kPa)

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


Page 174
Oedometer test for clay
Advanced Computational Geotechnics

0
Eur refers to when 3 = 100kPa
Testdata
0.1 During oedometer loading, when y
=100kPa, x<100kPa;
Verticalstrain(%)

0.2

When y loaded to about 300kPa and


0.3
unload to 100kPa, x is expected to be
0.418
0.4 closer to 100kPa. As such, we can
0.427
approximately accept the derived Eur.
0.5
1 10 30 100 1000
Verticalpressure(kPa) 120

log(120) log(30)
gradient _ k 66.9 Eur 2.3 100 66.9 15000 kPa
0.427 0.418
Jakys formula:
Eur 2.3 y '66.9
K 0NC 1 sin ' 1 sin 25 .58
Poissons ratio ur = 0.2

80
3 = 100kPa for
consolidation,
During shearing, 3 = 0
70

60
Excess pore pressure
Deviatorstress(kPa)

50 accumulates during
40 shearing 3 100kPa
3 = 100 kPa
30
Typically for NC clay, E50ref
20
may be about
Testdata 2~5 times
10 Eoedref or about 2800kPa~7000kPa.
Trial runs to fit the test data gives
0
0 0.01 0.02 0.03 0.04 E50ref =0.06
0.05 3500kPa
0.07

Axialstrain

Can we use the CIU test 1 ~ q test data to


derive the E50ref ?? Page 175
Advanced Computational Geotechnics

SOIL LAB TEST: FOR CLAY SAMPLE

SECTION 4

Summary of Hardening Soil Parameters

Page 176
FEM simulation using Plaxis SoilTest Facility
Advanced Computational Geotechnics

END

Page 177
Advanced Computational Geotechnics

H
HKCG 7

MOD
DELLIN
NG OF DEEP
P EXCA
AVAT
TIONS

Professo
P or Anton
nio Gens
s

Page 178
Advanced Computational Geotechnics

CG7: MODELLING OF DEEP EXCAVATIONS

Antonio Gens

Technical University of Catalunya, Barcelona

some of the slides were originally created by:


Helmut Schweiger (Technical University of Graz, Austria)
Dennis Waterman (Plaxis)

Outline
Introduction
Fundamentals of excavation modelling
Excavation
Hydraulic conditions
Thin wall vs. thick wall
Wall support
Interfaces

Simplified example
A note on factors of safety

Page 179
Advanced Computational Geotechnics

INTRODUCTION

Analysis of a deep excavation involves modelling of

various excavation stages


interaction wall / soil > interface elements
changes in groundwater level
struts or anchors (including load transfer to soil)
adjacent structures (buildings, tunnels, piles, )

Requires advanced constitutive model because

stress paths in soil are not monotonic (significant change in stress path
direction)
primary loading and unloading / reloading occurs in different parts of the domain
analysed
some areas will experience large strains with significant plastic deformations,
others will be in the very small strain range
> simple elastic - perfectly plastic models not suffcient

2D model
sufficient

Page 180
Advanced Computational Geotechnics

3D model required

Excavations: general
Typical excavation sequence:
I 1st excavation
II Pre-stressing anchor
III Final excavation

Page 181
Advanced Computational Geotechnics

Excavations: general
I 1st excavation

The domain excavated is removed from the mesh


In the nodes on the excavated boundary, out-of-balance forces appear:
they are exactly equivalent to the effect of excavation

Excavations: general
II Pre-stressing anchor

The anchor element is installed (created)


The pre-stress load is applied

Page 182
Advanced Computational Geotechnics

Excavations: general
III Final excavation

The excavated domain is removed and the corresponding out of balance


forces appear
Attention to hydraulic conditions!

Excavations: hydraulic conditions

Wet excavation
Simply click on the cluster or clusters to be excavated
The water remains in the excavated domain, water pressures automatically
appear
Hydraulic conditions do not change

Page 183
Advanced Computational Geotechnics

Excavations: hydraulic conditions

Dry excavation
Click on the cluster or clusters to be excavated
Use cluster dry option or cluster phreatic line
Set new hydraulic conditions

Excavations: hydraulic conditions


Setting of new hydraulic conditions
Consider each individual case carefully
Select the relevant situation: undrained, drained or consolidation
it may be different for different materials

Specifying pore pressure distributions


General phreatic level
Applies to all clusters that have not been separately defined
Cluster phreatic level
Applies to one specific cluster
Cluster dry
Makes a specific cluster dry
Interpolate
Interpolates pore pressures between clusters above and below
User-defined pore pressure
Specify pore pressure at reference level and the rate of increase in the y direction
Groundwater flow calculation
Gives the steady-state solution

Page 184
Advanced Computational Geotechnics

Excavations: hydraulic conditions


Groundwater flow calculation: example

Equal pore pressure values on both


sides at the foot of the wall!

Groundwater flow calculations


Plate elements are permeable
Interface elements are impermeable (unless
deactivated in the groundwater flow calculation)

Excavations: hydraulic conditions


Dewatering
1 2 3

General
Z-shape phreatic level gives wrong results:
General

6
General
7
No equilibrium in horizontal water pressures:
Local peak stresses
5
Local peak strains
Non-physical horizontal displacements
8

0 4
Non-physical excess pore pressures

Possible incorrect water pressure acting on


wall

Page 185
Advanced Computational Geotechnics

Walls thin wall vs. thick wall

Thin wall
Wall thickness << wall length
Shearing not important
No end-bearing, only friction
Plate element suffices

Thick wall
Wall thickness significant
Shearing important
End-bearing capacity needed
Use soil elements with material set representing wall material
In order to obtain structural forces a plate may be inserted

Walls thick wall


Soil elements with material set representing wall material
Difficult to obtain structural forces from soil elements, therefore introduce plate:
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
d

Soil elements: Esoil=Ewall, I = 1/12*d3 , d = wall thickness

Plate element: EI = EsoilI / x, choose x large (e.g. 10000)

uplate = usoil Mwall = x*Mplate, Qwall = x*Qplate

NOTE: Nwall = Nplate+y,0 , y,0 = initial vertical stress in soil elements

Page 186
Advanced Computational Geotechnics

Walls support
Lagging
Use short additional section of plate
perpendicular to the wall
Create short section with increased
stiffness using multiple chains
Struts
Full excavation: node-to-node anchor
Half (symmetric) excavation: fixed-end
anchor
Anchors
Grout anchor: node-to-node anchor +
geotextile for grout body
Ground anchor: node-to-node anchor
+ perpendicular plate element

Ground anchors
Axial forces in ground anchors:

Input geometry

real 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

Page 187
Advanced Computational Geotechnics

Interfaces
Soil-structure interaction
Wall friction
Slip and gapping between soil and structure
Interface material properties
Taken from soil using reduction factor Rinter
Cinter = Rinter * Csoil
tan(inter) = Rinter * tan(soil)
inter = 0 for Rinter < 1, else inter = soil
Ginter = (Rinter)2 * Gsoil
Some building codes prescribe soil-wall friction angle :
Rinter = tan()/ tan(soil)
Individual material set for interface

Interfaces

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 = Rinter 1.0
(interface may not be required)
Interaction soil/geotextile = Rinter 0.9 0.5 (foil, textile)

Stability should not be critical on Rinter !

Page 188
Advanced Computational Geotechnics

SIMPLIFIED EXAMPLE

Goal of study
assess influence of relative stiffness between wall and soil on results
assess influence of wall length
show influence of constitutive model
>show trends and qualitative behaviour rather than
quantitative comparison
Constitutive models
Mohr-Coulomb
Hardening Soil
HS_small
Soft Soil (excavation in clay only)
Modified Cam Clay (excavation in clay only)

Results
wall deflection
bending moments / strut forces
earth pressure distribution (active - passive)
vertical displacements behind wall

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

MODELS COMPARED
Mohr-Coulomb Model

q 1 3

perfectly plastic

elastic

Note: In this study stiffness for variations of MC analyses is assumed as:


MC_a > EMC = (E50 + Eur)HS at level of base of wall
MC_b > EMC = Eur, HS at level of base of wall
MC_c > EMC = E50, HS at level of base of wall
MC = 0.3

Page 189
Advanced Computational Geotechnics

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

MODELS COMPARED
Hardening Soil Model

q
plastic

elastic

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

MODELS COMPARED
Hardening Soil Small Model
additional input parameters for HS-small
G0 ref
40000 40000
TangentmodulusG[kN/m]
SecantmodulusG[kN/m]

30000 0.7 30000

20000 20000

Gurref
10000 10000
HSSmall
Hardin&Drnevich
0 0
1E5 0.0001 0.001 0.01 1E5 0.0001 0.001 0.01
Shearstrain[] Shearstrain[]

Thomas Benz, Small-Strain Stiffness of Soils and its Numerical Consequences, Mitteilung
55 des Instituts fr Geotechnik, Universitt Stuttgart, 2007.

Page 190
Advanced Computational Geotechnics

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

MODELS COMPARED
Soft Soil Model

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

PROBLEM SPECIFICATION

Wall length:
Dense sand and marl: 9 m
Medium dense sand: 10 m
Clay: 11 m

Page 191
Advanced Computational Geotechnics

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

DENSE SAND
Parameters for Hardening Soil model (HS)
Parameter Meaning Value
[kN/m] Unit weight (unsaturated) 18
sat [kN/m] Unit weight (saturated) 20
[] Friction angle (Mohr-Coulomb) 41
c [kPa] Cohesion (Mohr-Coulomb) 0
[] Angle of dilatancy 15
ur [-] Poissons ratio unloading-reloading 0.20
E50ref [kPa] Secant modulus for primary triaxial loading 30 000
Eoedref [kPa] Tangent modulus for oedometric loading 30 000
Eurref [kPa] Secant modulus for un- and reloading 90 000
m [-] Exponent of the Ohde/Janbu law 0.55
pref [kPa] Reference stress for the stiffness parameters 100
K0nc [-] Coefficient of earth pressure at rest (NC) 1-sin()
Tension [kPa] Tensile strength 0

Additional parameters for Hardening Soil Small model (HSS)


Parameter Meaning Value
G0 [kPa] Small-strain Shear modulus 112 500
0,7 [-] Reference shear strain where Gsec=0.7G0 0.0002

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

DENSE SAND

horizontal wall displacement [mm] bending moments [kNm/m]


15 12 9 6 3 0 -3 -6 -80 -60 -40 -20 0 20 40
0 0
HS HS
HSS HSS
MC_a 1 MC_a 1
MC_b MC_b
MC_c MC_c

2 2
depth below surface [m]

depth below surface [m]

3 3

4 4

5 5

6 6

7 7

8 8

9 9

Page 192
Advanced Computational Geotechnics

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

DENSE SAND 2
earth pressure [kN/m ]
-120 -100 -80 -60 -40 -20 0 20 40
0
HS
HSS
MC_a 1
distance from wall [m]
0 4 8 12 16 20 24
14
surface displacement [mm]

2
12
10 HS
HSS
8

depth below surface [m]


MC_a 3
6 MC_b
4 MC_c
2 4
0
-2
-4 5
-6
-8
6

Strut forces: 7
HS: -102 kN/m
HSS: -107 kN/m 8
MC_a: -78 kN/m
MC_b: -83 kN/m
9
MC_c: -72 kN/m

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

LOOSE TO MEDIUM DENSE SAND


Parameters for Hardening Soil model (HS)
Parameter Meaning Value
[kN/m] Unit weight (unsaturated) 18
r [kN/m] Unit weight (saturated) 20
[] Friction angle (Mohr-Coulomb) 34
c [kPa] Cohesion (Mohr-Coulomb) 0
[] Angle of dilatancy 0
ur [-] Poissons ratio unloading-reloading 0.2
E50ref [kPa] Secant modulus for primary triaxial loading 12 000
Eoedref [kPa] Tangent modulus for oedometric loading 16 000
Eurref [kPa] Secant modulus for un- and reloading 60 000
m [-] Exponent of the Ohde/Janbu law 0.75
pref [kPa] Reference stress for the stiffness parameters 100
K0nc [-] Coefficient of earth pressure at rest (NC) 1-sin()
Rf [-] Failure ratio 0.90
Tension [kPa] Tensile strength 0

Additional parameters for Hardening Soil Small model (HSS)


Parameter Meaning Value
G0 [kPa] Small-strain Shear modulus 87 500
0,7 [-] Reference shear strain where Gsec=0.7G0 0.0002

Page 193
Advanced Computational Geotechnics

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

LOOSE TO MEDIUM DENSE SAND

horizontal wall displacement [mm] bending moments [kNm/m]


35 30 25 20 15 10 5 0 -5 -10 -140 -120 -100 -80 -60 -40 -20 0 20 40
0 0
HS HS
HSS HSS
MC_a 1 MC_a 1
MC_b MC_b
MC_c MC_c
2 2

3 3

depth below surface [m]


depth below surface [m]
4 4

5 5

6 6

7 7

8 8

9 9

10 10

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

LOOSE TO MEDIUM DENSE SAND

distance from wall [m]


0 5 10 15 20 25
25
surface displacement [mm]

20
15
10
5
0
-5
HS
-10 HSS
-15 MC_a
-20 MC_b
-25 MC_c
-30

Strut forces:
HS: -151 kN/m
HSS: -154 kN/m
MC_a: -119 kN/m
MC_b: -127 kN/m
MC_c: -105 kN/m

Page 194
Advanced Computational Geotechnics

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

LOOSE TO MEDIUM DENSE SAND

Ratio G/Gur at final stage

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

CLAY
Parameters for Hardening Soil model (HS)
Parameter Meaning Value
[kN/m] Unit weight (unsaturated) 15
r [kN/m] Unit weight (saturated) 16
[] Friction angle (Mohr-Coulomb) 27
c [kPa] Cohesion (Mohr-Coulomb) 15
[] Angle of dilatancy 0
ur [-] Poissons ratio unloading-reloading 0.20
E50ref [kPa] Secant modulus for primary triaxial loading 4 300
Eoedref [kPa] Tangent modulus for oedometric loading 1 800
Eurref [kPa] Secant modulus for un- and reloading 14 400
m [-] Exponent of the Ohde/Janbu law 0.90
pref [kPa] Reference stress for the stiffness parameters 100
K0nc [-] Coefficient of earth pressure at rest (NC) 1-sin()
Rf [-] Failure ratio 0.90
Tension [kPa] Tensile strength 0

Additional parameters for Hardening Soil Small model (HSS)


Parameter Meaning Value
G0 [kPa] Small-strain Shear modulus 25 000
0,7 [-] Reference shear strain where Gsec=0.7G0 0.0003

Page 195
Advanced Computational Geotechnics

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

CLAY
Parameters for Soft Soil model (SS)
Parameter Meaning Value
[kN/m] Unit weight (unsaturated) 15
r [kN/m] Unit weight (saturated) 16
[] Friction angle (Mohr-Coulomb) 27
c [kPa] Cohesion (Mohr-Coulomb) 15
[] Angle of dilatancy 0
ur [-] Poissons ratio 0.20
* [-] Modified swelling index 0.0125
* [-] Modified compression index 0.0556
K0nc [-] Coefficient of earth pressure at rest (NC) 1-sin()
Tension [kPa] Tensile strength 0

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

CLAY

horizontal wall displacement [mm] bending moments [kNm/m]


120 110 100 90 80 70 60 50 40 30 20 10 0 -10 -150 -120 -90 -60 -30 0 30
0 0
HS HS
HSS HSS
1 1
MC_a MC_a
MC_b MC_b
MC_c MC_c
2 2
SS SS

3 3
depth below surface [m]

depth below surface [m]

4 4

5 5

6 6

7 7

8 8

9 9

10 10

11 11

Page 196
Advanced Computational Geotechnics

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

CLAY

distance from wall [m]


0 10 20 30 40 50 60 70
120
surface displacement [mm]

100 HS
80 HSS
MC_a
60 MC_b
40 MC_c
SS
20
0
-20
-40
-60

Strut forces:
HS: -115 kN/m
HSS: -120 kN/m
MC_a: -96 kN/m
MC_b: -93 kN/m
MC_c: -103 kN/m
SS -75 kN/m

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

MARL
Parameters for Hardening Soil model (HS)
Parameter Meaning Value
[kN/m] Unit weight (unsaturated) 17
r [kN/m] Unit weight (saturated) 20
[] Friction angle (Mohr-Coulomb) 30
c [kPa] Cohesion (Mohr-Coulomb) 47
[] Angle of dilatancy 10
ur [-] Poissons ratio unloading-reloading 0.20
E50ref [kPa] Secant modulus for primary triaxial loading 30 000
Eoedref [kPa] Tangent modulus for oedometric loading 15 000
Eurref [kPa] Secant modulus for un- and reloading 100 000
m [-] Exponent of the Ohde/Janbu law 0.90
pref [kPa] Reference stress for the stiffness parameters 100
K0nc [-] Coefficient of earth pressure at rest (NC) 1-sin()
Rf [-] Failure ratio 0.90
Tension [kPa] Tensile strength 0

highly overconsolidated > POP = 2 000 kPa, K0 = 2.0


Additional parameters for Hardening Soil Small model (HSS)
Parameter Meaning Value
G0 [kPa] Small-strain Shear modulus 150 000
0,7 [-] Reference shear strain where Gsec=0.7G0 0.0001

Page 197
Advanced Computational Geotechnics

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

MARL

Parameters for Soft Soil model (SS)


Parameter Meaning Value
[kN/m] Unit weight (unsaturated) 17
r [kN/m] Unit weight (saturated) 20
[] Friction angle (Mohr-Coulomb) 30
c [kPa] Cohesion (Mohr-Coulomb) 47
[] Angle of dilatancy 10
ur [-] Poissons ratio 0.20
* [-] Modified swelling index 0.0018
* [-] Modified compression index 0.0067
K0nc [-] Coefficient of earth pressure at rest (NC) 1-sin()
Tension [kPa] Tensile strength 0

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

MARL

horizontal wall displacement [mm] bending moments [kNm/m]


12 10 8 6 4 2 0 -2 -80 -60 -40 -20 0 20 40 60 80 100
0 0
HS HS
HSS HSS
MC_a 1 MC_a 1
MC_b MC_b
MC_c MC_c
SS SS
2 2
depth below surface [m]

depth below surface [m]

3 3

4 4

5 5

6 6

7 7

8 8

9 9

Page 198
Advanced Computational Geotechnics

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

MARL

distance from wall [m]


0 10 20 30 40 50 60 70
12
surface displacement [mm]

10 HS
HSS
8 MC_a
MC_b
6
MC_c
4 SS

-2

-4

Strut forces:
HS: -254 kN/m
HSS: -208 kN/m
MC_a: -212 kN/m
MC_b: -192 kN/m
MC_c: -239 kN/m
SS -195 kN/m

Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions

MARL

Ratio G/Gur at final stage

Page 199
Advanced Computational Geotechnics

Material behaviour

Mohr-Coulomb: unrealistic deformations


Overestimation over bottom heave
Often heave of soil behind the wall

Hardening Soil model: qualitative realistic deformations


Better bottom heave, but increases with model depth
Settlement trough behind wall, but often too shallow and too wide

HS-small model: qualitative and quantitative realistic deformations


Good bottom heave independent of model depth
More realistic settlement trough behind the wall (narrower and
deeper)

NOTE ON / c - REDUCTION
Material parameters soil layer:
sheet pile
type AZ-18 = 35

1.0 m c = 0.1 kN/m


1.5 m
= 17 kN/m
= 0
6.5 m Properties sheet pile wall:
EA = 3.008E6 kN/m
2
EI = 6.84E4 kNm /m

4.0 m
Mpl = 505 kNm/m
Properties strut:
2
E = 3.0E7 kN/m
A = 0.24 m2
Horizontal strut distance: 1 m

Page 200
Advanced Computational Geotechnics

NOTE ON / c - REDUCTION

wall elastic
Msf = 1.95

NOTE ON / c - REDUCTION

wall elastic - perfectly plastic


Msf = 1.73

Page 201
Advanced Computational Geotechnics

H
HKCG 8

STR
RUCTU
URAL E
ELEMENTS IN PL
LAXIS

Dr Jo
ohnny Cheuk
C

Page 202
Advanced Computational Geotechnics

STRUCTURAL ELEMENTS & IN PLAXIS 2D

Dr Johnny Cheuk
Aecom Asia

Some course notes:


Dr William Cheang, Plaxis AsiaPac
Dr Ronald Brinkgreve, Plaxis B.V.
Dr Shen Rui Fu, NUS

Contents
1. StructuralelementsavailableinPlaxis
2. UsageofstructuralelementsinFEmodelling
3. Plate elements(BeamandShellelement)
4. Anchor elements(Springelement)
5. Geotextile elements(Membraneelement)
6. Interface elements(Zerothicknesselement)

Page 203
2
1.Structural elements in Computational
Advanced PlaxisGeotechnics

1. Plateelement Section 3.42 & 14.5

2. Anchorelement Section 3.45, 3.46 & 14.1

3. Geogridselement Section 3.43 & 14.3

4. Interfaceelement
Section 3.44 & 14.1

2. Application of structural elements

wall strip footing tunnel

geotextile wall ground anchor cofferdam

strut anchored wall Page 204


4
Advanced Computational Geotechnics
3.1 Plate Element

Overview:
1. 3or5nodedlineelements(for6nodedor15nodedelementmesh)
2. 3degreesoffreedompernode
3. Plateshave:
o Axialforces
o Shearforces
o Bendingmoments
o Hoopforces(axisymmetry)
4. Elasticorelastoplasticbehaviour
5. Formodellingwalls,floors,tunnels

3.2 Plate Element

Plates elasticparameters
h3 b
EI E (b = 1 m)
12
EA E h b (b = 1 m)

EI (Equivalent rectangular
d h 12
EA plate thickness)

h h
b

b = 1 m in plane strain
b = 1 meter in axisymmetry
b Page 205
6
Advanced Computational Geotechnics
3.3 Plate Element
Plates elastoplasticbehaviour

Np

M
Mp

3.4 Plate Element


(Illustration:MpNp.P2D):

1200 10090 80 70 60 50 40 30 20 10 0 10 20 30 40 50 60 70 80 90 100


Envelope 0
Elastoplastic plate
1000
Elasticplate
5
800
Elastoplastic plate

10 Elasticplate
600
N

400 15

200 20

0 Page 206
25 8
200 0 200 400 600 800 1000 1200 M
M
3.5 Plate Element Advanced Computational Geotechnics

EffectonGlobalFOSbyc/phiReduction

CBPElastic,Failure CBP Elasto-Plastic


withnoPlastic Failure with Plastic
Hinge, Hinge, FOS=1.40
FOS=1.75

1. Elastic wall excludes possibility of wall plastic hinge; and over-estimate FOS=1.75
2. Allowing for wall plastic hinge (Elasto-plastic wall) gave lower FOS=1.40 and smaller soil yielded
zone behind the wall
9

3.6 Plate Element

Plates weight,insoil

Actual problem In the model

dreal

wreal = concrete d real wmodel = soil d real wplate Below GT


soil sat
wmodel = wreal w plate = ( concrete - soil ) d real Above GT
Page unsat
soil 10
207
Advanced Computational Geotechnics
3.7 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 wplate = ( concrete soil ) d real Above GT
2
soil 11 unsat

3.8 Plate Element


Plates connections
Spring data:
Stiffness
6 8 Min/Max moment

Rotation
spring

5 7

Hinged connection

Rigid connection
(default)

Page 208
12
Illustration: Connection.P2D
Advanced Computational Geotechnics
3.9 Plate Element
Walls thinwallvs.thickwall
Thinwall
Wallthickness<<walllength
Nomuchendbearing,onlyfriction
Plateelementsuffices

Thickwall
Wallthicknesssignificant
Endbearingcapacityneeded
Usesoilelementswithmaterialsetrepresen ngwallmaterial
Inordertoobtainstructuralforcesaplatewithfictitiousproperties
maybeinserted

13

3.10 Plate Element


Walls thick wall
1. Soilelementswithmaterialsetrepresentingwallmaterial
2. Difficulttoobtainstructuralforcesfromsoilelements,
thereforeintroduceveryflexibleplatewithinthesolidwall
elements:
Noinfluenceondeformation:lowstiffness,noweight
Locatedinontheneutralline(usuallythemiddle)
Tightbondingtotheconcreteelements:nointerfaces

(Illustration: Beam.P2D): d

Solid elements: Esoil=Ewall, I = 1/12*d3 , d = wall thickness

Plate element: EI = EsoilI / x, choose x large (e.g. 106)

uplate = usoil Mwall = x*Mplate, Qwall = x*Qplate

Page 209
14
2. Advanced
Anchor Element
Computational Geotechnics

Anchors fixedend
a) Tomodelsupports,anchorsandstruts
a) Elastoplasticspringelement
b) Oneendfixedtopointinthegeometry,otherendisfully
fixedfordisplacement
c) Positioningatanyangle
d) Prestressingoption

Anchors nodetonode
a) Tomodelanchors,columns,strutsandrods
a) Elastoplasticspringelement
b) Connectstwogeometrypointsinthegeometry
c) Nointeractionwiththemeshalongtheanchorrod
d) Prestressingoption

15

4.1Anchor Element
Anchors materialproperties
Axialstiffness,EA (foroneanchor) [kN]
Spacing,Ls (outofplanedistancebetweenanchors) [m]
Maximumanchorforceforcompressionandtension,
|Fmax,comp|and|Fmax,tens| [kN]

Ls

Page 210
16
4.2 Anchor ElementAdvanced Computational Geotechnics

Anchors prestressing
DefinedinStagedconstructionphase
Bothtension(groutanchor)orcompression(strut)
possible

Tension = positive

17

5.1.Geogrid Element

Geogrids
1. 3or5nodedlineelement
2. Elasticorelastoplasticbehaviour
3. Noflexuralrigidity(EI),onlyaxialstiffness(EA)
4. Onlyallowsfortension,notforcompression

Page 211
18
Advanced Computational Geotechnics
5.2 Anchor Element + Geogrid Element
Groundanchors

1. Combinationofnodetonodeanchorandgeogrid
2. Nodetonodeanchorrepresentsanchorrod(freelength)
(nointeractionwithsurroundingsoil)
3. Geogridrepresentsgroutedpart(fullinteractionwithsurroundingsoil)
4. Nointerfacearoundgroutedpart;interfacewouldcreateunrealisticfailuresurface
5. Workingloadconditionsonly nopullout
6. Ifpulloutforceisknownthiscanbeusedbylimitinganchorrodforce

19

5.3Groundanchors
Axial force distribution along fixed length (modelled using geogrid)

Nrod <>Ngrout duetosharednode


betweenanchor,geotextile
andsoil

Probableactualdistributionofaxial
forcesingroundanchor
axialforcesingeotextileelement

Input geometry Generated mesh Page 212


20
5.4GroundAnchors:Influenceofnodenumbers
Advanced Computational Geotechnics

alongstructuralelements

21

6.1InterfaceElement
Interfaces materialproperties
1. Soilstructureinteraction
1. Wallfriction
2. Slipandgappingbetweensoilandstructure
2. Soilmaterialproperties
A. TakenfromsoilusingreductionfactorRinter
3. Individualmaterialsetforinterfacepossible

Page 213
22
6.2 Interface ElementAdvanced Computational Geotechnics
Interfaces reductionfactor
SuggestionsforRinter:
Interactionsand/steel =Rinter 0.6 0.7
Interactionclay/steel =Rinter 0.5
Interactionsand/concrete =Rinter 1.0 0.8
Interactionclay/concrete =Rinter 1.0 0.7
Interactionsoil/geogrid(groutedbody) =Rinter1.0
(interfacemaynotberequired)
Interactionsoil/geotextile =Rinter0.9 0.5(foil,textile)

With reference to BS8002:

23

References

1. Brinkgreve, R., Engin, E, & Swolf, W. (2010), Plaxis 2d 2010

Page 214
24
Advanced Computational Geotechnics

25

Page 215
Advanced Computational Geotechnics

H
HKCG 9

EXE
ERCIS
SE 3
EX ATION OF A BUILD
XCAVA DING P IT
ee Siew
Dr Le w Wei

Page 216
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

EXCAVATIONOFBUILDINGPIT(MCModel)

FranzTschuchnigg

COMPUTIONALGEOTECHNICS

CourseComputionalGeotechnics 1
Page 217
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

CourseComputionalGeotechnics 2
Page 218
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

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 60 cm diameter bored piles; the wall is anchored by two rows of pre
stressedgroundanchors.Inthisexercisetheconstructionofthisbuildingpitissimulatedandthedeformation
andbendingmomentsofthewallareevaluated.
Theupper40mofthesubsoilconsistsofamoreorlesshomogeneouslayerofmediumdensefinesandwitha
unitweightof18kN/m.TriaxialtestdataofarepresentativesoilsampleisgiveninFigure2.Underneaththis
layerthereisverystifflayerofgravel,whichisnottobeincludedinthemodel.Thegroundwatertableisvery
deepanddoesnotplayaroleinthisanalysis.

AIMS
Determinationofsoilstiffnessparameters
Usinginterfaceelements
Usinggroundanchors
Prestressingofanchors
Combinationofstructuralelements

Stage 1 Diaphragm wall

Stage 2 Anchor rod top sand

Stage 3

Grout body middle sand

lower sand

deep sand 1

deep sand 2


Figure1:Geometryfortiedbackexcavation.

CourseComputionalGeotechnics 3
Page 219
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

DETERMINATIONOFSTIFFNESS&STRENGTHPROPERTIES(SAND)
Use the MohrCoulomb model and extract model parameters for the sand layer from the triaxial test data.
Concerningstiffnessbehaviour,takeintoaccountthefactthattheexcavationinvolvesunloadingratherthan
primaryloading.Itissuggestedtodividethesubsoilintothreeregions:TopSand(0m15m),MiddleSand
(15m 25m) and Lower Sand (25m 35m). This allows for the input of an average stiffness in each
individualregion.AsthesimpleMohrCoulombmodelcannottakeintoaccountthestressdependencyofthe
stiffness,theinputofanaveragestiffnessperregion,bytheuser,isthenextbestoptiontoenhancethemodel.
Please note there are other models (HS and SS model) available in PLAXIS that take into account the stress
dependency of the soil stiffness. These advanced models will be introduced in a later exercise. The soil
parameterscanbeenteredintoTable1.


Figure2:MohrCoulombmodelforthesandlayer.

Table1:Soilmaterialsetparameters.
Parameter Symbol Topsand Middle Lower Deepsand Deepsand Unit
sand sand 1 2
Mohr Mohr Mohr Mohr Mohr
Materialmodel Model
Coulomb Coulomb Coulomb Coulomb Coulomb
Typeofbehaviour Type Drained Drained Drained Drained Drained
Dryweight unsat 18.0 18.0 18.0 18.0 18.0 kN/m3
Wetweight sat 18.0 18.0 18.0 18.0 18.0 kN/m3
Permeabilityxdir. kx 0.0 0.0 0.0 0.0 0.0 m/d
Permeabilityydir. ky 0.0 0.0 0.0 0.0 0.0 m/d
Young'smodulus* E ? ? ? 280.000 470.000 kN/m2
Poisson'sratio 0,33 0,33 0,33 0,33 0,33
Cohesion c ? ? ? 1,0 1,0 kN/m2
Frictionangle ? ? ? 35 35
Dilatancyangle ? ? ? 5 5
Interfacestrength
Rinter 0.6 0.6 Rigid(1.0) Rigid(1.0) Rigid(1.0)
reduction

Sinceanexcavationisconsideredinthisexercise,theinputoftheYoungsmodulusEshouldbebasedonunloading,
ratherthanonprimaryloading.

CourseComputionalGeotechnics 4
Page 220
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

GEOMETRYINPUT
Startanewproject
Entergeneralsettings
Accept the default values in the Project tab sheet of the General settings (15node elements). For the
dimensionsseeFigure3.


Figure3:Generalsettings,tabsheetdimensions.

EnterGEOMETRY,PLATE,interfaces,anchors,andgeotextiles
Enterthegeometryasproposedinfigure4.

Clicktheplatebutton ,tointroducethediaphragmwall.

ClicktheGeogridbutton ,tointroducethetwogeotextileelementsthatrepresentthegroutbody.

ClicktheInterfacebutton ,whichwillpresentthecursorintheInterfacemode.Asinterfacescanbe
introducedonbothsidesofageometryline,oneshouldpayattentiontothearrowsonthecursor.These
arrowsindicatewheretheprogramwilllocatetheinterfaces.Clickonthetopofthediaphragmwallfirst
andthenonthebottomtoentertheinterfaceattheleftsideofthewall.Clickonthebottomandthenon
thetoptoentertheinterfaceontherightside.

Hint: Pleasenotethattheinterfaceisextendedforashortdistanceunderneaththebeam.Thisisdone
toovercomeasingularpointatthebottomofthewall(seeFigure4).

Click the Nodetonode anchor button , and introduce the two anchors. These anchors connect the
beginningofthegroutbodytothewall.

Enterfixities

Clickthestandardfixitiesbutton ,forthestandardboundaryconditions.

CourseComputionalGeotechnics 5
Page 221
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

Figure4:Geometryofthemodel

INPUTMATERIALPROPERTIES
Soilandinterfaces
Enterthematerialpropertiesforthethreesoildatasets,asdeterminedinthefirsttableofthisexercise.
After entering all properties for the three soil types, drag and drop the properties to the appropriate
clusters.
Entermaterialpropertiesfortheplates,anchorsand'geogrids'asindicatedinTable2,Table3andTable4.
Table2:Propertiesofthediaphragmwall(plate)
Parameter Symbol Diaphragmwall Unit
MaterialModel Model LinearElastic
Normalstiffness EA 8.0*107 kN/m
Flexuralrigidity EI 1.5*106 kNm2/m
Weight w 8.0 kN/m/m
Poisson'sratio 0.0

Table3:Propertiesoftheanchorrod(nodetonodeanchors).
Parameter Symbol Anchorrod Unit
MaterialModel Model Elastic
Normalstiffness EA 2.0*105 kN/m
Spacing Ls 1.0 m

Table4:Propertyofthegroutbody(geogrid).
Parameter Symbol Groutbody Unit
5
Normalstiffness EA 2.0*10 kN/m

CourseComputionalGeotechnics 6
Page 222
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

MESHGENERATION
FromtheMeshmenu,settheGlobalcoarsenesstoMedium.
Clickonthemeshgeneratorbutton,whichwillpresenttoFEmeshshowninFigure5.

Figure5:MediumFiniteElementmesh

SelecttheClustersaroundthediaphragmwallandpressRefineClustersfromtheMeshmenu.Thiswill
resultinarefinementaroundtheselectedClustersasshowninFigure6.

CourseComputionalGeotechnics 7
Page 223
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)


Figure6:RefinedFiniteElementmesh

CALCULATION
StarttheCalculationprogramandchooseClassicmodeintheSelectcalculationmodewindow(Figure7).


Figure7:Selectcalculationmodewindow

Initialconditions(Figure9)
Initialstresses
ChooseK0procedurefromtheCalculationtypecomboboxoftheGeneraltabsheetinthecalculation
window.

Initialgeometryconfiguration
Checkthatthestructureelementsarenotactive.Inactiveelementsareindicatedbythethickgreylines.If
necessary click on the plate chain to deactivate them. Structure elements are not part of the initial
situationandthereforetheyneedtobedeactivated.

CourseComputionalGeotechnics 8
Page 224
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

Porepressuregeneration
Asthephreaticsurfaceislocatedbelowthedeepsend2layer.ThegenerationofPorepressurescanbe
skipped.

Constructionprocess
The entire construction process consists of five phases. Define the phases, as shown graphically below. For
eachphase,usethePlasticcalculation,Stagedconstruction.
Inthefirstphase(Figure10)thediaphragmwallisactivatedandthefirstexcavationtakesplace.Notethat
though the the interfaces along the wall are activated automatically with the activation of the wall; the
extensionsbelowthediaphragmwallhavetobeactivatedmanually.
Inthesecondphase(Figure11),anewoptionisused,namelytheprestressingofanchors.
First the groutbody (the geogrid) is switched on by clicking on the 'geogrid' element. The element will
appearinyellowassoonasitisswitchedon.Thelightgreycolorindicatesnonactiveelements.
Now the groutbody is active and the anchor element needs to be prestressed. By double clicking on a
nodetonode anchor a window (Figure 8) will appear as indicated below. In the nodetonode anchor
window one can enter the prestress force. A black nodetonode anchor indicates that the anchor is
activated.TheletterPindicatesthataprestressforcewillbeactiveintheanchor.

Createtheremainingphases,asindicatedinFigure12,Figure13andFigure14.Selectsomenodesfortheload
displacementcurves(e.g.topofwall).

Figure8:Nodetonodeanchorwindow

Hint: Whenprocessingananchorinacertaincalculationphasetheanchorforcewillexactlymatchthe
prestressforceattheendofthatphase.Infollowingcalculationphaseswithoutprestressing,the
anchorforcewillbeinfluencedbytheexcavationprocess.

Figure9:Initialconditions,diaphragmwallandanchorsare Figure10:Phase1,excavationandactivationofthe

CourseComputionalGeotechnics 9
Page 225
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

notactive. diaphragmwall.

Figure11:Phase2,activationoftheanchorandgeotextile Figure12:Phase3,excavationofthesecondpart.
(grouting).Prestressingoftheanchortoavalueof300
kN/m.

Figure13:Phase4,activationofthesecondanchorand Figure14:Phase5,excavationofthethirdpart.
geotextile(grouting).Prestressingoftheanchortoavalue
of300kN/m.

CourseComputionalGeotechnics 10
Page 226
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

INSPECTOUTPUT
Bydoubleclickingonthenodetonodeanchors,Plaxiswillpresentabox,inwhichthestressintheanchormay
beinspected.

15,3cmmaximumheave

Figure15:Deformedmesh(MCmodelwithE50modulus)

4,4cmmaximumheave

Figure16:Deformedmesh(MCmodelwithEurmodulus)

CourseComputionalGeotechnics 11
Page 227
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

Figure17:Soleheaveafterthelastexcavationstep

CourseComputionalGeotechnics 12
Page 228
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

Figure18:Bendingmomentsafterthelastexcavationstep

DETERMINATIONOFPARAMETERSFROMTRIAXIALTEST


Figure19:MohrCoulombmodelforthesandlayer.

STRENGTHPARAMETERS
Fillin1and3intheMohrCoulombcriteria:
1 3 ( 1 3 ) sin 2 c cos

Sincecwillbesmall,assumec=0:

1 3
sin
1 3

370 100
sin
370 100

35 Forreasonsofnumericalstabilityusec=1kPa
30 5




CourseComputionalGeotechnics 13
Page 229
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(MCModel)

Figure20:StrengthparameteroftheMohrCoulombmodel.

STIFFNESSPARAMETERS
Sinceexcavationisconsideredinthisexercise,theinputofYoungsmodulusEshouldbebasedonunloading,
ratherthanonprimaryloading.

3=100kPa(correspondswithreferencepressure,E50=E50ref)

v 135
E 50 2.0 10 4 kN/m assumeEur4E50
v 0.675%

For improved modelling, three regions are distinguished, for each individual layer an average stiffness is
determinedbyusing:
3
E 50 E 50
ref
E50EloadEurEunload
p ref

Wedistinguish:
Toplayer(fromtopto15m)
Middlelayer(from15mto25m)
Lowerlayer(from25mto35m)

Toplayer:
3=depthsoilK0
3=averagehorizontalstress(referencepointat7.5m)
3=7.518(1sin)=58kPa
top 58
E 50 2.0 104 1.5 104 kPa
100
Eurtop4E50top=6.0104kPa

Middlelayer:
3=2018(1sin)=153kPa
middle 153
E 50
2.0 104 2.5 104 kPa
100
Eurmiddle4E50middle=1.0105kPa

Lowerlayer:
3=3018(1sin)=230kPa
bottom 230
E 50
2.0 104 3.0 104 kPa
100
Eurbottom4E50bottom=1.2105kPa

Assumeforalllayers=0.33.

CourseComputionalGeotechnics 14
Page 230
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(HS,HSS)

EXCAVATIONOFBUILDINGPIT(HSModelandHSSModel)

FranzTschuchnigg

COMPUTIONALGEOTECHNICS

CourseComputionalGeotechnics 1
Page 231
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(HS,HSS)

CourseComputionalGeotechnics 2
Page 232
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(HS,HSS)

INTRODUCTION
In the next exercise the geometry from Excavation of building pit in Limburg is used. Instead of the Mohr
Coulombmodel,theHardeningSoilandtheHardeningSoilSmallmodelisused.
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 60 cm diameter bored piles; the wall is anchored by two rows of pre
stressedgroundanchors.Inthisexercisetheconstructionofthisbuildingpitissimulatedandthedeformation
andbendingmomentsofthewallareevaluated.
Theupper40mofthesubsoilconsistsofamoreorlesshomogeneouslayerofmediumdensefinesandwitha
unitweightof18kN/m3.TriaxialtestdataofarepresentativesoilsampleisgivenintheexerciseExcavationof
buildingpitinLimburg.Thegroundwatertableisverydeepanddoesnotplayaroleinthisanalysis.

GEOMETRYINPUT
SelecttheexistingprojectoftheExcavationofbuildingpitinLimburgexercise.
FromtheFilemenuselectSaveAsandsavetheexistingprojectunderanewfilename.

GEOMETRYINPUTHSMODEL
ModifythematerialpropertiesbyselectingtheitemSoils&InterfacesfromtheMaterialsmenu.Usethe
materialpropertiesasgiveninTable1.AssigntheSandmaterialsettoallsoilclusters.

Stage 1 Diaphragm wall

Stage 2 Anchor rod top sand

Stage 3

Grout body middle sand

lower sand

deep sand 1

deep sand 2


Figure1:Geometryfortiedbackexcavation.

CourseComputionalGeotechnics 3
Page 233
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(HS,HSS)

Table1:SoilmaterialsetparametersforHSmodel.
Parameter Symbol Top/Middle Lowersand Deepsand1 Deepsand2 Unit
sand
Materialmodel Model HSmodel HSmodel HSmodel HSmodel
Typeofbehaviour Type drained drained drained drained
Dryweight unsat 18 18 18 18 kN/m3
Wetweight sat 18 18 18 18 kN/m3
ref
Young'smodulus E50 20.000 20.000 40.000 60.000 kN/m2
Oedometermod. Eoedref 20.000 20.000 40.000 60.000 kN/m2
Unloadingmodulus Eurref 80.000 80.000 160.000 240.000 kN/m2
Power m 0,5 0,5 0,5 0,5
Poisson'sratio 0,2 0,2 0,2 0,2
Referencestress Pref 100 100 100 100 kN/m2
Cohesion c 1,0 1,0 1,0 1,0 kN/m2
Frictionangle 35 35 35 35
Dilatancyangle 5,0 5,0 5,0 5,0
Interfacestrength
Rinter 0,6 1,0 1,0 1,0
reduction
nc
K0 K0=1sinf 0,426 0,426 0,426 0,426

CALCULATION
Constructionprocess
The entire construction process consists of five phases. Define the phases, as shown graphically below. For
eachphase,usethePlasticcalculation,Stagedconstruction.
Createtheremainingphases,asindicatedinFigure2,Figure3,Figure4,Figure5,Figure6andFigure7.Select
somenodesfortheloaddisplacementcurves(e.g.topofwall).

Figure2:Initialconditions,diaphragmwallandanchorsare Figure3:Phase1,excavationandactivationofthe
notactive. diaphragmwall.

CourseComputionalGeotechnics 4
Page 234
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(HS,HSS)

Figure4:Phase2,activationoftheanchorandgeotextile Figure5:Phase3,excavationofthesecondpart.
(grouting).Prestressingoftheanchortoavalueof300
kN/m.

Figure6:Phase4,activationofthesecondanchorand Figure7:Phase5,excavationofthethirdpart.
geotextile(grouting).Prestressingoftheanchortoavalue
of300kN/m.

CourseComputionalGeotechnics 5
Page 235
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(HS,HSS)

GEOMETRYINPUTCHANGEMATERIAL(HSSMODEL)
ModifythematerialpropertiesbyselectingtheitemSoils&InterfacesfromtheMaterialsmenu.Usethe
materialpropertiesasgiveninTable2.AssigntheSandmaterialsettoallsoilclusters.

Table6:SoilmaterialsetparametersforHSSmodel.
Parameter Symbol Topsand/Middle Lowersand/Deepsand1 Unit
sand /Deepsand2
Materialmodel Model HSSmodel HSmodel
Typeofbehaviour Type drained drained
Dryweight unsat 18 18 kN/m3
Wetweight sat 18 18 kN/m3
Young'smodulus E50ref 20.000 20.000 kN/m2
Oedometermod. Eoedref 20.000 20.000 kN/m2
Unloadingmodulus Eurref 80.000 80.000 kN/m2
Power m 0,5 0,5
Poisson'sratio 0,2 0,2
Referencestress Pref 100 100 kN/m2
Cohesion c 1,0 1,0 kN/m2
Frictionangle 35 35
Dilatancyangle 5,0 5,0
Interfacestrengthreduction Rinter 0,6 1,0
K0nc K0=1sinf 0,426 0,426
Shearmodulus G0ref 100.000 100.000 kN/m2
Shearstrain g 0.7 0,0001 0,0001

CourseComputionalGeotechnics 6
Page 236
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(HS,HSS)

INSPECTOUTPUT
ThemainresultsforthisexerciseareillustratedinFigure8,Figure9andFigure10.Themaximumvaluesare
listedbelowinTable7.


Figure8:heaveafterthelastexcavationstep[mm]

CourseComputionalGeotechnics 7
Page 237
Advanced Computational Geotechnics
EXCAVATIONOFBUILDINGPIT(HS,HSS)

Figure 9: Bending moments after the last excavation step Figure10:lateraldeflectionofthesheetpilewallafterthelast


[kNm/m] excavationstep[mm]

Table7:Maximumvalueoftheresultsafterthelastexcavationstep
Parameter Symbol MC(E50) MC(Eur) HS HSS Unit
Max.Soleheave u y 153 44 58 56 mm
Max.Lateraldeflection ux 46 13 16 13 mm
Max.Surfacesettlements u y 8 0 7 8 mm
Max.Bendingmoments My 536 318 465 416 kNm/m

CourseComputionalGeotechnics 8
Page 238
Advanced Computational Geotechnics

EXERCISE 3: EXCAVATION OF BUILDING PIT


Exercise created by: Dr Franz Tschuchnigg
Slides from: Dr Lee Siew Wei & Dr William Cheang

CONTENTS

A. INTODUCTION

B. THE MODELS (EXERCISE 3A & 3B)

C. RESULTS

D. DISCUSSIONS

Page 239
Advanced Computational Geotechnics

PART 1

EXERCISE 3A

INTRODUCTION

Page 240
INTRODUCTION
Advanced Computational Geotechnics

GEOMETRY AND COORDINATES

Page 241
STRENGTH AND
Advanced MODULUS
Computational GeotechnicsVALUES

WHAT VALUES ARE SUPPOSE TO USE?

CAN YOU DETERMINE THE VALUES?

LAYER PHI C DILATANCY E

TOP SAND ? ? ? ?

MIDDLE SAND ? ? ? ?

LOWER SAND ? ? ? ?

DENSE 1
The parameters for layer Dense 1 & Dense 2 are given
DENSE 2

Page 242
Lets usComputational
Advanced find thisGeotechnics
values

Find E50 & Eur

Page 243
STAGES
Advanced Computational Geotechnics

RESULTS

Page 244
Inspect Output 1:Advanced
Deformed mesh
Computational for MC-E50 (Stage 5)
Geotechnics

MC-model (E50) 25X

Inspect Output 1: Deformed mesh MC-EUR Stage 5

MC-model (Eur) 25X

Page 245
Inspect Output 2: Displacements (Shading)
Advanced Computational Geotechnics for MC E50 Stage 5

MC-model (E50)
MC-model (Eur)

Inspect Output 2: Displacements (Shading) for MC-EUR Stage 5

MC-model (Eur)

Page 246
Advanced Computational Geotechnics

MOHR-COULOMB : OPERATIVE MODULUS EUR vs E50

EXERCISE 3A

Lateral deflection after Phase 5


0
-1
-2
-3
-4
Basal Heave After Stage 5 -5
200 -6
-7
180
-8
160 -9
-10
140
-11
Depth [m]
Heave (mm)

120 -12
-13
100
-14
80 -15
-16
60
-17
40 -18
-19
20
-20
0 -21
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 -22
-23
Distance from retaining wall
-24
MC_E50 MC_EUR -25
-50-45-40-35-30-25-20-15-10 -5 0 5 10 15 20 25 30 35 40 45 50
Lateral deflection [mm]

Disp-MC-E50 Disp MC-EUR

Page 247
Advanced Computational Geotechnics

Bending Moments After Phase 5 Bending Moments Envelope


0 0
-1 -1
-2 -2
-3 -3
-4 -4
-5 -5
-6 -6
-7 -7
-8 -8
-9 -9
-10 -10
Depth [m]

-11 -11

Depth [m]
-12 -12
-13 -13
-14
-14
-15
-15
-16
-16
-17
-17
-18
-18
-19
-19
-20
-20
-21
-22 -21
-23 -22
-24 -23
-25 -24
-600 -500 -400 -300 -200 -100 0 100 200 300 400 500 600 -25
Bending Moments [kNm/m] -600 -500 -400 -300 -200 -100 0 100 200 300 400 500 600
Bending Moments [kNm/m]

BMD @ Phase 5 [MC-E50] BMD @ Phase 5 [MC-EUR]


BMD-min [MC-E50] BMD-min [MC-EUR]
BMD @ Phase 5 [MC-EUR] BMD @ Phase 5 [MC-E50] BMD-max [MC-E50] BMD-max [MC-EUR]

PART 2

EXERCISE 3B

Page 248
Advanced Computational Geotechnics
SAME MODEL, BUT CONSTITUTIVE MODEL IS HS & HS-small

MODEL: EXERCISE 3B-HS

Page 249
MODEL: EXERCISE
Advanced 3B HS-small
Computational Geotechnics

RESULTS

EXERCISE 3B

Page 250
INSPECT THE MODEL
Advanced Computational Geotechnics

A. DISPLACEMENTS
1. Type of movements?
2. Lateral wall displacements?
3. Basal heave?

B. FORCES &BENDING MOMENTS


1. Anchor forces
2. Bending moments from PHASE 1 to 5 (How are they fluctuating?)
3. Whats the effect of pre-stress on the shape (response) and magnitudes.
4. Compare all these aspects for MC-E50, MC-EUR, HS & HS-small.

C. For excavations where the design is govern by serviceability how


should you model the problem?
D. What have we learned?

Basal Heave After Stage 5


Lateral deflection after Phase 5
200
0
180 -1
-2
160
-3
140 -4
-5
Heave (mm)

120
-6
100 -7
-8
80
-9
60 -10
-11
Depth [m]

40
-12
20 -13
-14
0 -15
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
-16
Distance from retaining wall -17
MC_E50 MC_EUR HS-small HS -18
-19
-20
-21
-22
-23
-24
-25
-50-45-40-35-30-25-20-15-10 -5 0 5 10 15 20 25 30 35 40 45 50
Lateral deflection [mm]

Disp-Hssmall Disp-HS Disp-MC-E50 Disp MC-EUR

Page 251
Advanced Computational Geotechnics

Bending Moments After Phase 5


0
-1
-2
-3
-4
-5
-6
-7
-8
-9
-10

Depth [m]
-11
-12
-13
-14
-15
-16
-17
-18
-19
-20
-21
-22
-23
-24
-25
-600 -500 -400 -300 -200 -100 0 100 200 300 400 500 600
Bending Moments [kNm/m]

BMD @ Phase 5 [Hssmall] BMD @ Phase 5 [HS]


BMD @ Phase 5 [MC-EUR] BMD @ Phase 5 [MC-E50]

Settlement Behind RE-Wall after Phase 5


10
Settlemen (-ve), Heave (+ve) [mm]

-10
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Distance from wall [m]

Settle-MC E50 Settle-Hssmall

Page 252
Advanced Computational Geotechnics

HKCG 10
1

DRAINED AND
A U
UNDRA
AINED ANAL
LYSIS

Professo
P or Anton
nio Gens
s

Page 253
Advanced Computational Geotechnics

CG10: DRAINED AND UNDRAINED ANALYSIS

Antonio Gens

Technical University of Catalunya, Barcelona

some of the slides were originally created by:


Helmut Schweiger (Technical University of Graz, Austria)
Ronald Brinkgreve (Plaxis bv / Delft University of Technology)
Cino Viggiani (Laboratoire 3S, Grenoble, France)

outline

drained / undrained (conditions and analysis)


drained / undrained soil behaviour
typical results from drained and undrained triaxial tests
strength parameters in effective stresses and total stresses
what is the critical case: drained or undrained?
modelling undrained behaviour with Plaxis
general procedure
three methods
Method A: effective stresses
Method C: total stresses
Method B (effective stress, hybrid method)
undrained shear strength (for Method A)
undrained behaviour with Mohr-Coulomb Model
undrained behaviour with advanced models
influence of dilatancy
summary

Page 254
Advanced Computational Geotechnics

drained / undrained (conditions and analysis)

in undrained conditions, no water movement takes place and,


therefore, excess pore pressures are built up
u 0, '

in drained conditions, no excess pore pressures are built up


u = 0, = '

drained analysis appropriate when


permeability is high
rate of loading is low
short term behavior is not of interest for problem considered

undrained analysis appropriate when


permeability is low and rate of loading is high
short term behavior has to be assessed

triaxial test (NC soils) drained / undrained

drained undrained

Page 255
Advanced Computational Geotechnics

triaxial test (OC soils) drained / undrained

drained undrained

stress paths in undrained triaxial test NC / OC

1 3
t
2
3
s' 1
2
3
s 1
2

Page 256
Advanced Computational Geotechnics

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

q
ine
el
advanced ur
f ail
models

advanced
models
elastic-perfectly
cu,3
2c u,3 plastic models
cu,2
c2cu,1u,1 2c u,2

pc p

Results from undrained triaxial tests using simple and advanced constitutive models

Strength parameters
Mohr-Coulomb parameters in terms of effective stress


c tan

in terms of principal effective stresses

1 3 3 c c
1 sin t s sin
tan tan
;
2 2

Page 257
Advanced Computational Geotechnics

Strength parameters
Mohr-Coulomb parameters in terms of principal effective stress
tan = sin

t
1 3
2 a = ccos

s'
'1 '3 , s 1 3
2 2

1 3 3 c c
1
sin ; t s
sin ; t s sin c cos
2 2 tan tan

Strength parameters
Mohr-Coulomb parameters in terms of total stresses
Only undrained conditions!

c tan total stresses

Cu
1 3 1 3
2
F
2
F cu ,

-Cu
Effective stresses

Soil behaves as if it was cohesive


cu ( su ) : undrained shear strength
cu only changes if drainage occurs (no change if undrained conditions prevail)
Page 258
Advanced Computational Geotechnics

What is the critical case: drained or undrained?


t

OC
OC

NC
NC
Loading

Unloading
s, s

note that for soils in general:


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

What is the critical case: drained or undrained?


t

OC
OC

NC
NC
Loading

Unloading
s, s

For very soft NC soil, factor of safety against failure may


be in some cases lower for short term (undrained)
conditions for unloading problems (e.g. excavations)
For very stiff OC soil, factor of safety against failure may
be in some cases lower for short term (undrained)
conditions for loading problems (e.g. embankment)

Page 259
Advanced Computational Geotechnics
FE modeling of undrained behavior
undrained analysis in terms of effective stress (Plaxis 2010)
type of material behaviour: undrained (Plaxis < 2010)
the issue:
constitutive equations are formulated in terms of
' D'
we need to compute D
D

FE modeling of undrained behavior

principle of effective stress ' f


with f pw pw pw 0 0 0
T

since the strains are the same in each phase,


' D '
13 03 Kf
f D f D f Ke Ke
03 03 n
pore fluid stiffness, related to the
bulk modulus of pore fluid (water) Kf

We need D D
' f D D ' D f ( D ' D f )

D D' D f
Page 260
Advanced Computational Geotechnics

FE modeling of undrained behavior


Example: linear elastic model + plane strain

D D' D f K
E
G
E
3(1 2 ) 2(1 )
4 2 2 G G'
K ' 3 G K ' G K ' G 0
xx
3 3
xx
e

K ' 2 G
K ' G 0 eyy
4 2
K ' G
yy 3 3 3 D
zz zz
e

K ' 2 G K ' G 0 e
2 4
xy K ' G
3 3 3 xy
0 G
0 0

4 2 2
K 3 G K G K G 0
xx
3 3
xx
e

K 2 G
K G 0 eyy
4 2
yy K G

zz
3 3 3
zz
e D
K 2 G K G 0 e
2 4
xy K G
3 3 3 xy
0 G
0 0

FE modeling of undrained behavior


Example: linear elastic model + plane strain

D D' D f
4 2 2
K ' 3 G K ' G
3
K ' G 0
3
Ke Ke Ke 0
K ' 2 G K 0
K ' G 0
4 2
K ' G e Ke Ke
D 3 3 3
Ke Ke Ke 0
K ' 2 G 2
K ' G
4
K ' G 0
3 3 3 0 0 0 0
0 G
0 0

4 2 2
K ' 3 G K e K ' G K e
3
K ' G K e
3
0

K ' 2 G K 4
K ' G K e
2
K ' G K e 0
D 3
e
3 3

K ' 2 G K e 2
K ' G K e
4
K ' G K e 0
3 3 3
G
0 0 0

Page 261
Advanced Computational Geotechnics

FE modeling of undrained behavior


Example: linear elastic model + plane strain
D D' D f
4 2 2
K ' 3 G K e K ' G K e
3
K ' G K e
3
0

K ' 2 G K 4
K ' G K e
2
K ' G K e 0
D D ' De 3
e
3 3

K ' 2 G K e 2
K ' G K e
4
K ' G K e 0
3 3 3
G
0 0 0

4 2 2
K 3 G K G
3
K G 0
3
4 4
K 2 G 4
K G K G 0
2 K ' G K e K G
D 3 3 3 3 3

K 2 G 2
K G
4
K G 0
3 3 3
0
0 0 G
K K ' K e

FE modeling of undrained behavior

all the above (which is valid for any soil (or model) for which the
principle of effective stress applies) can be easily combined
with the FEM

instead of specifying the components of D, specify D' and Ke


D D' D f
when calculating stresses,
f K e v
' f
' D'

Page 262
Advanced Computational Geotechnics

FE modeling of undrained behavior

a value must be set for Ke


real value of Ke = Kw/n ~ 1106 kPa (for n = 0.5)
in fact, the pore-fluid is assigned a bulk modulus that is
substantially larger than that of the soil skeleton (which
ensures that during undrained loading the volumetric strains
are very small)

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 31 2 u 31 2 u

E' 1 u
K total assuming u = 0.495
31 2 u 1 '
Note: this procedure gives reasonable results only for ' < 0.35 !

modeling undrained behavior with PLAXIS

method A (analysis in terms of effective stresses):


effective strength parameters (MC: c', ', )
effective stiffness parameters (MC: E50', )

method B (analysis in terms of effective stresses):


total strength parameters c = cu, = 0, = 0
effective stiffness parameters E50', '

method C (analysis in terms of total stresses):


total strength parameters c = cu, = 0, = 0
total stiffness parameters Eu, u = 0.495

Page 263
Advanced Computational Geotechnics
FE modeling of undrained behavior (method A)
analysis in terms of effective stress
type of material behaviour: undrained method A (version 2010)
undrained (version < 2010)
u changes (excess pore water pressures generated)
constitutive equations are formulated in terms of
' D '
In the case of Mohr Coulomb model:
effective strength parameters c, ,
effective stiffness parameters E50', '

the total stiffness matrix is computed as: D D' D f

available for all models (including user defined models)!

FE modeling of undrained behavior (method A)


t

uf
u
TSP
cu ESP

s, s

single set of parameters in terms of effective stress (undrained,


drained, consolidation analysis consistent)
realistic prediction of pore pressures (if model is appropriate)
the undrained analysis can be followed by a consolidation
analysis (correct pore pressures, correct drained parameters
and Cu is updated automatically)
Cu is a consequence of the model, not an input parameter!!
Page 264
Advanced Computational Geotechnics

modeling undrained behavior with PLAXIS

method A (analysis in terms of effective stresses):


effective strength parameters (MC: c', ', )
effective stiffness parameters (MC: E50', )

method B (analysis in terms of effective stresses):


total strength parameters c = cu, = 0, = 0
effective stiffness parameters E50', '

method C (analysis in terms of total stresses):


total strength parameters c = cu, = 0, = 0
total stiffness parameters Eu, u = 0.495

FE modeling of undrained behavior (method C)


analysis in terms of total stress
type of material behaviour: undrained method C (version 2010)
drained (version < 2010) (in spite of
modelling an undrained situation)
u does not change
constitutive equations are formulated in terms of
total strength parameters c = cu, = 0, = 0
total stiffness parameters Eu, u = 0.495

available for Linear elastic, Mohr Coulomb, NGI-ADP models

Page 265
Advanced Computational Geotechnics

FE modeling of undrained behavior (method C)


t

TSP=ESP
cu
s, s

parameters in terms of total stress


no prediction of pore pressures (only total stresses
are obtained)
the undrained analysis can not be followed by a
consolidation analysis
Cu is an input parameter!!

modeling undrained behavior with PLAXIS

method A (analysis in terms of effective stresses):


effective strength parameters (MC: c', ', )
effective stiffness parameters (MC: E50', )

method B (analysis in terms of effective stresses):


total strength parameters c = cu, = 0, = 0
effective stiffness parameters E50', '

method C (analysis in terms of total stresses):


total strength parameters c = cu, = 0, = 0
total stiffness parameters Eu, u = 0.495

Page 266
Advanced Computational Geotechnics

FE modeling of undrained behavior (method B)


analysis in terms of effective stress
type of material behaviour: undrained method B (version 2010)
undrained (version < 2010)
u changes
constitutive equations are formulated in terms of (but strength
in total stresses!)
total strength parameters c = cu, = 0, = 0
effective stiffness parameters E50', '
' D '
D
D D' D f
3 E
Resulting undrained stiffness parameters Eu ; u 0.495
2 1
available for Mohr Coulomb, HSM, HSSM, NGI-ADP models

FE modeling of undrained behavior (method B)


t

ESP TSP
cu
s, s

parameters in terms of total stress and effective stress


prediction of pore pressures (generally unrealistic)
the undrained analysis should not be followed by a
consolidation analysis (pore pressures unrealistic and
Cu is not updated automatically)
Cu is an input parameter!!

Page 267
Advanced Computational Geotechnics
Undrained shear strength (for Method A)
In method A, Cu is a consequence of the model, not an input parameter!!
Therefore, it is necessary to know what value of Cu we are actually using

In the case of the Mohr-Coulomb model (in plane strain), it is easy to compute
Cu analytically
For plane strain: the undrained effective stress path rises vertically
2 0 '2 ' ('1 '3 ) ; '2 ' ('1 '3 )
1
E
Linear Elasticity
p '
0 p ' 0 p ' '1 '2 '3 '1 '3 (1 ' ) 0
1 1
v
K' 3 3
1
'1 '3 s' 0 tan = sin
2
t
1 3
2

Effective Stress
a = ccos

Path, ESP
A to , so s'
'1 '3 , s 1 3
2 2

Undrained shear strength from the Mohr Coulomb model

Plane strain: effective stress path rises vertically

tan = sin

t
1 3 Effective Stress
2
Path, ESP B
a = ccos

cu
A
to , so

s'
'1 '3 , s 1 3
2 2

1
cu c 'cos ' so 'sin ' c 'cos ' vo ho sin '
2
cu c 'cos ' 1
1 K 0 sin ' , K 0 ho
'v 0 'v 0 2 vo
Page 268
Advanced Computational Geotechnics

Undrained shear strength from the Mohr Coulomb model

The Mohr Coulomb model in terms of effective stresses


OVERESTIMATES the undrained shear strength of soft clays!

cuMC

cu real
s, s

Undrained shear strength from advanced models

Although it is possible, in a few simple cases, to obtain an analytical


expression for cu, it is advisable to perform a numerical laboratory
test to check the value of undrained shear strength actually supplied
by the model
It is important to perform the numerical laboratory test under the
same condition as in the analysis
Plane strain, triaxial, simple shear
Correct initial stresses
Compression, extension, simple shear
Not all cu values are achievable with a particular model

Page 269
Advanced Computational Geotechnics
Soft soil model

Parameters
c ' 0.1 kPa ' 23 K 0NC 1 sin ' 0.609 ur 0.15 * 0.11 * 0.0275

140.00
140.00

120.00
120.00

100.00 100.00
1-3 (kPa)

80.00

1-3 (kPa)
80.00

60.00 60.00

40.00
40.00

20.00
20.00

0.00
0 20 40 60 80 100 120 140 160 0.00
0.00 20.00 40.00 60.00 80.00 100.00 120.00 140.00 160.00
p' (kPa)
p' (kPa)

Triaxial (compression) Triaxial (extension)

cu/v=0.279 cu/v=0.214

Soft soil model

Parameters
c ' 0.1 kPa ' 23 K 0NC 1 sin ' 0.609 ur 0.15 * 0.11 * 0.0275

140.00

120.00

100.00
1-3 (kPa)

80.00

60.00

40.00

20.00

0.00
0 20 40 60 80 100 120 140 160
p' (kPa)

Plane strain (compression) Plane strain (extension)

cu/v=0.279 cu/v=0.277

Page 270
Advanced Computational Geotechnics

influence of dilatancy on undrained shear strength

if we set 0 then, negative volumetric plastic deformations


occur at failure:

v ve vp (elastic-plastic behavior)

v 0 (undrained conditions)

vp 0 ve 0 p ' K ' ve 0
At failure: q M p ' q 0
t s sin t 0

result: unlimited increase of q (or t), i.e. infinite strength!!

Therefore, in undrained analysis, dilatancy, , must be set to zero!

influence of constitutive model and dilatancy

simulation of undrained triaxial compression test MC / HS model - q vs 1

300

275

250

225

200

175
q [kN/m ]
2

150

125

100

75
MC non dil
50 MC dil
HS_1 non dil
25 HS_1 dil

0
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00

1 [%]

Page 271
Advanced Computational Geotechnics

influence of constitutive model and dilatancy

simulation of undrained triaxial compression test MC / HS model - q vs p'

300

275 MC non dil


MC dil
250 HS_1 non dil
HS_1 dil
225 total stress path
200

175
q [kN/m ]
2

150

125

100

75

50

25

0
0.00 25.00 50.00 75.00 100.00 125.00 150.00 175.00 200.00 225.00 250.00
2
p' [kN/m ]

summary
FEM analysis of undrained conditions can be performed in effective
stresses and with effective stiffness and strength parameters
(Method A)
Method A must be used:
if consolidation/long term analysis are required
advanced soil models are adopted
undrained shear strength is a result of the constitutive model
care must be taken with the choice of the value for dilatancy angle
Methods B and C provide alternative ways to analyze undrained
problems but:
the constituive model dos not generally represent the true soil
behaviour (before failure)
potentially useful for stability problems in undrained conditions
(specification of undrained shear strength is straightforward)

Page 272
Advanced Computational Geotechnics

HKCG 11
1

CONS
SOLIDA
ATION
N ANA
ALYSIS
S

Professo
P or Anton
nio Gens
s

Page 273
Advanced Computational Geotechnics

CG11: CONSOLIDATION ANALYSIS

Antonio Gens

Technical University of Catalunya, Barcelona

some of the slides were originally created by:


Helmut Schweiger (Technical University of Graz, Austria)
Dennis Waterman (Plaxis)

CONSOLIDATION: OUTLINE

Introduction
Theory of consolidation
Permeability
FEM for consolidation analysis
New features in PLAXIS 2D 2010

Page 274
Advanced Computational Geotechnics

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

Page 275
Advanced Computational Geotechnics

OTHER EXAMPLES

1973 1984

Ekofisk tank

OTHER EXAMPLES

Construction at intermediate rates

Change of hydraulic conditions

Page 276
Advanced Computational Geotechnics

CONSOLIDATION

Equation of continuity (with soil deformation)

q x q y n
div q
x y t
n
n q x q y t
0
t x y

change of n depends on changes of effective stresses (flow


problem depends on the mechanical problem)
change of pore pressures (effective stresses) is derived from the
flow problem (mechanical problem depends on the flow problem)
mechanical and flow problem interact and must be solved
simultaneously
COUPLED PROBLEM!

Theory of consolidation

Considering:

Homogeneous, fully saturated, isotropic linear elastic soil


Incompressible particles

Volumetric strain of saturated soil is caused by:


n
1. Pore pressure change: v1 pw
Kw

2. Storage of pore water:


v 2 qt

v = volumetric strain Kw = bulk stiffness of water


pw = (excess) pore pressure q = pore water flow
n = porosity t = time increment

Page 277
Advanced Computational Geotechnics
Theory of consolidation

Pore water flow:


k k
Darcys law: q pw v 2 pw t
w w

Total change of volumetric strain in time,


considering homogeneous permeability:

v ( v1 v 2 ) n pw k 2
pw
t t K w t w

k = permeability
w = unit weight of water

Theory of consolidation

General 3D case:

v 1 p ' 1 ( p pw ) 1 p 1 pw

t K ' t K ' t K ' t K ' t

1 p 1 n pw k 2
pw
K ' t K ' K w t w

E'
where K' = bulk stiffness of soil skeleton and p = mean total stress
3 1 2 '

kK' pw p
Considering incompressible water: 2 pw
w t t

Page 278
Advanced Computational Geotechnics
Theory of consolidation

1D consolidation:
v 1 ' 1 ( pw )
2H
t Eoed t Eoed t
1 1 n pw k 2
pw
Eoed t Eoed K w t w
(1 ') E '
where Eoed (1 ')(1 2 ') = constrained modulus of soil skeleton

k Eoed pw
Considering incompressible water: 2 pw
w t t

Theory of consolidation

1D consolidation, considering a constant total stress : 0
t
pw k Eoed
cv 2 pw where cv = consolidation coefficient =
w
t

cv t
T
H2

Page 279
Advanced Computational Geotechnics

PERMEABILITY
Dependence on grain (pore) 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)

Permeability often anisotropic



q x k x q y k y
x y

PERMEABILITY
Dependence on void ratio

Page 280
Advanced Computational Geotechnics

PERMEABILITY

PLAXIS allows consideration of change of permeability with void ratio

k e
log
k0 ck
Default value for ck 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

FEM for consolidation analysis

Soil properties:
Unit weight (sat,unsat)
Stiffness (E,)
Strength (,c)
Permeability (k)
Change of permeability (Ck)

k Eoed
Note: Consolidation coefficient is more or less constant C
v
w
But: k decreases with load (compaction of soil)
Eoed increases with load (stress level)
Use realistic Ck only with advanced models k e
log
k 0 Ck

Page 281
Advanced Computational Geotechnics
FEM for consolidation analysis

Classical formulation
Consolidation analysis based on excess pore pressure (EPP):
pactive psteady pexcess

Assumptions:
Steady state pore pressure is constant in time (horizontal phreatic level or
steady state pore pressure from groundwater flow calculation)
Excess pore pressure can change in time
Fully saturated soil (above and below phreatic level)

Limitation:
Time dependent hydraulic boundary is not possible (variable phreatic level)

FEM for consolidation analysis

Node
Element

Stress
point

Mesh:
Elements: Interpolation of primary variables
Nodes: Primary variables (displacements, pore pressures)
Stress points: Derived variables (strains, stresses, Darcy velocities)

Same order of interpolation in PLAX

18
Page 282
Advanced Computational Geotechnics
FEM for consolidation analysis

Calculations:
Stiffness matrix
Coupling matrix
K v L p f Forces Equilibrium
Pore pressures
Displacements

Flow matrix
Displacements
T dv dp
H pL S q Net flow
dt dt Continuity
Compressibility of water
Transposed coupling matrix

K L v 0 0 v0 f

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

S t H S q q0 q
* *

Solution: Displacements and (excess) pore pressures

FEM for consolidation analysis


Calculations:

K v L p f Equilibrium
dv dp
H pL S q
T
Continuity
dt dt
K L v 0 0 v0 f

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

S t H S q q0 q
* *

H: permeability matrix H ( N )T k ( N ) dV
V
T
n
S: compressibility matrix S N N dV
V
Kw
T
K: stiffness matrix K B D B dV
V

L: coupling matrix T
L B m N dV
V

Solution: Displacements and (excess) pore pressures


Page 283
Advanced Computational Geotechnics
FEM for consolidation analysis

Boundary conditions (mechanical)


Fixities
Loads
Boundary conditions (hydraulic )
Prescribed groundwater head
No flow
Others: drains, wells, infiltration

FEM for consolidation analysis

INITIAL PORE PRESSURE AND EXCESS PORE PRESSURE

Sand

Clay pw

Sand pw0

e.g. t=0 t=1 t = 10 t = 40 t = 100 days

pw pw 0 pw pw 0 steady state pore pressure

Consolidation in classical mode in terms of excess pore pressures


Steady state pore pressures have to be established

Page 284
Advanced Computational Geotechnics

FINITE ELEMENT FORMULATION FOR CONSOLIDATION (5)

Time step
Automatic time stepping is required
Critical time step (smalle steps may cause stress oscillations)

l2
tcritical
Cv
l = element length
= 80 for 15-node triangles
= 40 for 6-node triangles

FEM for consolidation analysis

Calculations:
Consolidation Staged construction > Time interval t
Consolidation Minimum pore pressure > |p-stop|
Consolidation Incremental multipliers > Time increment

Typical: Plastic calculation (staged construction) with undrained materials, followed


by consolidation analysis
Rate loading: Time increment and load increment give loading rate

Page 285
Advanced Computational Geotechnics
FEM for consolidation analysis

Output:
Deformations
Stresses
Excess pore pressure
History curves
(e.g. pore pressure as function of time)

FEM for consolidation analysis

Validation: One-dimensional consolidation

Page 286
Advanced Computational Geotechnics
FEM for consolidation analysis

Validation: One-dimensional consolidation

FEM for consolidation analysis

New features of PLAXIS 2D 2010


PLAXIS 2010 offers two types of consolidation analysis: classical
mode and advanced mode

Page 287
Advanced Computational Geotechnics

TYPES OF CONSOLIDATION ANALYSIS

PLAXIS 2010 offers two types of consolidation analysis: classical


mode and advanced mode
Classical mode Advanced mode
Consolidation in terms of excess Consolidation in terms of total pp
pp (generated in undrained or Fully coupled analysis between
consolidation analyses) deformations and pore pressures
Coupling between deformation and Change in steady state pp,
excess pp excess pp and deformation
Steady state pp generated prior to influence each other
calculation More boundary conditions for flow
Changes in steady state pp may Soil can be saturated, dry or
change excess pp and unsaturated
deformation but not viceversa
Bishop stress is adopted
(semi-coupled)
Soil fully saturated or dry

FEM for consolidation analysis

Conclusions

FEM is quite suitable for 2D and 3D consolidation analysis


2D or 3D coupled consolidation is different from 1D or uncoupled consolidation
PLAXIS has several options for consolidation based on excess pore pressure
Recent development: Fully coupled flow-deformation analysis and unsaturated
soil behaviour

Page 288
Advanced Computational Geotechnics

HKCG 12
1

MO
ODELL
LING OF
O GR
ROUND
DWAT
TER IN
N PLAX
XIS

ee Siew
Dr Le w Wei

Page 289
Advanced Computational Geotechnics

Modelling of Groundwater in PLAXIS

Dr Lee Siew Wei

CONTENTS

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

Page 290
Advanced Computational Geotechnics

Modelling of Groundwater in Plaxis


A. INTRODUCTION

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

Page 291
Advanced Computational Geotechnics

Modelling of Groundwater in Plaxis


B. DEFINITIONS OF POREWATER
PRESSURES IN PLAXIS

A. DEFINITIONS AND MODES

A. Definition of Porewater Pressure Terms in Plaxis

1. Active State Porewater Pressures

2. Steady State Porewater Pressures (Background)

3. Excess Porewater Pressures

B. Calculation Modes (since 2010)

1. Classical (same with previous versions prior 2010)

2. Advanced

3. Flow

Page 292
Advanced Computational Geotechnics

A1.ACTIVE POREWATER PRESSURE

A. In Classical model there three porewater pressure terms

B. Active (Total) = Steady-state + Excess


1. ACTIVE porewater pressures is combination of STEADY-STATE and EXCESS porewater
pressures (see Reference Manual 5.9).

2. Steady-state pre pressures are generated due to water conditions (hydraulic boundaries)
assigned to soil clusters (layers)

3. Excess pore pressures are calculated as a result of undrained or consolidation analysis

A2.STEADY-STATE POREWATER PRESSURE

A. Steady-state porewater pressures (Background PWP)


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

Page 293
Advanced Computational Geotechnics

A3.EXCESS POREWATER PRESSURE

A. Excess pore pressures are generated during calculation:


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

B. CALCULATION MODES

Page 294
Advanced Computational Geotechnics

B1.CLASSICAL MODE
A. Steady-state pore pressures
1. Phreatic lines
2. Steady-state groundwater flow analysis
3. Transient-state groundwater flow analysis =
Steady-state background pore pressure
B. Excess pore pressures
1. Undrained material type in combination with
Plastic calculation
2. Consolidation analysis

INPUT KERNEL

Steady State Excess Porewater Deformation


Note:
Note:
1. Undrained Analysis
1. Hydrostatic
2. Consolidation Analysis
2. GWF calculation
3. Excess (Soil Model)

Active Porewater = Steady-state + Excess

B2.ADVANCED MODE

A. Consolidation analysis
B. Transient groundwater flow analysis

KERNEL

Active Porewater Deformation


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

Page 295
Advanced Computational Geotechnics

B3.FLOW MODE

Flow mode:
Similar to PlaxFlow but with huge improvements in the
kernel (see Galavi, 2010)
All functionalities of PlaxFlow rewritten in PLAXIS code
(new)
Steady state groundwater flow

Transient groundwater flow

All types of boundary conditions

New features in wells and drains

Faster calculation (new)

Modelling of Groundwater in Plaxis


B. GENERATION OF POREWATER
PRESSURES

Page 296
Advanced Computational Geotechnics

GENERATION OF STEADY-STATE PORE PRESSURES

Steady-state (Background) pore pressures can be generated by:

1. Phreatic and Cluster Approach (Hydrostatic)

2. Groundwater Flow Analysis

PHREATIC AND CLUSTER APPROACH

A. Phreatic Level (Reference 5.9.2)


1. General Phreatic Level

2. Cluster Phreatic Level


B. Cluster Pore Pressure Distribution (Reference 5.9.5)
1. Interpolation (Adjacent to clusters or phreatic lines)

2. Cluster Dry

3. User-defined Pore Pressure Distribution

Page 297
Advanced Computational Geotechnics

C1. PHREATIC LINE

a) Porewater pressures are hydrostatic


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

Hydrostatic steady-state pressure along interface

GENERAL PHREATIC LINE

Page 298
Advanced Computational Geotechnics

GENERAL & CLUSTER PHREATIC LINES

Cluster: Dry

Page 299
Advanced Computational Geotechnics

Cluster: User-defined Pore Pressure Distribution

Combination: Phreatic and Cluster Options

Page 300
Advanced Computational Geotechnics

COMBINATION: PHREATIC & CLUSTER OPTIONS

CLUSTER DRY + INTERPOLATION COMBO

Cluster Dry

Interpolated
Case Histories

Page 301
Advanced Computational Geotechnics

GROUNDWATER FLOW ANALYSIS


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

E. Transient-state groundwater flow :


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

Page 302
Advanced Computational Geotechnics

GROUNDWATER FLOW: STEADY STATE

Calculation based on setup of:

1. Boundary conditions:

a) Prescribed water levels (constant)

b) Closed flow boundaries (bottom, axis of symmetry)

c) Wells and drains (constant)

d) Interface elements (on=impermeable, off=permeable)

e) Inflow / outflow (constant)

2. Soil permeabilities

3. Phreatic level in the soil is being calculated for t=

GROUNDWATER FLOW: TRANSIENT-STATE

Transient groundwater flow:


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

2. Flow field changes in time:


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

Page 303
Advanced Computational Geotechnics

Pore pressures steady-state

Steady-state flow
3 28 29 6 9 30 31 2

4 8 11 5
General General

General
16 17

21 26
19 18
20 23 24 27

22 25
13 14 12
15 7 10

0 1

SOME POINTS: FINITE ELEMENT MODELING

1. GWF calculation generally needs finer mesh than deformation analysis


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

Page 304
Advanced Computational Geotechnics

SOME POINTS- FINITE ELEMENT MODELING

1. Qualitative evaluation:
Flow field
Location of phreatic line
2. Quantitative evaluation:
Heads, pore pressures compared to hydrostatic,
Compare with measurements or field experience

Modelling of Groundwater in Plaxis


D. HYDRAULIC MODELS IN PLAXIS

Page 305
Advanced Computational Geotechnics

Groundwater flow flow in unsaturated soil

Water content and permeability in unsaturated zone

k k rel k sat , k rel f h p , S


(h p )
S (h )
n

hp= => pressure head

Groundwater flow flow in unsaturated soil


A. Linear Model

krel
1
hp = -
hp = 0
m
0 hp

1 hp 0 Saturated For numerical stability



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

Page 306
Advanced Computational Geotechnics

Groundwater flow flow in unsaturated soil


Soil Water Characteristic Curve (SWRC or RC)

Groundwater flow flow in unsaturated soil

van Genuchten model

1 g n


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

2
g n 1


g n

g n
krel S Se l 1 1 Se n
g g 1

with S Sres
Se
Ssat Sres

Ssat,Sres: saturated and residual saturation


ga, gn and gl: curve fitting parameters
Page 307
Advanced Computational Geotechnics

Groundwater flow flow in unsaturated soil

Approximate van Genuchten model

1 if hp 0
Linear in Saturation


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

Groundwater flow - material data sets


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

Page 308
Advanced Computational Geotechnics

Groundwater flow - material data sets

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

Groundwater flow - material data sets

Soil classification: Hypres

Hydraulic Properties of
European Soils
Particle distribution:
< 2m
2m - 50m
50m 2mm

5 upper soils data sets


5 lower soil data sets
1 organic soil data set

Page 309
Advanced Computational Geotechnics

Groundwater flow - material data sets

Soil classification: USDA


United States Department
of Agriculture

Particle distribution:
< 2m
2m - 50m
50m 2mm
12 soils data sets
No difference between
upper and lower soils

Groundwater flow - material data sets

Soil classification and Van Genuchten parameters

Relative permeability

Degree of saturation

Page 310
Advanced Computational Geotechnics

Modelling of Groundwater in Plaxis


E. EXAMPLES

Page 311
Advanced Computational Geotechnics

Page 312
Advanced Computational Geotechnics

Page 313
Advanced Computational Geotechnics

Page 314
Advanced Computational Geotechnics

REFERENCES

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

Page 315
Advanced Computational Geotechnics

H
HKCG 8

EXERCIS SE 4
DEWAT
TERING IN EXCAV
E VATIO
ON
Dr Jo
ohnny Cheuk
C

Page 316
Advanced Computational Geotechnics

Excavation and dewatering

EXCAVATION AND DEWATERING

Computational Geotechnics 1

Page 317
Advanced Computational Geotechnics

Excavation and dewatering

2 Computational Geotechnics

Page 318
Advanced Computational Geotechnics

Excavation and dewatering


INTRODUCTION
This example involves the dry construction of an excavation. The excavation is supported by concrete diaphragm
walls. The walls are tied back by pre-stressed ground anchors. The Hardening Soil model is used to model the
soil behaviour. Special attention is focused on the output, which provides us insight in the soil behaviour and its
interaction with structural elements. It is noted that the dry excavation involves a groundwater ow calculation
to generate the new water pressure distribution.

INPUT
The excavation is 20 m wide and 10 m deep. 15 m long concrete diaphragm walls of 0.35 m thickness are used to
retain the surrounding soil. Two rows of ground anchors are used at each wall to support the walls. The upper
anchor has a total length of 14.5 m and an inclination of 33.7o (2:3). The lower anchor is 10 m long and is installed
at an angle of 45o . The excavation is symmetric so only one half of the problem needs to be modelled.

Figure 1: Excavation supported by tie back walls

The relevant part of the soil consists of three distinct layers. From the ground surface to a depth of 3 m there is
a ll of relatively loose ne sandy soil. Underneath the ll, down to a minimum depth of 15 m, there is a more or
less homogeneous layer consisting of dense well graded sand. This layer is particular suitable for the installation
of the ground anchors. In the initial situation there is a horizontal phreatic level at 3 m below the ground surface,
(i.e. at the base of the ll layer) Below the sand layer there is a loam layer which extends to large depth.

Geometry model
The symmetric problem can be modelled with a geometry model of 32 m width and 20 m depth. The proposed
geometry model is given in gure 2. A ground anchor can be modelled by a combination of a node-to-node anchor
and a geogrid (yellow line). The geogrid simulates the grout body whereas the node-to-node anchor simulates
the anchor rod. The diaphragm wall is modelled as a plate. The interfaces around the plate are used to model
soil-structure interaction eects. They are extended under the wall for 1.0 m to allow for sucient exibility and
accurate reaction forces. Interfaces should not be used around the geogrids that represent the grout body. In
general, it is a good habit to extend interfaces around corners of structures in order to allow for sucient freedom
of deformation and to obtain a more accurate stress distribution. When doing so, make sure that the extended
part of the interface is always turned o in the water conditions mode.

Computational Geotechnics 3

Page 319
Advanced Computational Geotechnics

Excavation and dewatering


(10,0)
(0,0) (45,0)
0 1 2

(0,-3) 10 12 3
(45,-3)

(0,-7)
9 13 (19,-9)
(0,-10) 16
8 14 (22,-11)
17 18

19
(17,-14)
(0,-17) 7 11 4
(45,-17)
15
(10,-18)

(14,-11)

(0,-35) (45,-35)
6 5

Figure 2: Geometry model of building pit

Material properties
The soil consists of three distinct layers. The parameters of the dierent layers are shown in table 1. The interfaces
around the wall will be left impermeable in order to block the ow through it. Since the interfaces in the loam
layer below the wall (the extended part of the interfaces) do not inuence the soil behaviour, therefore their
strength is not reduced and the permeability must be changed to permeable. This will be achieved during the
denition of the staged construction phases.

Table 1: . Soil and interface properties.


Parameter Symbol Fill Sand Loam Unit
Material model Material model HSsmall HSsmall HSsmall
Drainage type Drainage type Drained Drained Drained
Unsaturated soil weight unsat 16.0 17.0 17.0 kN/m3
Saturated soil weight sat 20.0 20.0 19.0 kN/m3
ref
Reference secant stiness from triaxial test E50 20.5 103 38.5 103 20.0 103 kN/m2
ref
Reference tangent stiness from oedometer test Eoed 20.5 103 35.0 103 20.0 103 kN/m2
Reference unloading/reloading stiness ref
Eur 61.5 103 115.5 103 60.0 103 kN/m2
Power for stress-dependent stiness m 0.5 0.5 0.7 
Cohesion c' 1.0 1.0 8.0 kN/m2
Friction angle 30.0 34.0 29.0 o

Dilatancy angle 0.0 4.0 0.0 o

Threshold shear strain 0.7 1.0 104 1.0 104 1.5 104 
Reference small-strain shear modulus Gref
0 180.0 103 350.0 103 180.0 103 kN/m2
Advanced parameters Default Default Default
Horizontal permeability kx 1.0 0.5 0.1 m/day
Vertical permeability ky 1.0 0.5 0.1 m/day
Interface strength reduction Rinter 0.65 0.7 Rigid 
Coecient for initial horizontal stress K0 Automatic Automatic Automatic 

4 Computational Geotechnics

Page 320
Advanced Computational Geotechnics

Excavation and dewatering


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

Table 2: Properties of the diaphragm wall


Parameter Symbol Diaphragm wall Unit
Material type Material type
Axial stiness EA 1.2 107 kN/m
Flexural stiness EI 1.2 105 kN/m2 /m
Weight w 8.3 kN/m/m
Poisson's ratio 0.15 

For the properties of the ground anchors, two material data sets are needed: One of the Anchor type (anchor rod)
and one of the Geogrid type (grout body). The Anchor data set contains the properties of the anchor rod and
the Geogrid data set contains the properties of the grout body. The data are listed in tables 3 and 4.

Table 3: Properties of the anchor rod


Parameter Symbol Anchor rod Unit
Material type Material type Elastic
Axial stiness EA 2.5 105 kN
Spacing Ls 2.5 m

Table 4: Property of the grout body


Parameter Symbol Grout Unit
Material type Material type Elastic
Axial stiness EA 1.0 105 kN/m

Mesh generation
For the generation of the mesh it is advisable to set the Global coarseness parameter to Medium. In addition, it
is expected that stress concentrations will occur around the two grout bodies and in the lower part of the wall,
hence local renements are proposed there.
After generating the mesh, continue to the calculation.

Computational Geotechnics 5

Page 321
Advanced Computational Geotechnics

Excavation and dewatering


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

Method 1: Steady-state groundwater ow


In this method a so-called semi-coupled analysis will be performed. This means that the groundwater ow eld
is generated rst and used as input to the deformation analysis. In other words, the groundwater ow will have
an eect on the deformations in the soil, but the deformations in the soil will not change the ow eld. This
assumption is reasonable if the ow eld will not be disturbed by excess pore pressures resulting from undrained
behaviour or by signicant changes in permeability due to large deformations. In this excavation problem indeed
permeabilities are high and undrained behaviour should be of little or no importance.
For this method the Calculation mode should be set to Classical mode in the Select calculation mode window that
appears directly after opening PLAXIS Calculations. If the incorrect mode is chosen one can still change this by
selecting the Calculation mode option from the Tools menu.
All calculation phases are dened as Plastic calculations of the Staged construction type and standard settings
for all other parameters. The instructions given below are limited to a description of how the phases are dened
within the Staged construction mode.

Initial phase
Set the Calculation type to K 0 procedure for calculating the initial stresses.
Press the Dene button on the Parameters tabsheet to dene the initial situation
In Staged construction mode make sure that all soil is activated and all structural elements are deactivated,
then continue to Water conditions mode.
Draw a horizontal phreatic level from (x,y) = (-2,-3) to (20,-3), (30,-3) and (47,-3).
Pore pressures will be generated based on this phreatic line. To do so, make sure the Generate by phreatic
level button is selected.

Press the Water pressures button to view the pore pressures.


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

6 Computational Geotechnics

Page 322
Advanced Computational Geotechnics

Excavation and dewatering


Phase 1: Construction of the diaphragm wall
Construction of the diaphragm wall takes 5 days. Therefore, ll in a Time interval of 5 days on the
Parameters tabsheet.
Go to the phase denition by pressing the Dene button on the Parameters tabsheet.

In Staged construction mode activate the wall and the interface extensions below the wall. The interfaces
along the wall are activated automatically.

Phase 2: First excavation stage


On the Parameters tabsheet select the option Reset displacements to zero as we're not interested in the
displacements caused by the installation of the wall.

Also on the Parameters tabsheet, enter a construction time of 7 days in the Time interval eld.

Go to the phase denition by pressing the Dene button on the Parameters tabsheet.

In Staged construction mode deactivate the rst excavation part.

Press Update to return to the Calculations program.

Phase 3: Prestress rst anchor


Prestressing the rst row of anchors will take 1 day, hence enter a 1 day construction time.

In Stage construction mode of the phase denition activate the upper geotextile representing the grout body
of the rst anchor.

Double-click on the top node-to-node anchor, the properties window for the node-to-node anchor appears.

Select the option Adjust prestress and enter a 120 kN/m prestress force.

Close the properties window and return to the Calculations program.

Phase 4: Second excavation stage and dewatering


The second excavation stage including dewatering will take 10 days, hence enter a 10 days construction time.

In Staged construction mode deactivate the second excavation stage.

Switch to Water conditions mode.

No water ow can occur through a axis of symmetry. Therefore the axis of symmetry must be a closed ow
boundary. To do so, select the Closed boundary button and draw a closed boundary from (x,y) = (0,0)
to (0,-35). Check that the bottom of the geometry is also a closed boundary.

During excavation the water level will be lowered. Due to high permeabilities water will be drawn from
outside the excavation, hence a groundwater ow analysis has to be performed. Therefore, make sure the
Groundwater button is set to Groundwater ow steady state by clicking the down arrow and choosing
the correct option.

The groundwater head boundary conditions needed for the groundwater ow analysis can be applied in a
simple manner by using the general phreatic level. In order to do so, make sure no cluster is selected (for
instance by clicking completely outside the geometry so that the general phreatic line is red) and then draw
a new general phreatic level from (x,y) = (-2,-7) to (20,-7), (30,-3) and (47,-3).

Press Update to return to the Calculations program.

Computational Geotechnics 7

Page 323
Advanced Computational Geotechnics

Excavation and dewatering


Phase 5: Prestress second anchor
Prestressing the second row of anchors will take 1 day, hence enter a 1 day construction time.
In Stage construction mode of the phase denition activate the lower geotextile representing the grout body
of the second anchor.
Double-click on the lower node-to-node anchor, the properties window for the node-to-node anchor appears.

Select the option Adjust prestress and enter a 200 kN/m prestress force.
Close the properties window and continue to Water conditions mode.

The phreatic line should be still the same as in the previous calculation phase and also the option Ground-
water ow steady-state should still be selected.
Return to the Calculations program.

Phase 6: Third excavation stage and dewatering


The third excavation stage including dewatering will take 7 days, hence enter a 7 days construction time.

In Staged construction mode deactivate the third excavation stage.


Switch to Water conditions mode.

Check that both the axis of symmetry and the bottom of the model are closed boundaries.

Make sure the Groundwater button is set to Groundwater ow steady-state .

Draw a new general phreatic level from (x,y) = (-2,-10) to (20,-10), (30,-3) and (47,-3).

Nodes for load displacement curves


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

8 Computational Geotechnics

Page 324
Advanced Computational Geotechnics

Excavation and dewatering


Method 2: Transient groundwater ow
In this method a fully coupled analysis will be performed. This analysis couples transient groundwater ow,
consolidation and deformations implying that the groundwater ow eld, development and dissipation of excess
pore pressures and deformation are calculated simultaneously all inuencing each other. This type of analysis
should be performed if developement of excess pore pressures is expected inuencing the ow eld or when
signicant changes in permeability due to large deformations are likely to occur. In this excavation problem the
main reason to use this analysis is to take into account that the ow eld will not reach a steady-state during
excavations. The addidional eects of coupling the ow eld with undrained behaviour will probably be small as
this project deals with high permeabilities. Note that a fully coupled analysis requires that the calculation type
is Consolidation.
It is possible to re-use the project made for the calculation using the method of steady-state analysis:

In PLAXIS Calculation save the project under a dierent name

From the menu Tools select the option Calculation mode. In the window that now appears select Advanced
mode.

Change the calculation phases according to the description below. Note that only the changes relative to
the steady-state calculation method are mentioned.

Initial phase
No changes have to be made

Phase 1: Construction of the diaphragm wall


Set the calculation type to Consolidation on the Parameters tabsheet.

Phase 2: First excavation stage


Set the calculation type to Consolidation on the Parameters tabsheet.

Phase 3: Prestress rst anchor


Set the calculation type to Consolidation on the Parameters tabsheet.

Phase 4: Second excavation stage and dewatering


Set the calculation type to Consolidation on the Parameters tabsheet.

Dene the staged construction phase and switch to Water conditions mode.

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

Return to the Calculations program.

Computational Geotechnics 9

Page 325
Advanced Computational Geotechnics

Excavation and dewatering


Phase 5: Prestress second anchor
Set the calculation type to Consolidation on the Parameters tabsheet.
Dene the staged construction phase and switch to Water conditions mode.

Though the phreatic level in the excavation doesn't change, the ow eld is not steady-state yet outside the
excavation. Therefore this phase needs transient ow analysis without making further changes.

Make sure the option Groundwater ow transient is selected.


Return to the Calculations program.

Phase 6: Third excavation stage and dewatering


Set the calculation type to Consolidation on the Parameters tabsheet.

Also on the Parameters tabsheet, set the number of Additional steps to 500.
Dene the staged construction phase and switch to Water conditions mode.

Make sure the Groundwater button is set to Groundwater ow transient .


Return to the Calculations program.

Start the calculation

10 Computational Geotechnics

Page 326
Advanced Computational Geotechnics

Excavation and dewatering


OUTPUT
Figure 3 gives the total displacements for the nal phase for both the calculation with steady-state groundwater
ow and the transient groundwater ow.
The excavation using steady-state ow gives a maximum displacements of about 24 mm while excavation using
transient ow gives a maximum displacement of about 23 mm.

Figure 3: Total displacements for the steady state ow analysis (left) and the transient ow analyis (right)

Figure 4 shows the vertical displacements for the nal phase for both calculations. For the displacements behind
the wall the excavation using steady-state analysis clearly gives more vertical displacements over a larger distance
from the excavation than the excavation with transient ow.

Figure 4: Vertical displacements for the steady state ow analysis (left) and the transient ow analyis (right)

The extreme bending moments are about -165 kNm/m and 75 kNm/m for the excavation using steady-state
groundwater ow analysis while the extremen bending moments for the excavation using transient groundwater
ow are about -170 kNm/m and 95 kNm/m.

Figure 6 shows the horizontal displacements of the top of the wall as a function of construction time for both the
excavation using steady-state ow and transient ow.

Computational Geotechnics 11

Page 327
Advanced Computational Geotechnics

Excavation and dewatering

Figure 5: Bending moments in the wall for the steady state ow analysis (left) and the transient ow analyis
(right)

Figure 6: Horizontal wall displacements for the excavation

12 Computational Geotechnics

Page 328
Advanced Computational Geotechnics

HKCG 14
1

UN
NSATU
URATEED SOILS
AND
BARC
CELON
NA BA
ASIC MODEL
M L
Professo
P or Anton
nio Gens
s

Page 329
Advanced Computational Geotechnics

CG14: UNSATURATED SOILS AND BARCELONA


BASIC MODEL

Antonio Gens

Technical University of Catalunya, Barcelona

some of the slides were originally created by:


Vahid Galavi (Plaxis)

Unsaturated soils: outline

Introduction: characteristic problems in unsaturated soils


Suction in unsaturated soils
Flow and consolidation in unsaturated soils
Stress variables
Mechanical behaviour of unsaturated soils
Shear strength
Volume change

The BBM model for unsaturated soils

Page 330
Advanced Computational Geotechnics

Unsaturated soils

Solid

Liquid Gas

V pores Vliquid Vgas Vliquid Vliquid


n Sr 1 Sg
Vtotal Vtotal V pores Vliquid Vgas
Porosity Degree of saturation

Unsaturated soils: failure

Shum Wan Road landslide, Hong Kong Island on August 13th,1995


Photographs from Geotechnical Engineering Office, Hong Kong
Page 331
Advanced Computational Geotechnics

Engineering problems involving unsaturated soils: collapse

Fig. 4
Collapse in Via Luigi Settembrini, Naples (15-09-2001)

Unsaturated soils: a bit of history


1950s, 1960s
Relevance of suction recognized
Interpretation in terms of single effective stress
Late1960s, 1970s
Unsaturated soils as difficult soils, `special soils, regional soils
Late1970s, 1980s
Recognition of need for two stress variables
State surface approach
Late 1980s, 1990s onwards
Large expansion of research
Suction control and measurement
Elastoplastic models
Incorporation into mainstream Soil Mechanics

Page 332
Advanced Computational Geotechnics

Unsaturated soils: outline

Introduction: characteristic problems in unsaturated soils


Suction in unsaturated soils
Flow and consolidation in unsaturated soils
Stress variables
Mechanical behaviour of unsaturated soils
Shear strength
Volume change

The BBM model for unsaturated soils

Suction in unsaturated soils


New additional variable: suction
Water potential, : work required to transport a unit mass from a
reference pool of pure water to the soil water under consideration

c o g z
Matric Osmotic Gas Gravitational

c ( uw ua ) : Matric (capillary) potential


o cm RT : Osmotic potential
g ( ua uatm ) : Gas pressure potential
z w z : Gravitational potential

Review panel (1965)


Page 333
Advanced Computational Geotechnics

Suction in unsaturated soils

Gravitational
potential

Gas pressure
potential

Matric potential

Suction in unsaturated soils

SEMIPERMEABLE
MEMBRANE

ns
os RT
V
SOLUTE
PURE
WATER

Osmotic potential

Page 334
Advanced Computational Geotechnics

Suction in unsaturated soils


New additional variable: suction
Water potential, : work required to transport a unit mass from a
reference pool of pure water to the soil water under consideration

c o g z (Review panel, 1965)


Matric Osmotic Gas Gravitational
Total water potential controls water flow
Water potential affects mechanical behaviour. Not all potential
components have, however, the same effect

s w c :Matric suction w o :Osmotic suction


st s : Total suction
Total suction is directly related to relative humidity (psychrometric law)
In most cases, only matric suction is relevant for mechanical behaviour

Unsaturated soils: outline

Introduction: characteristic problems in unsaturated soils


Suction in unsaturated soils
Flow and consolidation in unsaturated soils
Stress variables
Mechanical behaviour of unsaturated soils
Shear strength
Volume change

The BBM model for unsaturated soils

Page 335
Advanced Computational Geotechnics

Unsaturated soils: flow and consolidation


Saturated soils: equation of continuity (with soil deformation)

n q x q y n
0
t x y t

Unsaturated soils: equation of continuity (with soil deformation)

(n S r ) q x q y
0
t x y (n S r )
t
n (, s ) S r (, s ) q x q y
Sr n 0
t t x y

Unsaturated soils: flow and consolidation

Unsaturated soils: equation of continuity (with soil deformation)

(n S r ) q x q y (n S r )
0
t x y t

n (, s ) S (, s ) q q y
Sr r n x 0
t t x y

Constitutive Retention Darcys


law curve law

Page 336
Advanced Computational Geotechnics

Unsaturated soils: flow and consolidation


Unsaturated soils: retention curve
(also called soil water characteristic curve, SWCC)
Sr f pa pw = f s

Unsaturated soils: flow and consolidation


Unsaturated soils: retention curve

1. Retention curves exhibit hysteresis effects


2. This value of negative water pressure is called the air entry value
for that soil
3. Negative pore pressures can exist in saturated soils

UNSATURATED

SATURATED

SATURATED

Page 337
Advanced Computational Geotechnics

Unsaturated soils: flow and consolidation


Unsaturated soils: retention curve

UNIFORM

(Brooks and Corey, 1964)

Unsaturated soils: flow and consolidation


Retention curve: analytical expressions
Plaxis uses the Van Genuchten expression


1 g n

gn
gn
S ( p ) S residu ( S sat S residu ) 1 g a p

p s ( pa pw ) in m

S sat 1.0 S residu 0.027


g a 2.24 m 1 g n 2.286

An alternative linearized VG model is provided in Plaxis


Parameter values are provided depending on soil classification
Used-defined models are possible
Page 338
Advanced Computational Geotechnics

Unsaturated soils: flow and consolidation


Unsaturated soils: Darcys law

h pw
q k k h y
dy dy w

Relative permeability

k k rel ( S r ) k sat

Fredlund &
Rahardjo (1993)

Unsaturated soils: flow and consolidation


Relative permeability: analytical expressions
Plaxis uses the Van Genuchten expression
2

g n 1

g n g n
S r S residu

k rel ( S r ) ( S e ) g l 1 1 S e g n 1 S e
S sat S residu


S sat 1.0 S residu 0.027
g a 2.24 m 1 g n 2.286
g l 0.0

An alternative linearized VG model is provided in Plaxis


Parameter values are provided depending on soil classification Page 339
Advanced Computational Geotechnics

Unsaturated soils: flow and consolidation


Hydraulic parameters: data set

Standard (Topsoil part of Hypres)


Hypres (International soil
classification system)
USDA (International soil classification
system)
Staring (Dutch soil classification
system)
User-defined

Parameter values are provided depending on soil classification


Hydraulic parameters must be chosen very carefully!

Unsaturated soils: flow and consolidation


Finite element formulation for coupled consolidation analysis
(advanced model only!)
The same shape functions are used for pore water pressure and
displacement.
i 1 i 1
K Q v 0 0 vi fu
T p 0 t H
i t (G q )
Q S t H w p w p

p

k rel k rel
H ( N )T G p ( N )T k w g ( N ) dV
sat sat
k ( N ) dV
H: permeability matrix V
w V
w
T nS dS
S: compressibility matrix S N
K
n
dp
N dV
V w w

K: stiffness matrix T
K B M B dV
V

Q: coupling matrix Q S B m N dV
T

Page 340
Advanced Computational Geotechnics

Unsaturated soils: flow and consolidation

Calculation procedure
The presented system is highly non-linear. The non-linearity occurs
because of suction dependent degree of saturation (Q and S) and the
suction dependent permeability (H).
A fully implicit scheme of integration is used to solve the fully coupled
flow-deformation analysis.
k rel
H ( N )T
sat
H: permeability matrix k ( N ) dV
V
w

T nS dS
S: compressibility matrix S N n N dV
V K w dp w
T
Q: coupling matrix Q S B m N dV
V

Unsaturated soils: outline

Introduction: characteristic problems in unsaturated soils


Suction in unsaturated soils
Flow and consolidation in unsaturated soils
Stress variables
Mechanical behaviour of unsaturated soils
Shear strength
Volume change

The BBM model for unsaturated soils

Page 341
Advanced Computational Geotechnics

A fundamental variable for unsaturated soils: (matric) suction


Matric suction is often associated with capillary phenomena

s ua u w

Intergranular capillary forces

In most cases, only matric suction is relevant for mechanical behaviour

Stress variables for unsaturated soils


It is necessary to use two stress variables to characterize the
behaviour of unsaturated soils
Conventional effective stress concept does not apply!
ua 1 ( s, Sr ) 2 ( s, Sr )
Class I ua (1 0)
Includes net stress and suction (BBM): ua , s
Class II ua 1 ( s)
Class III ua 1 ( s, Sr )
Includes Bishops stress and suction (Plaxis BBM): ua S r s , s

Page 342
Advanced Computational Geotechnics

Stress variables for unsaturated soils


Class I
Includes net stress and suction (BBM): ua , s
Easy representation of conventional stress paths (including laboratory test
paths)
Difficulties in the transition saturated-unsaturated
Hysteresis and hydraulic effects difficult to incorporate
Independent function required to model the increase of strength with suction

Class III
Includes Bishops stress and suction (Plaxis BBM): ua S r s , s
Representation of conventional stress paths not straightforward, sometimes
impossible
No difficulties in the transition saturated-unsaturated (it recovers Terzaghis
effective stress on reaching Sr=1)
Hysteresis and hydraulic effects can be naturally incorporated
The increase of strength with suction results from stress variable definition

Stress variables
Isotropic plane using net stress and suction

: Wetting at constant (net) stress


: Loading a saturated soil
: Drying at constant (net) stress
: Loading at constant suction
: Stress path during a swelling pressure test
Page 343
Advanced Computational Geotechnics

Unsaturated soils: outline

Introduction: characteristic problems in unsaturated soils


Suction in unsaturated soils
Flow and consolidation in unsaturated soils
Stress variables
Mechanical behaviour of unsaturated soils
Shear strength
Volume change

The BBM model for unsaturated soils

Unsaturated soils: features of behaviour


Shear strength

Suction-controlled direct shear cell (Escario and Sez, 1980).


Page 344
Advanced Computational Geotechnics

Behaviour of unsaturated soils: shear strength


Shear strength increases with suction

Moderate suctions

Escario & Sez (1986)

Behaviour of unsaturated soils: shear strength


Shear strength increases with suction: a bilinear relationship

f c ' n pa tan ' s tan b

Variation of
apparent cohesion
and friction with
suction
Fredlund & Rahardjo (1985)

Shortcoming: close to saturation should be equal to '


b
Page 345
Advanced Computational Geotechnics
Behaviour of unsaturated soils: shear strength
Shear strength

s
s Large suctions
s

MATRIC SUCTION, s (kg/cm2) Escario & Juc(1990)

The bilinear relationship is not valid and must be modified


The rate of increase of strength close to saturation must be tan
The increase of strength is not linear but it becomes asymptotic at high
suctions
A single stress variable may account for the strength increase with suction
' pa ( pa pw ) ; ( Sl ) (Bishop stress)

Behaviour of unsaturated soils: shear strength


Bishop (1959) proposal for effective stress:
' ua (ua uw ) ; ( Sr )

Page 346
Advanced Computational Geotechnics

Behaviour of unsaturated soils: shear strength


Bishop (1959) proposal for effective stress: ' ua (ua uw ) ; ( Sr )

Often ( Sr ) Sr leading to Bishop stress: ' ua Sr (ua uw )

Sr

Behaviour of unsaturated soils: shear strength

Assuming (ua 0) and Sr leads to the simplified expression:

' ( S r u w )
f c' n tan ' c' ( n S r uw ) tan ' c' n tan ' S r uw tan '
The variation of shear strength with suction depends on the variation of Sr with
suction

Often Sr is replaced by Se (PLAXIS!)


S r S res Se: effective saturation
Se
S sat S res Sr: degree of saturation
Sres: residual saturation (suction )
Ssat: saturation when (suction=0)

Page 347
Advanced Computational Geotechnics

Behaviour of unsaturated soils: shear strength

Sand (USDA): (Sres=10,5%)

S r pw

S e pw

Unsaturated soils: outline

Introduction: characteristic problems in unsaturated soils


Suction in unsaturated soils
Flow and consolidation in unsaturated soils
Stress variables
Mechanical behaviour of unsaturated soils
Shear strength
Volume change

The BBM model for unsaturated soils

Page 348
Advanced Computational Geotechnics

Behaviour of unsaturated soils: volume change

Suction-controlled oedometer cell

Behaviour of unsaturated soils: consolidation lines

Suction increases the apparent preconsolidation stress


The soil can sustain a higher void ratio at the same stress value

(Oedometer
tests on a
Brazilian
residual soil;
Fig. 5 Lemos, 1998)
Page 349
Advanced Computational Geotechnics
Behaviour of unsaturated soils: volume change
Collapse

Jennings & Burland


(1962)

Two stress variables are required to describe collapse

Behaviour of unsaturated soils: volume change


Volume change behaviour on saturation depends on applied stress level

Page 350
Advanced Computational Geotechnics
Behaviour of unsaturated soils: volume change
Triaxial wetting tests on Ko-consolidated samples of Lower Cromer Till
(Maswoswe, 1985)

swelling collapse

After collapse, the saturated void ratio is recovered


Collapse (compression) is observed on wetting at high stresses, but a
(small) swelling is observed upon wetting at low stresses

Behaviour of unsaturated soils: volume change

Specimens of compacted clay at 90% of Normal Proctor energy and two


different water contents (Escario & Sez, 1973)

Sample A

Sample B

During collapse, volume strain may change sign (it can only be observed in
suction controlled tests)
Page 351
Advanced Computational Geotechnics

Unsaturated soils: outline

Introduction: characteristic problems in unsaturated soils


Suction in unsaturated soils
Flow and consolidation in unsaturated soils
Stress variables
Mechanical behaviour of unsaturated soils
Shear strength
Volume change

The BBM model for unsaturated soils

Elasto plastic model for unsaturated soils


The Barcelona Basic Model
(BBM)
Alonso, E.E., Gens, A. & Josa, A. (1990). A constitutive model for partially saturated soil,
Gotechnique, 40, No3, 405-430.

For partially saturated soils which are slightly or moderately expansive..

Implemented as UDSM in PLAXIS


by Gonzalez & Gens (2008) using
Bishops stress and suction as
stress variables.

Page 352
Advanced Computational Geotechnics

Elasto plastic model for unsaturated soils

The Barcelona Basic Model (BBM)

Net stress: ( pa ) Suction: s ( pa pw )


Isotropic plane Yield curve
S=0 Yield LC
Elastic
VOID RATIO, e

domain

SUCTION, s
S3
S3
S2 S2
S1
S1
S=0
MEAN NET STRESS, p MEAN NET STRESS, p

Elasto plastic model for unsaturated soils


The Barcelona Basic Model (BBM)

Yield curve Yield curve


LC1 LC2
Elastic
domain Loading
SUCTION, s

p
o 1
L
S1

C
Collapse

p p
*
o 1
*
o 2 MEAN NET STRESS, p

Isotropic plane
Page 353
Advanced Computational Geotechnics

Elasto plastic model for unsaturated soils


The Barcelona Basic Model (BBM)

Elastic Yield curve


LC LCB LCC
domain

A B C

SUCTION, s
SUCTION, s

plastic
compression
elastic
swelling

C B A vol
p p p
*
o
*
o B
*
o C compression swelling

MEAN NET STRESS, p


Isotropic plane

Elasto plastic model for unsaturated soils


The Barcelona Basic Model (BBM)
s SUCTION,

q
LC
s CSL (s)

Elastic CSL (s=0)


domain
s
s=0
po* po po* po
MEAN NET STRESS, p MEAN NET STRESS, p

Isotropic plane Deviatoric plane

Page 354
Advanced Computational Geotechnics

Elasto plastic model for unsaturated soils


The Barcelona Basic Model (BBM)

Three-dimensional view

Unsaturated soils

FEATURES OF BEHAVIOUR
Suction increases the apparent preconsolidation pressure (yield
stress) and (often) soil stiffness
Volume change behaviour depends on stress level. Swelling or
compression (collapse) may occur depending on applied load
Collapse behaviour
After collapse soil lies on saturated consolidation line
Volume change reversal may occur during collapse
Volume change behaviour is path independent only for a certain
class of stress paths
Shear strength increases with suction

Page 355
Advanced Computational Geotechnics

Elasto plastic model for unsaturated soils: BBM

Yield surface
g
2

f 3 J 2 p ps P0 p
2

g 30
Cam-clay ellipse (other choices are possible!)
g is function of lode angle () and J is the second deviatoric stress tensor.

Cohesion increases linearly with suction (not


really required in Plaxis implementation!):
ps k s S
ks is an input parameter.

53

Elasto plastic model for unsaturated soils: BBM

Yield surface: LC curve


Preconsolidation stress varies with suction:
*0 *
P * *s *
P0 Pr 0

Pr
Pr : a reference mean stress (fitting LC curve).
0* : modified compression index of saturated soil.
* : modified swelling index of (un)saturated soil.

*s *0 1 r e S r

r & : input parameter (fitting LC curve).

Page 356 54
Advanced Computational Geotechnics

Elasto plastic model for unsaturated soils: BBM

Plastic potential
g 2
2

g 3J
2
p ps P0 p
g 30
is used to obtain K0 path for normally consolidated soils.

1
M M 9 M 3 *
1 *
96 M 0

Hardening law:
P0
dP0 d vp
0
* *

55

Elasto plastic model for unsaturated soils: BBM

Elastic behaviour:
The elastic behaviour of the model is the same as Modified Cam-Clay
model:


d ve , p dp ; G
p
: elastic stiffness due to mean stress (input parameter) .
In addition to that, change in suction may produce volumetric elastic
strain according to (not really required in Plaxis implementation!):
s
d ve , s dS
3 s patm

s : elastic stiffness due to suction (input parameter) .

Page 357 56
Advanced Computational Geotechnics

Elasto plastic model for unsaturated soils

BBM: Triaxial tests

70 60
C3
s=0 kPa s=100 kPa s=200 kPa s=0 kPa s=100 kPa s=200 kPa
60 50
C2
50
40
CSL

q (kPa)
40
q (kPa)

30
30
C1
20
20

10
10
A B2 B3
0 0
0 0.1 0.2 0.3 0.4 0.5 0 20 40 60 80
Axial strain p' (kPa)

250 2.3
s=0 kPa s=100 kPa s=200 kPa
B3 C3 2.25
200
A
2.2
150 s=0 kPa s=100 kPa s=200 kPa
B2
s (kPa)

=1+e
B3
2.15
B2
100 C2
2.1

50 C3
2.05 C2
C1
A C1
0 2
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80
q (kPa) p' (kPa)

Gonzalez (2008)

57

Elasto plastic model for unsaturated soils

BBM: drying-wetting test

125 2.3
YLD A
YLD C 2.2
C NCL 0
100 B
2.1 NCL 1
A
75 2.0 B
s (kPa)

v=1+e

1.9 C
50 1.8
1.7
25
1.6 D
A D 1.5
0
0 50 100 150 200 250 300
0 50 100 150 200 250 300
p' (kPa)
p' (kPa)

Gonzalez (2008)
Page 358
Advanced Computational Geotechnics

Elasto plastic model for unsaturated soils


The Barcelona Basic Model Capable of reproducing
(BBM) main features of
unsaturated soil
behaviour
Provides a consistent
framework for an
integrated understanding
of unsaturated soil
behaviour
Complete/consistent but
based on drastically
simplified assumptions
Compatible with classical
models of saturated soils
Lots of simplifications!

Page 359
Advanced Computational Geotechnics

HKCG 15
1

IN L STRESSES
NITIAL S
AND
S
SLOPEE STA
ABILITY
Y ANA
ALYSIS
S
Pro
ofessor Helmut Schweiger

Page 360
Advanced Computational Geotechnics

CG15-A INITIAL STRESSES

AND PHI-C REDUCTION - PART 1

Helmut F. Schweiger

Graz University of Technology, Austria

most slides were originally created by:


Dennis Waterman (Plaxis bv)
William Cheang (Plaxis Asia)

Outline

Initial stresses
general
Ko procedure
gravity loading
special cases
Phi-c reduction
safety factor
safety factor analysis
examples
final advice

Page 361
Advanced Computational Geotechnics

INITIAL STRESSES

Initial stresses represent the equilibrium state of the


undisturbed soil and consist of:
Soil weight
Loading history

In Plaxis two possibilities exist:


K0 procedure
Gravity loading

INITIAL STRESSES

K0-procedure

Generation of initial stresses as first step in calculation procedure

Needed: coefficient for lateral earth pressure K0

Disadvantage:
No equilibrium for inclined surface (a nil-step can be used)

Advantage:
No displacements are generated, only stresses
Can take into account initial OCR or POP

'h 'v K0

Page 362
Advanced Computational Geotechnics

INITIAL STRESSES

Gravity loading

Calculation of initial stresses by gravity loading

Disadvantage
Non-physical displacements are created
Difficult to influence K0 value
Difficult to introduce initial OCR or POP


For 1D compression: 'h 'v so K0
1 1
Maximum value of K0 1 ( 0.5)

Advantage
Equilibrium satisfied in all case

Attention: ignore undrained


> dont produce excess pore water pressures

INITIAL STRESSES

Cases where gravity loading should be used instead of K0-procedure:

Page 363
Advanced Computational Geotechnics

INITIAL STRESSES

Special cases


Gravity loading needed due to geometry, but K0
1

Gravity loading needed due to geometry,


but initial OCR or POP required

K0 procedure using Phase 1: Phase 2:


desired K0, OCR, POP Excavate excess soil Reset displacements

Check K0, OCR !

INITIAL STRESSES

Special cases
For complex initial situations like inner city building projects it may
be needed to use several calculation phases to model the current
situation before starting the calculation for the actual project.

existing buildings

our
our project project

initial phase 1 phase 2 our project

reset displacements to 0

Page 364
Advanced Computational Geotechnics

PHI-C-REDUCTION

FACTOR OF SAFETY
Many possible definitions

available soil resistance


1.8
mobilized soil resistance
failure load
5.9
working load

PLAXIS: safety factor on soil resistance

PHI-C-REDUCTION

Strength reduction method: Phi/c reduction


Same numerical tool as for serviceability design
Automatically detects most critical failure mechanism

Page 365
Advanced Computational Geotechnics

PHI-C-REDUCTION

Reduction of strength parameters c and tan() until failure is reached.


The factor of safety is then defined as:

c tan
Msf Lowered incrementally

creduced tanreduced

PHI-C-REDUCTION

Calculation procedure:
Create a Phi/c reduction (<V2010) or Safety (V2010) phase
Accept the default increment for Msf = 0.1 from the multiplier tab-sheet
Calculate
Carefully examine Msf vs. displacement curve in Plaxis Curves to assure
that failure is indeed reached
If so, the value of Msf is assumed to be the factor of safety on soil
resistance

Notes:
In order to check failure, select a control point within the (expected)
failing body
Use sufficient number of load steps
Choose elasto-plastic behaviour for wall, anchors and geotextiles with
realistic full plastic values in order to prevent excessive high structural
forces

Page 366
Advanced Computational Geotechnics

PHI-C-REDUCTION

Main advantages of Phi-c-reduction method:


Requires no a-priori assumptions on the failure mechanism

Critical surface is found automatically as slope failure occurs naturally


through the zones due to insufficient shear strength to resist shear
stresses.

No requirement of assumptions on the inter-slice shear force distribution

Applicable to complex conditions

Information such as stresses, movements and pore-pressures and


numerical tool as for deformation analysis

Powerful alternative approach

PHI-C-REDUCTION

Influence of mesh discretization

15-noded elements Factor of Safety

5 1.90
11 1.62
(very coarse) 38 1.52
(coarse) 82 1.51
(medium) 170 1.50
(fine) 414 1.45
(very fine) 871 1.43
3733 1.43
15749 1.43

Page 367
Advanced Computational Geotechnics

PHI-C-REDUCTION
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 0.3 0.6 0.9 1.2
0.0
displacement
displacement

1. Arrows of 2. Shadings of 3. Shadings of


incremental incremental incremental shear
displacements displacements strains

PHI-C-REDUCTION

Example 1 stability of a drained slope


Sf=1.57
1.6
= = 16 kN/m
16 kN/m3
=20
= 20
c=5 kPa
c= 5 kPa
1.4
Sum-Msf

1.2

1
0 0.025 0.05 0.075 0.1
Total displacement node X

Bishop analysis: FoS = 1.54


Plaxis analysis: FoS = 1.57

Page 368
Advanced Computational Geotechnics

PHI-C-REDUCTION

Example 2, undrained stability of a slope:

H = 12m 2
D 1 cu = 50 kPa
cu = 100 kPa Plaxis:
F = 1,35

cu 50 D
Stability charts: F N0 6.6 1.38 , N0 f ( , ) (Taylor,1948)
Pd 12 20 H

PHI-C-REDUCTION

Example 2, undrained stability of a slope:


Automatic detection of most critical shear surface:

Plaxis:
F = 1,34
cu = 50 kPa

cu = 60 kPa

Page 369
Advanced Computational Geotechnics

PHI-C-REDUCTION

Example 2, undrained stability of a slope:


Automatic detection of most critical shear surface:

Plaxis:
F = 1,19
cu = 50 kPa

cu = 50 kPa

PHI-C-REDUCTION

Example 3: 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-15n) Model Slope G3D cu2/cu1=0.8 (3436 elements-15n)
1.0

FOS= 1.428 (1.471, 1.473) FOS= 1.384 (1.451) 0.8

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-15n) Model Slope G3F cu2/cu1=0.5 (3436 elements-15n) 1 Griffiths & Lane (1999)
FOS= 1.319(1.401, 1.404) FOS= 1.112(1.251) 2 Bishop & Morgenstern (1960)
3.Taylor (1937) (green line)
4.Janbu

Model Slope G3E cu2/cu1=0.4 (3436 elements-15n) Model Slope G3c cu2/cu1=0.2 (3436 elements-15n) :
FOS= 0.903(1.051) FOS= 0.470 (0.591, 1.304)

Page 370
Advanced Computational Geotechnics

PHI-C-REDUCTION

Some final advice:


Always inspect the incremental displacements or strains as
computed in the last load step to make sure that failure is
reached.

The mesh used in the calculation needs to be sufficiently fine.

Mesh: Refine and redo the phi-c analysis until the factor of
safety remains constant upon further refinement of the mesh.

Always use the arc-length time stepping procedure within the


Phi-C reduction (default)

Use a small tolerated error (maximum should be the default


error of 1% )

Beware of three dimensional effects

Page 371
Advanced Computational Geotechnics

CG15-B
SLOPE STABILITY
(considering rainfall infiltration)

Helmut F. Schweiger, Indra Noer Hamdhan


Computational Geotechnics Group
Institute for Soil Mechanics and Foundation Engineering
Graz University of Technology

PLAXIS HONG KONG COURSE, 6-9 November 2012

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

Rainfall is one of the main factors


causing slope failures.

Rainfall induced slope failures are


common problems in steep
residual soil slopes in the tropics.

Residual soils frequently exist in


an unsaturated state.

The infiltration of rainfall will


increase the groundwater level
and water pressure and decrease
matric suction of unsaturated
soils.

Slope failures caused by heavy rainfall


(Indonesia, 2010)
PLAXIS HONG KONG COURSE, 6-9 November 2012 2

Page 372
1
Advanced Computational Geotechnics

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

Hydraulic characteristics such as saturated coefficient of permeability,


initial degree of saturation, intensity and duration of rainfall are
important parameters in the analysis of slope stability considering rain
infiltration.

The main objective of this study is to evaluate the effects of hydraulic


characteristics of soils (hydraulic conductivity and initial degree of
saturation) and rainfall in slope stability calculations performed with the
finite element method in a fully coupled flow-deformation analysis.

PLAXIS HONG KONG COURSE, 6-9 November 2012 3

SLOPE STABILITY
FOS OF SLOPE DURING RAINFALL INFILTRATION

Hydraulic Characteristics
Van Genuchten (1980) presented a set of closed-form equations of hydraulic
characteristics for unsaturated soils which is based on the capillary model of
Mualem (1976). The Van Genuchten model introduced the relation between
saturation and suction pore pressure head (p):


gc
S

Sr

Sr uw

1
ga

gn

e
s
i
d
u

s
a
t

e
s
i
d
u

p
g

Sresidu: residual degree of saturation of the soil that describes the part of water
that remains in the soil even at high suction heads.

Ssat: degree of saturation of the soil when the pores are filled with water.

ga, gn and gc are empirical parameters, and it is assumed that:


1
gn n
gc


PLAXIS HONG KONG COURSE, 6-9 November 2012 4

Page 373
2
Advanced Computational Geotechnics

SLOPE STABILITY
FOS OF SLOPE DURING RAINFALL INFILTRATION

Hydraulic Characteristics

The effective degree of saturation (Se) is obtained as:

S Ss
S rS
S

e
s
i
d
u

a t

r
e
s
i
d
u
The relative permeability based on Mualem - Van Genuchten is:

2
gng
1

gngne

kr
S

1
1
S

n
gl

1

e
l

where gl is an empirical parameter.

PLAXIS HONG KONG COURSE, 6-9 November 2012 5

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

SHEAR STRENGTH OF UNSATURATED SOILS


Bishop (1959) has modified Terzaghis classical effective stress theory and
proposed the following expression for the effective stress of unsaturated soils:

= ( - ua) + (ua - uw)


and : effective and total stress respectively
ua : pore air pressure
uw : pore water pressure
(ua uw) : matric suction
: matric suction coefficient, varies from 0 to 1 covering the range from dry to
fully saturated conditions.

Assuming (ua 0) and S leads to the simplified expression:

= ( - S uw)

PLAXIS HONG KONG COURSE, 6-9 November 2012 6

Page 374
3
Advanced Computational Geotechnics

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

SIMPLE BENCHMARK EXAMPLE


Homogeneous slope.
The international soil classification system USDA series is used for determining
the hydraulic parameters for the Van Genuchten model.
Mohr Coulomb failure criterion.
The height of the slope is 10 m and the gradient (horizontal to vertical) is 2:1.

20m 20m 20m

10m

15m

PLAXIS HONG KONG COURSE, 6-9 November 2012 7

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

SOIL PARAMETERS FOR MOHR COULOMB MODEL

Description Symbol Unit Value

Unit weight kN/m3 20

Elastic modulus E kPa 7 500

Poisson's ratio - 0.35

Effective cohesion c' kPa 20

Effective friction angle ' o 20

PLAXIS HONG KONG COURSE, 6-9 November 2012 8

Page 375
4
Advanced Computational Geotechnics

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION
PARAMETERS FOR VAN GENUCHTEN MODEL
Soil Ksat (m/s)ga (1/m) gn (-) gl (-)
Sand 8.25E-05 14.50 2.68 0.50
Hydraulic characteristics
Loamy Sand 4.05E-05 12.40 2.28 0.50 for different soils based on
Sandy Loam 1.23E-05 7.50 1.89 0.50 USDA series with Van
Genuchten parameters
Loam 2.89E-06 3.60 1.56 0.50
Silt 6.94E-06 1.60 1.37 0.50
Silty Loam 1.25E-06 2.00 1.41 0.50
Sandy Clay Loam 3.63E-06 5.90 1.48 0.50
Clayey Loam 7.22E-06 1.90 1.31 0.50
Silty Clay Loam 1.94E-06 1.00 1.23 0.50
Sandy Clay 3.33E-06 2.70 1.23 0.50
Silty Clay 5.50E-07 0.50 1.09 0.50
Clay 5.50E-08 0.80 1.09 0.50 Source: Plaxis 2D Reference Manual 2010
PLAXIS HONG KONG COURSE, 6-9 November 2012 9

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

INITIAL CONDITIONS (kN/m2)


100 kN/m2
100.00
90.00

Initial suction in the model is 80.00


70.00

assumed to increase linearly 0 kN/m2


60.00
50.00

above ground water level 40.00


30.00
until ground surface. 20.00
10.00
0.00

0 kN/m2 -10.00

1.2

1.0

0.8
Saturation

0.6

0.4
Soil Water Characteristic
Clay
0.2
Sandy Clay
Silt
Curve (SWCC)
Loamy Sand

0.0
0.1 1 10 100 1000

Soil Suction (kPa)

PLAXIS HONG KONG COURSE, 6-9 November 2012 10

Page 376
5
Advanced Computational Geotechnics

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

MODELLING ASSUMPTIONS

Initial ground water level horizontal at level of the toe of the slope.
Rainfall intensity: 10 mm/hour lasting 3 days (72 hours).
Minimum and the maximum pore pressure head: -0.1 m (min) and 0.1m (max).
Left, right and lower boundary are impervious boundaries.

Rainfall 10 mm/hour

General

PLAXIS HONG KONG COURSE, 6-9 November 2012 11

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

INITIAL CONDITIONS
Initial degree of saturation for the four different hydraulic parameters
(suction is the same)
85.43% 60.73%

100.00% 100.00%

(%)
(a) Clay (b) Sandy Clay 100.00
95.00
90.00
85,00
80.00
40.39% 14.08% 75,00
70.00
65,00
60.00
55,00
100.00% 100.00% 50.00
45,00
40.00
35,00
30.00
25,00
20.00
15,00
10.00
5.00
(c) Silt (d) Loamy Sand 0.00

saturation

PLAXIS HONG KONG COURSE, 6-9 November 2012 12

Page 377
6
Advanced Computational Geotechnics

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

RESULTS CLAY
Before rain infiltration: suction
FOS = 1.709 (kN/m2)
100.00
90.00
80.00
70.00
60.00
50.00
40.00
30.00

After 3 hours rain infiltration: 20.00


10.00

FOS = 1.709 0.00


-10.00

saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00

After 6 hours rain infiltration: 65,00


60.00

FOS = 1.708 55,00


50.00
45,00
40.00
35,00
30.00
25,00
20.00
15,00
10.00
5.00
0.00
Failure mechanism Distribution of suction Degree of saturation

PLAXIS HONG KONG COURSE, 6-9 November 2012 13

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

RESULTS CLAY
After 48 hours rain infiltration: suction
FOS = 1.694 (kN/m2)
100.00
90.00
80.00
70.00
60.00
50.00
40.00
30.00

After 60 hours rain infiltration: 20.00


10.00

FOS = 1.692 0.00


-10.00

saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
After 72 hours rain infiltration: 65,00
60.00

FOS = 1.686 55,00


50.00
45,00
40.00
35,00
30.00
25,00
20.00
15,00
10.00
5.00
0.00
Failure mechanism Distribution of suction Degree of saturation

PLAXIS HONG KONG COURSE, 6-9 November 2012 14

Page 378
7
Advanced Computational Geotechnics

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

RESULTS SANDY CLAY


Before rain infiltration: suction
FOS = 1.680 (kN/m2)
100.00
90.00
80.00
70.00
60.00
50.00
40.00
30.00

After 3 hours rain infiltration: 20.00


10.00

FOS = 1.679 0.00


-10.00

saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
After 6 hours rain infiltration: 65,00
60.00

FOS = 1.678 55,00


50.00
45,00
40.00
35,00
30.00
25,00
20.00
15,00
10.00
5.00
0.00
Failure mechanism Distribution of suction Degree of saturation

PLAXIS HONG KONG COURSE, 6-9 November 2012 15

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

RESULTS SANDY CLAY


After 48 hours rain infiltration: suction
FOS = 1.644 (kN/m2)
100.00
90.00
80.00
70.00
60.00
50.00
40.00
30.00

After 60 hours rain infiltration: 20.00


10.00

FOS = 1.634 0.00


-10.00

saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
After 72 hours rain infiltration: 65,00
60.00

FOS = 1.623 55,00


50.00
45,00
40.00
35,00
30.00
25,00
20.00
15,00
10.00
5.00
0.00
Failure mechanism Distribution of suction Degree of saturation

PLAXIS HONG KONG COURSE, 6-9 November 2012 16

Page 379
8
Advanced Computational Geotechnics

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

RESULTS SANDY SILT


Before rain infiltration: suction
FOS = 1.656 (kN/m2)
100.00
90.00
80.00
70.00
60.00
50.00
40.00
30.00

After 3 hours rain infiltration: 20.00


10.00

FOS = 1.649 0.00


-10.00

saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00

After 6 hours rain infiltration: 65,00


60.00

FOS = 1.644 55,00


50.00
45,00
40.00
35,00
30.00
25,00
20.00
15,00
10.00
5.00
0.00
Failure mechanism Distribution of suction Degree of saturation

PLAXIS HONG KONG COURSE, 6-9 November 2012 17

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

RESULTS SANDY SILT


After 48 hours rain infiltration: suction
FOS = 1.536 (kN/m2)
100.00
90.00
80.00
70.00
60.00
50.00
40.00
30.00

After 60 hours rain infiltration: 20.00


10.00

FOS = 1.492 0.00


-10.00

saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
After 72 hours rain infiltration: 65,00
60.00

FOS = 1.423 55,00


50.00
45,00
40.00
35,00
30.00
25,00
20.00
15,00
10.00
5.00
0.00
Failure mechanism Distribution of suction Degree of saturation

PLAXIS HONG KONG COURSE, 6-9 November 2012 18

Page 380
9
Advanced Computational Geotechnics

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

RESULTS LOAMY SAND


Before rain infiltration: suction
FOS = 1.603 (kN/m2)
100.00
90.00
80.00
70.00
60.00
50.00
40.00
30.00

After 3 hours rain infiltration: 20.00


10.00

FOS = 1.596 0.00


-10.00

saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00

After 6 hours rain infiltration: 65,00


60.00

FOS = 1.592 55,00


50.00
45,00
40.00
35,00
30.00
25,00
20.00
15,00
10.00
5.00
0.00
Failure mechanism Distribution of suction Degree of saturation

PLAXIS HONG KONG COURSE, 6-9 November 2012 19

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION

RESULTS LOAMY SAND


After 48 hours rain infiltration: suction
FOS = 1.502 (kN/m2)
100.00
90.00
80.00
70.00
60.00
50.00
40.00
30.00

After 60 hours rain infiltration: 20.00


10.00

FOS = 1.344 0.00


-10.00

saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
After 72 hours rain infiltration: 65,00
60.00

FOS = 1.229 55,00


50.00
45,00
40.00
35,00
30.00
25,00
20.00
15,00
10.00
5.00
0.00
Failure mechanism Distribution of suction Degree of saturation

PLAXIS HONG KONG COURSE, 6-9 November 2012 20

Page 381
10
Advanced Computational Geotechnics

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION - SUMMARY

Clay

Sandy
Clay

Silt

Loamy
Sand

Initial degree of saturation Distribution of suction Degree of saturation after


after 72 hours infiltration 72 hours infiltration
PLAXIS HONG KONG COURSE, 6-9 November 2012 21

SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION - SUMMARY

FOS Analysis Slope Stability


Time during Infiltration (R = 10 mm/hr)
(hours) Clay Sandy Silt Loamy 1.8
Clay Sand
0 1.709 1.680 1.656 1.603
1.7
3 1.709 1.679 1.649 1.596
6 1.708 1.678 1.644 1.592
1.6
9 1.705 1.674 1.637 1.590
12 1.704 1.673 1.634 1.586
FOS

1.5
18 1.702 1.667 1.617 1.582
24 1.699 1.662 1.602 1.577
1.4
36 1.699 1.653 1.573 1.550
Clay
48 1.694 1.644 1.536 1.502 Sandy Clay
1.3 Silt
60 1.692 1.634 1.492 1.344 Loamy Sand
72 1.686 1.623 1.423 1.229
1.2
0 20 40 60 80

time (hours)

PLAXIS HONG KONG COURSE, 6-9 November 2012 22

Page 382
11
Advanced Computational Geotechnics

SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST

The site area is located on


the intake canal of the
Dagangpo second-level
pumping station in
Zaoyang, Hubei, China.
31m
The site is in a semi-arid
area with an average
annual rainfall of about
800mm, and 70% of the
annual rainfall is distributed
from May to September.

The area has a uniform slope angle of 22o and a uniform slope height of 11m.
There was a 1 m wide berm at the middle of the slope. The slope surface was well
grassed, but no trees were present.

PLAXIS HONG KONG COURSE, 6-9 November 2012 23

SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST

Artificial Rainfall Simulations


Rainfall was artificially produced using a
specially designed sprinkler system. The
sprinkler system comprised a pump, a main
water supply pipe, five branches and 35
sprinkler heads. The system could produce
three levels of rainfall intensity: 3, 6 and 9
mm/h.
Two rainfall events were simulated during
the one month monitoring period, from 13
August to 12 September 2001, with
average daily rainfall of 62 mm.
90
Average daily rainfall (mm)

80
70
60
50
40
30
20
10
0
13-Aug
14-Aug
15-Aug
16-Aug
17-Aug
18-Aug
19-Aug
20-Aug
21-Aug
22-Aug
23-Aug
24-Aug
25-Aug
26-Aug
27-Aug
28-Aug
29-Aug
30-Aug
31-Aug
1-Sep
2-Sep
3-Sep
4-Sep
5-Sep
6-Sep
7-Sep
8-Sep
9-Sep
10-Sep
11-Sep
12-Sep

Date (2001)

PLAXIS HONG KONG COURSE, 6-9 November 2012 24

Page 383
12
Advanced Computational Geotechnics

SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST

FE-Model
Homogeneous slope with angle of 22, berm in the middle of the slope.
The international soil classification system USDA series is used for determining
the hydraulic parameters for the Van Genuchten model.
Mohr Coulomb failure criterion in terms of Bishops effective stress .

1m berm
11m

PLAXIS HONG KONG COURSE, 6-9 November 2012 25

SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST

Soil Parameters:
Description Symbol Unit Value

Unit weight kN/m3 15.6

Young modulus E (kPa) 50000

Poisson's ratio (-) 0.35

Cohesion c' (kPa) 16.7

Friction angle (o) 28.7

Hydraulic Characteristics for Clay soils (USDA series with Van


Genuchten parameters) :
Texture Ksat (m/s) ga (1/m) gn (-) gl (-)

Clay 5.50E-08 0.80 1.09 0.50

Source: Zhan, L. T., Ng, C. W. W., and Fredlund, D. G. (2007). Field study of rainfall infiltration into a grassed
unsaturated expansive soil slope. Can. Geotech. J.44, 392-408.

PLAXIS HONG KONG COURSE, 6-9 November 2012 26

Page 384
13
Advanced Computational Geotechnics

SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST

Calculation in PLAXIS is performed using fully coupled flow-deformation analysis


based on total pore pressures with 2 periods of rain infiltration from 13 August to
12 September 2001 (31 days). The rain infiltration is applied on the entire ground
surface. The steady state condition of the ground water table is as shown.

Period 1 Period 2
90

Average daily rainfall (mm)


80
Rainfall 70
60
50
40
30
20
10
0

13-Aug
14-Aug
15-Aug
16-Aug
17-Aug
18-Aug
19-Aug
20-Aug
21-Aug
22-Aug
23-Aug
24-Aug
25-Aug
26-Aug
27-Aug
28-Aug
29-Aug
30-Aug
31-Aug
1-Sep
2-Sep
3-Sep
4-Sep
5-Sep
6-Sep
7-Sep
8-Sep
9-Sep
10-Sep
11-Sep
12-Sep
GWT
Date (2001)

PLAXIS HONG KONG COURSE, 6-9 November 2012 27

SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST
90
Average daily rainfall (mm)

80
70
60
50
40
30
20
10

Soil suction :
0
13-Aug
14-Aug
15-Aug
16-Aug
17-Aug
18-Aug
19-Aug
20-Aug
21-Aug
22-Aug
23-Aug
24-Aug
25-Aug
26-Aug
27-Aug
28-Aug
29-Aug
30-Aug
31-Aug
1-Sep
2-Sep
3-Sep
4-Sep
5-Sep
6-Sep
7-Sep
8-Sep
9-Sep
10-Sep
11-Sep
12-Sep

Date (2001)
68 kPa

44 kPa

24 kPa

0 kPa 5th day (17th Aug 2001) 9th day (21st Aug 2001) 13th day (25th Aug 2001)
Degree of saturation:
100%

95%

90%

85%
PLAXIS HONG KONG COURSE, 6-9 November 2012 28

Page 385
14
Advanced Computational Geotechnics

SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST
90

Average daily rainfall (mm)


80
70
60
50
40
30
20
10
0

Soil suction :

13-Aug
14-Aug
15-Aug
16-Aug
17-Aug
18-Aug
19-Aug
20-Aug
21-Aug
22-Aug
23-Aug
24-Aug
25-Aug
26-Aug
27-Aug
28-Aug
29-Aug
30-Aug
31-Aug
1-Sep
2-Sep
3-Sep
4-Sep
5-Sep
6-Sep
7-Sep
8-Sep
9-Sep
10-Sep
11-Sep
12-Sep
Date (2001)
68 kPa

44 kPa

24 kPa

0 kPa 29th day (10st Sept 2001) 31st day (12th Sept 2001)
26th day (7th Sept 2001)
Degree of saturation:
100%

95%

90%

85%
PLAXIS HONG KONG COURSE, 6-9 November 2012 29

SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST

Calculated factor of safety (FOS):


time (day) FOS
13-Aug-01 2.30
17-Aug-01 2.30
18-Aug-01 2.28
19-Aug-01 2.27
20-Aug-01 2.17
21-Aug-01 2.12
22-Aug-01 2.06
23-Aug-01 2.01
24-Aug-01 1.96
25-Aug-01 1.92
7-Sep-01 2.25
8-Sep-01 2.23
9-Sep-01 2.10
10-Sep-01 2.05
12-Sep-01 2.19

PLAXIS HONG KONG COURSE, 6-9 November 2012 30

Page 386
15
Advanced Computational Geotechnics

SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST - RESULTS

R2--1.2

R2--1.6

Generally, the comparison of


measured and calculated soil
suction and pore water
pressures is very reasonable.

R2-TC-1.6

PLAXIS HONG KONG COURSE, 6-9 November 2012 31

SLOPE STABILITY
CONCLUSIONS

Safety factors from FE slope stability analysis compare well with


factors obtained from limit equilibrium analysis

Influence of element type and discretization is significantly larger than


for serviceability limit state analysis

The stability of an unsaturated slope will be affected by the


distribution of negative pore water pressures (suction). Slope stability
increases when the shear strength contributed by matric suction is
taken into account.

During the time of rain infiltration, suction decreases and thus the
FOS of the slope reduces, whereas (as expected) the reduction is
faster for soils with high permeability than for soils with low
permeability.

PLAXIS HONG KONG COURSE, 6-9 November 2012 32

Page 387
16
Advanced Computational Geotechnics

SLOPE STABILITY

References for field test


Ng, C. W. W., Springman, S. M. and Alonso, E. E. (2008). Monitoring the
performance of unsaturated soil slopes Geotech Geol Eng 26, 799-816.

Ng, C. W. W., Zhan, L. T., Bao, C. G., Fredlund, D. G. and Gong, B. W. (2003).
Performace of an unsaturated expansive soil slope subjected to artificial rainfall
infiltration. Geotechnique 53, No. 2, 143-157.

Zhan, L. T., Ng, C. W. W. and Fredlund, D. G. (2007). Field study of rainfall


infiltration into a grassed unsaturated expansive soil slope. Can. Geotech. J.44,
392-408.

Zhan, L. T., Ng, C. W. W. and Fredlund, D. G. (2006). Instrumentation of an


unsaturated expansive soil slope. Geotechnical Testing Journal, Vol. 30, No. 2,
1-11.

PLAXIS HONG KONG COURSE, 6-9 November 2012 33

Page 388
17
Advanced Computational Geotechnics

HKCG 16
1

EXERCISSE 5
SLOPE
E STA
ABILITY
Y EXEERCISE
E
Dr Jo
ohnny Cheuk
C

Page 389
Advanced Computational Geotechnics

Slope stability for a road construction project

SLOPE STABILITY FOR A ROAD


CONSTRUCTION PROJECT

Computational Geotechnics 1

Page 390
Advanced Computational Geotechnics

Slope stability for a road construction project

2 Computational Geotechnics

Page 391
Advanced Computational Geotechnics

Slope stability for a road construction project

INTRODUCTION
On the North Island of New Zealand a new road section has to be constructed along the shore line of a tidal bay,
see gure 1.

Figure 1: Situation overview for the newly constructed road

Though the easiest solution would have been to construct the road at a larger distance from the bay as the slope
gradients are easier there, this is not possible as the upper land is privately owned which for historic reasons
cannot be changed. The new road therefore had to be constructed along the steeper gradient just next to the
shore line of the tidal bay.

The hillside is mainly siltstone, weathered at the surface but intact at certain depth. Construction will take place
in summer when the ground water level is low. However, in winter the hillside side almost fully saturates due to
heavy rainfall, which has a signicant inuence on the stability. For the construction of the new road part of the
slope was excavated. The excavated material is crushed and mixed with sand and gravel to make ll material to
support the road.

During the rst winter after the road construction the road started to tilt towards the tidal bay and after assessing
the winter situation the factor of safety was considered too low. The decision was taken to stabilize the ll and
hillside below the road using so-called launched soil nails: long steel reinforcement bars that are shot with high
speed into the ground.

MAIN GOAL OF THE ANALYSIS


Determine the factor of safety of the original hillside

Construct the new road under dry (summer) conditions and calculate its factor of safety

Simulate wet (winter) conditions and calculate its factor of safety

Apply stabilising soil nails and calculate the factor of safety in wet conditions

INPUT
Start a new project and select appropriate General settings according to the size of the geometry (see gure 2)
and make sure to use a snap distance of 0.25m. Use 15-node elements as basic element type since in this exercise
we will deal with failure behaviour.

Computational Geotechnics 3

Page 392
Advanced Computational Geotechnics

Slope stability for a road construction project

Figure 2: Geometry model (a) and position of the road surface and soil nails (b)

Geometry
Enter the geometry as indicated in gure 2a. The order in which geometry points are created is arbitrary.

Introduce the 3 soil nails by using geogrids according to the coordinates given in gure 2b.

Introduce the road surface by using a plate element from (22,16) to (28,16)

Introduce the trac load by applying a vertical distributed load of -10 kN/m2 on the road surface.

Apply default boundary conditions.

Material properties
Soil and interfaces
Enter the material properties for the three soil data sets specied in table 1.

After entering all properties for the three soil types, drag and drop the properties to the appropriate clusters,
as indicated in gure 3.

4 Computational Geotechnics

Page 393
Advanced Computational Geotechnics

Slope stability for a road construction project

Table 1: Soil material set parameters


Parameter Symbol Intact siltstone Weathered siltstone Reinforced ll Units

Material model Model Mohr-Coulomb Mohr-Coulomb Mohr-Coulomb


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

Friction angle 35 19 30

Dilatancy angle 0 0 0
3
Permeability x-dir. kx 110 0.01 0.1 m/d
3
Permeability y-dir. ky 110 0.01 0.1 m/d
Tension cut-o Tension cut-o Disabled Enabled Enabled -

Weath
ered s
iltston
e
A A A

Intact siltstone

Figure 3: Geometry model showing the soil material sets

Road surface
The road surface is modelled with a plate element. Therefore, create a new plate material set using the parameters
as specied in table 2 and assign it to the plate representing the road surface.

Table 2: Properties of the road surface (plate)


Parameter Symbol Road surface Unit

Material model Model Elastic 


5
Axial stiness EA 2.510 kN/m
2
Flexural stiness EI 500 kNm /m
Weight w 3.0 kN/m/m
Poisson's ratio 0.0 

Soil nails
The 3 soil nails are modelled using geogrid elements. Hence, create a new geogrid material set with parameters
as specied in table 3 and assign the material to all 3 soil nails.

Computational Geotechnics 5

Page 394
Advanced Computational Geotechnics

Slope stability for a road construction project

Table 3: Properties of the soil nails (geogrids)


Parameter Symbol Soil nail Unit

Material model Model Elastoplastic 


Axial stiness EA 5275 kN/m
Max. tension force Np 150 kN/m

Mesh generation
Set the Global coarseness to Medium.

Select all clusters that fall within the boxed area (12 clusters in total) while keeping the <Shift> button
pressed and then select the option Rene cluster from the Mesh menu in order to rene the mesh in the
selected area. This will give a mesh as given in gure 4.

Figure 4: Medium coarse generated mesh with renement

6 Computational Geotechnics

Page 395
Advanced Computational Geotechnics

Slope stability for a road construction project

CALCULATION
The calculation consists of the initial phase and 12 calculation phases more in order to model the proper con-
struction sequence and the determination of the factors of safety at key moments in the construction process.

When starting the Calculations program select Classical mode for calculating undrained behaviour and consoli-
dation.

Initial phase
The initial situation consists of the intact hill side and a phreatic level representing typical summer conditions as
construction starts in summer. In order to dene the initial situation, follow these steps:

The geometry has a non-horizontal soil layering, hence the K0 -procedure cannot be used. Therefore, set the
Calculation type to Gravity loading.

Dene the Staged construction settings and make sure only the clusters representing the original hillside are
activated. Also make sure all structural elements (road surface and soil nails) are switched o.

In Water conditions mode, eEnter a phreatic level by two coordinates (-1, 10) and (56, 10).

Click the Update button to return to the Calculations program.

Phase 1 - Stability prior to the construction


Before the construction is started the factor of safety is determined of the initial situation

One the General tabsheet make sure this calculation phase is Safety.

Accept all default settings

Phase 2 - Road excavation


The road excavation should continue from the initial situation and not from the results of the safety factor
determination. Therefore, on the General tabsheet, set the eld Start from phase on  Initial phase .

Set the Calculation type to Plastic of loading type Staged construction.

In order to discard the displacements during gravity loading make sure the option Reset displacements to
zero is selected on the Parameters tabsheet.

On the Parameters tabsheet press the Dene button to dene the phase

Switch o the upper part of the road excavations, see gure 5.

Press the Update to return to the Calculations program.

Computational Geotechnics 7

Page 396
Advanced Computational Geotechnics

Slope stability for a road construction project

Figure 5: Phase 2, road excavation

Phase 3 - Construction of the ll


This calculation phase is again a Plastic calculation, loading type Staged construction.
For all the other settings the defaults should be used.

Switch on the additional ll

Open the material set database and assign the  reinforced ll material set to the 4 clusters of the ll area,
see gure 6.

Close the material sets database and press the Update to return to the Calculations program.

Figure 6: Phase 3, Construction of the ll

Phase 4 - Construction of the road


This calculation phase again is a Plastic calculation, loading type Staged construction. Keep all default
settings.

Switch on the plate representing the road by clicking on it. Make sure the distributed load representing the
trac load remains switched o.

Return to the Calculations program.

8 Computational Geotechnics

Page 397
Advanced Computational Geotechnics

Slope stability for a road construction project

Phase 5 - Apply the trac load


Again a Plastic calculation, topmost level of loading type Staged construction. Keep all default settings.

Switch on both parts (left nd right) of the distributed load representing the trac load. The plate repre-
senting the road surface remains switched on.

Return to the Calculations program.

We are now nished with the road construction.

Phase 6 - Factor of safety of the road in summer conditions


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

Phase 7 - Winter conditions


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

Set the Calculation type to Plastic, loading type Staged construction.


Go to the Staged construction denition and from there to the Water conditions mode.

Select the Phreatic level button and draw a new phreatic line from (-1,20) to (5,20) and further to (20,10)
and (56,10).

If there is no closed ow boundary yet on the bottom of the geometry (indicated with a thick black line)
then select the Closed boundary button and draw a closed ow boundary at the full bottom of the geometry.

Select Steady state as groundwater analysis type and press the Update button to return to the Calculations
program.

Phase 8 - Factor of safety of the road in winter conditions


In order to determine the factor of safety directly in winter conditions create a Safety phase. Keep all default
settings.

In winter conditions the factor of safety appears to be rather low and therefore it is decided to improve stability
by applying launched soil nails.

Phase 9 - Apply top level soil nails


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

Set the Calculation type to Plastic, loading type Staged construcion.


Switch on the topmost soil nail, see gure 7.

Return to the Calculations program

Computational Geotechnics 9

Page 398
Advanced Computational Geotechnics

Slope stability for a road construction project

Figure 7: Phase 9, Road construction with trac load and topmost level of soil nails

Phase 10 - Factor of safety in winter conditions with top level soil nails
In order to determine the factor of safety directly in winter conditions with the topmost level of soil nails
installed create a Safety phase. Keep all default settings

Phase 11 - Apply additional soil nails


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

Set the Calculation type to Plastic, loading type Staged construction.


Switch on the 2 other soil nails

Return to the Calculations program

Phase 12 - Factor of safety in winter conditions with all soil nails in-
stalled
In order to determine the factor of safety directly in winter conditions with the all soil nails installed create
a Safety phase. Keep all default settings

Load-displacement curves
Before starting the calculation choose some points for node-displacement curves. In order to check failure for the
phi/c reduction phases the chosen points should be in the expected failure zone. As there are several possible
slope instabilities, chose at least points at (15,20), (25,16), (28,16) and (33,11).

Now start the calculation by pressing the Calculate button.

10 Computational Geotechnics

Page 399
Advanced Computational Geotechnics

Slope stability for a road construction project

INSPECT OUTPUT
Factors of safety

Check the factors of safety in the Curves program. Create a new curve of displacements vs. Sum-Msf for the
point at coordinates (25,16). See gure 8.

1.70

1.60
FoS 1.6 (all nails installed)

1.50

FoS 1.4 (top nails installed)


1.40
Msf


1.30

FoS 1.2 (summer conditions, no nails)


1.20
FoS 1.15 (before construction)

1.10
FoS 1.15 (winter conditions, no nails)

1.00
0.00 0.50 1.00 1.50 2.00
|U| [m]

Figure 8: Factors of safety for key moments in the project.

Critical failure mechanisms

For all situations check the failure mechanism. This can be done by for instance with the graph of incremental
shear strains (s ) of every Safety phase. This will show the change in shear strains in the last calculation step,
hence when failure occurred, and will show any shear bands that may have occurred. See gures 9-13. It becomes
clear that installing the top nails disturbs the failure mechanism. However, it is only after having installed the
lower nails as well that the sliding of the road ll no longer is the most critical mechanism. Both failure of the
slope above the road and a very large hillside sliding mechanism with considerably higher factor of safety (almost
1.6) are now the critical mechanisms.

Computational Geotechnics 11

Page 400
Advanced Computational Geotechnics

Slope stability for a road construction project

Figure 9: Most critical failure mechanisme prior to construction

Figure 10: Most critical failure mechanisme after construction, summer conditions.

Figure 11: Most critical failure mechanisme after construction, winter conditions.

12 Computational Geotechnics

Page 401
Advanced Computational Geotechnics

Slope stability for a road construction project

Figure 12: Most critical failure mechanisme after installing top soil nails

Figure 13: Most critical failure mechanisme after installing all soil nails

Computational Geotechnics 13

Page 402
Advanced Computational Geotechnics

HKCG 15
1

HO
OEK-B
BROWN
N AND
D ROC
CK JOIINTED
D MOD
DEL

ofessor Helmut Schweiger


Pro

Page 403
Advanced Computational Geotechnics

S C I E N C E P A S S I O N T E C H N O L O G Y

CG17
HOEK-BROWN
JOINTED ROCK MODEL

Helmut F. Schweiger
Computational Geotechnics Group
Institute for Soil Mechanics and Foundation Engineering
Graz University of Technology

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


2
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

CONTENTS

Continuum model for jointed rock


Plaxis jointed rock model
Validation with data from literature
Examples
Hoek-Brown model
Comparison with analytical solution
Practical Example: FE-modelling of grouting pressures for
pressure tunnels
Concluding remarks on discontinua

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 404
1
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


3
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

JOINTED ROCK BEHAVIOUR


Behaviour of rock = Behaviour of intact rock + Joints (joint sets)
approx. even, unfilled peak and residual shear strength depend on joint conditions
T . friction angle joint
n tanT T* .residual friction angle joint
F . friction angle joint material
residual n tanT* F* .residual friction angle joint material
cF . cohesion joint material
approx. even, filled
cF* . residual cohesion joint material
G . friction angle intact rock
n tanF c F G* residual friction angle intact rock
cG cohesion intact rock
residual n tan F* c F*
tensile strength is usually assumed
uneven, (un)filled
to be 0 for all cases

for n n tan (T i) residual n tan ( T* i)

for n n tan G c G residual n tan G*


PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


4
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

JOINTED ROCK BEHAVIOUR

Schematic model of "rough" joint

for n

n tan (T i)

for n

n tan G c G

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 405
2
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


5
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

JOINTED ROCK BEHAVIOUR

Alternative formulation for joint behaviour

JCS
n tan JRC log10 r

n

JRC .. joint roughness coefficient


JCS .. joint compressive strength
R residual friction angle

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


6
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

JOINTED ROCK

distance of joints or fault zones large > explicit modelling required

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 406
3
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


7
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

JOINTED ROCK

distance of joints small (sets of joints) > explicit modelling not


necessarily required, approximation with suitable constitutive
model possible (smeared model, homogeneous model)

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


8
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

Assumption:
approximately parallel, continuous, unfilled joint sets

m
d rock d int act rock d joj int set
j 1

d rock d int act rock d jo int set


for each joint set a failure (yield) function is defined:

F j , n 0

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 407
4
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


9
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

Validity of model
- intact rock
- parallel joint sets
- distance between individual joints
small compared to typical
dimension of structure
- no fault gouge in joints

Features of model

behaviour of joint sets can be taken into account


Max. 3 joint sets (Plane 1, 2, 3)
anisotropic, elastic perfectly-plastic model
Mohr-Coulomb criterion on joints

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


10
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

Parameters required

Anisotropic elastic parameters


E1 Young's modulus
1 Poisson's ratio
E2 Young's modulus in Plane 1 direction
G2 Shear modulus in Plane 1 direction
2 Poisson's ratio in Plane 1 direction

Strength parameters in joint directions (Planes i = 1, 2, 3 )


ci Cohesion
i Friction angle
i Dilatancy angle
t, i Tensile strength

Definition of joint orientation (Planes i = 1, 2, 3)


n Number of joint directions
1i Dip angle
2i Dip direction
PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 408
5
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


11
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

8000

7000

6000

1-3 5000
2
[KN/m ]
4000

3000
45+(/2)

2000

1000

0
20 30 40 50 60 70 80 90

orientation
Winkel of joint 1
der Trennflchen

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


12
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

Qualitative comparison of contour lines of stresses with data from


literature ( = 0)

from Wittke 1984

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 409
6
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


13
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

Qualitative comparison of contour lines of stresses with data from


literature ( = 90)

from Wittke 1984

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


14
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

DEEP TUNNEL (K0 = 1)

Parameters
E1 [kN/m2] 1.3e6
1 - 0.25
E2 [kN/m2] 4.0e5
2 - 0.25
G2 [kN/m2] 1.33e5
c'1 [kN/m2] 50
'1 [] 22
1 [] 0 / 90 / 45

Tension cut off is assumed zero.

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 410
7
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


15
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

DEEP TUNNEL (K0 = 1)

Dip Direction = 0

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


16
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

DEEP TUNNEL (K0 = 1)

Dip Direction = 90

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 411
8
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


17
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

DEEP TUNNEL (K0 = 1)

Dip Direction = 45

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


18
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

INFLUENCE OF DIRECTION OF JOINT SET

full excavation
MStage (until failure)
unsupported

overburden: 60 m
width of tunnel: approx. 14 m
height of tunnel: approx. 12 m

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 412
9
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


19
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

INFLUENCE OF DIRECTION OF JOINT SET


PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


20
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

INFLUENCE OF DIRECTION OF JOINT SET


PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 413
10
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


21
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

INFLUENCE OF K0

influence of K0 (=32; 2652 elements)

1.0

0.9
0.8
0.7
0.6
MStage

0.5
0.4
0.3 Ko = 0,3
0.2 Ko = 0,7
0.1
Ko = 1,0
0.0
0 22.5 45 67.5 90
orientation of joint ( 1)

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


22
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

INFLUENCE OF DILATANCY

joints 1 = 45; = 0

MStage = 0.13

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 414
11
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


23
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

INFLUENCE OF DILATANCY

joints 1 = 45; = 5

MStage = 0.18

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


24
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

INFLUENCE OF DILATANCY

joints 1 = 45; = 10

MStage = 0.23

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 415
12
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


25
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

INFLUENCE OF DILATANCY

Influence of (K0=0,3; =32; 2652 elements)


1,0
0,9
0,8
0,7
0,6
MStage

0,5
0,4
0,3 Dilat. = 0

0,2 Dilat. = 5
0,1 Dilat. = 10
0,0
0 22,5 45 67,5 90
orientation of joints

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


26
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

1, 3 .. principal stresses
c .. uniaxial compressive strength
m, s . material parameters
s =1 > intact rock, s = 0 > heavily jointed
m > .. brittle behaviour, m < .. ductile behaviour

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 416
13
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


27
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

1, d 1
p

HB failure line

{dp}
gf
=0
{d }
p

=+60

{d }
p

max=+90 3, d 3
p

t +90 0
Hoek-Brown criterion: flow rule

1, d 1
p
MC failure line

{d }
p

=0
gf

{dp}
gt = ft

=90

3, d 3
p

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


28
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

GENERALISED HOEK-BROWN CRITERION


Hoek E., Carranza-Torres C., Corkum B., 2002. HoekBrown failure criterion - 2002 edition. In: Proceedings
of the North American rock mechanics Symposium, Toronto.
3
1 3 ci (mi s) a
ci
GSI 100 GSI 100
mb mi exp s exp
28 14 D 9 3D

a e GSI / 15 e 20 / 3
1 1
2 6
GSI geological strength index
mi depends on type of rock
D disturbance factor
(due to blasting and/or stress relaxation)

Failure function can be written as:


~ ~
f HB 1 3 f ( 3 ) with f ci (mb 3 s) a
ci
PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 417
14
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


29
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

GEOLOGICAL STRENGTH INDEX

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


30
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

GENERALISED HOEK-BROWN CRITERION - VALIDATION


Carranza-Torres C., 2004. Elasto-plastic solution of tunnel problem using the generalized form of the HoekBrown failure
criterion. Int J Rock Mech Min Sci; 41(3), 4801.
Intact rock parameters:
HB constant, mi [-] 10 Rock mass parameters:
Uniaxial compression strength, sci [MPa] 30 HB constant mb 1.6767
Geological strength index, GSI [-] 50 Parameter s 0.0038
Hydrostatic pressure, p0 [MPa] 15 Parameter a 0.5057
Young's modulus, E [MPa] 5700 Parameter D 0
Poisson's ratio,n [-] 0.3

Plastic radius = 2.58 m


Support pressure = 2.5 MPa

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 418
15
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


31
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

GENERALISED HOEK-BROWN CRITERION - VALIDATION

Relative shear stresses Plastic points

Plastic radius = 3.79 m


Support pressure = 0 MPa
PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


32
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

GENERALISED HOEK-BROWN CRITERION - VALIDATION

Elasto-Plastic Stress Distribution (after Carranza-Torres)

25.0

Support pressure = 0 MPa


20.0

15.0
Stress [MPa]

10.0

Radial Stress (Exact)

Tangential Stress (Exact)


5.0
Plaxis HB-Model Radial Stress

Plaxis HB-Model Tangential Stress


0.0
0.0 5.0 10.0 15.0 20.0 25.0 30.0
Distance from Tunnel Center [m]

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 419
16
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


33
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

GENERALISED HOEK-BROWN CRITERION - VALIDATION

Elasto-Plastic Stress Distribution (after Carranza-Torres)

30.0

Support pressure = 2.5 MPa


25.0

20.0
Stress [MPa]

15.0

10.0
Radial Stress (Exact)
Tangential Stress (Exact)
5.0
Plaxis HB-Model Radial Stress
Plaxis HB-Model Tangential Stress
0.0
0.0 5.0 10.0 15.0 20.0 25.0 30.0
Distance from Tunnel Center [m]

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


34
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

SHORT PROJECT DESCRIPTION

Longitudinal section

1. Project description

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 420
17
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


35
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

SHORT PROJECT DESCRIPTION

Typical cross section Detail A:


Detail A

12/21/2011Project description Assignment of tasks 35


Analytical solution
PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


36
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

PLAXIS MODEL
132m

120m
PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 421
18
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


37
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

Detail of numerical model

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


38
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

Modelling of grout pressure


Whirlpool

activated pore
tunnel interior

pressure

Queenston
Q10

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 422
19
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


39
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

Calculation phases:

0. Initial phase (K0 = 1,5)

1. Excavation (Mstage = 0,2)

2. Activation of shotcrete (Mstage = 1,0)

3. Activation of final lining

4. Pressure phase 1(8 bar, Egap <<)

5. Pressure phase 2 (15bar)

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


40
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

RESULTS

Constant pressure application: p = 15bar

Analytical PLAXIS
Solution
Normal force 10334 kN/m 10220 kN/m 1%

Displacements r = 4,06 mm Pux = 4,06 mm 0%

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 423
20
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


41
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

Diametrical displacements for unsymmetrical pressure application as


measured on site

2 1

1
4
3


D =

6
5
8

8 7

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


42
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

Unsymmetrical pressure
application in Plaxis

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 424
21
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


43
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

RESULTS

Unsymmetrical pressure application: resulting diametrical strains

D [mm] D D
1-8 3-6 4-5 2-7 [mm] [-]

Displacements on site -9,66 -3,73 -3,44 -6,43 -5,82 -4,6151E-04

Results in PLAXIS -9,60 -3,92 -3,45 -6,79 -5,94 -4,7139E-04


K0 = 1,5

Comparison with Hoek- Brown model

D [mm] D D
1-8 3-6 4-5 2-7 [mm] [-]

MC -9,60 -3,92 -3,45 -6,79 -5,94 -4,7139E-04

HB -9,60 -3,92 -3,45 -6,79 -5,94 -4,7141E-04

PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


44
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

DISCONTINUUM METHODS (E.G. DISCRETE ELEMENT METHOD)

Allows modelling of finite displacements and rotations of deformable or rigid bodies


Individual bodies can separate and new contacts are detected automatically

Advantages
Modelling of blocky structures (discontinua)
For explicit solution algorithms no equation system required
Suitable for studying micromechanical behaviour of granular materials

Disadvantages
In 3D long calculation times
For static problems artificial damping required
Influence of various input parameters difficult to judge, i.e. joint stiffnesses
(> may cause numerical problems, lot of experience required)

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 425
22
Advanced Computational Geotechnics

Hoek-Brown / Jointed Rock model


45
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

DISCONTINUUM METHODS ( E.G. DISCRETE ELEMENT METHOD)

Such displacement patterns cannot be obtained


from continuum models as discussed
PLAXIS Hong Kong Course, 6 - 9 November 2012

Hoek-Brown / Jointed Rock model


46
Continuum Model for Rock | Plaxis Jointed Rock Model | Examples | Hoek-Brown Model | Example Pressure Tunnel | Concluding Remarks

WHICH MODEL
Selection often done according to availability / familiarity rather than problem oriented
With simplification relevant mechanisms often lost, and thus neglected in the support
design

discontinuum model jointed rock model (continuum) isotropic continuum model

Acknowledgement: W. Schubert

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 426
23
Advanced Computational Geotechnics

HKCG 18
1

MO
ODELLING OF TU
UNNEL
LS IN 2
2D

ofessor Helmut Schweiger


Pro

Page 427
Advanced Computational Geotechnics

S C I E N C E P A S S I O N T E C H N O L O G Y

CG18

MODELLING OF TUNNELS IN 2D

Helmut F. Schweiger
Computational Geotechnics Group
Institute for Soil Mechanics and Foundation Engineering
Graz University of Technology

PLAXIS Hong Kong Course, 6 - 9 November 2012

2D Modelling of Tunnels
2
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

CONTENTS

Introduction
Typical excavation sequence for NATM-tunnels
Modelling 3D-effects in plane strain analysis
Calibration of pre-relaxation factors
Conclusion

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 428
1
Advanced Computational Geotechnics

2D Modelling of Tunnels
3
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

INTRODUCTION

3-D Models

easy modelling of excavation sequence


computational effort high
essential for analysis of face stability

Plane Strain

assumption of pre-relaxation factors


excavation sequence in cross section
face stability cannot be considered
"state of the art" in practice

PLAXIS Hong Kong Course, 6 - 9 November 2012

2D Modelling of Tunnels
4
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

TYPICAL NEW AUSTRIAN TUNNELLING METHOD (NATM) EXCAVATION


0 1 2 3 4 5 6 7

0 2 4 6

primary support primary support


initial stress state tunnel invert
of top heading of bench

1 3 5 7

unsupported zone transition top head transition bench final


at tunnel face - bench excavation invert excavation lining
PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 429
2
Advanced Computational Geotechnics

2D Modelling of Tunnels
5
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

DEFORMATION AHEAD OF FACE


tunnel face monitoring section
excavation
sequence

Chainage

settlements ahead
of tunnel face
Settlements

settlements of
unsupported zone
settlements after
installation of
monitoring
section

PLAXIS Hong Kong Course, 6 - 9 November 2012

2D Modelling of Tunnels
6
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

2D MODELLING - LOAD REDUCTION METHOD

E0 0( )
approximate values for :

= 0.2 0.5
for top heading
excavation

s = . 0 = 0.4 0.8
for side drift
excavation
(Laabmayr & Swoboda 1986)

tunnel lining

PRE-RELAXATION
EXCAVATION PLAXIS: Mstage 1-

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 430
3
Advanced Computational Geotechnics

2D Modelling of Tunnels
7
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

2D MODELLING - STIFFNESS REDUCTION METHOD

E0

approximate values for :

= 0.3 0.5

Ec = . E0 (Schikora & Fink 1982)

tunnel lining

PRE-RELAXATION
EXCAVATION

PLAXIS Hong Kong Course, 6 - 9 November 2012

2D Modelling of Tunnels
8
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

PRE-RELAXATION

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 431
4
Advanced Computational Geotechnics

2D Modelling of Tunnels
9
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

FULL EXCAVATION WITH LINING

PLAXIS Hong Kong Course, 6 - 9 November 2012

2D Modelling of Tunnels
10
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

FULL EXCAVATION WITHOUT LINING

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 432
5
Advanced Computational Geotechnics

2D Modelling of Tunnels
11
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

CHOICE OF AND
values depend on:
ground conditions
length of unsupported section
advance rate
time of construction of invert
experience of personnell (workmanship)
...........

Alternatively use 3D analysis to


estimate pre-relaxation factors
Note: these are different for
displacements and lining stresses

PLAXIS Hong Kong Course, 6 - 9 November 2012

2D Modelling of Tunnels
12
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

WHICH METHOD ?

Working Group 1.6 of DGGT


> Load Reduction Method

Stiffness Reduction Method


Influence of
Poisson ratio
Constitutive model

Correlation of and very difficult

PLAXIS Hong Kong Course, 6 - 9 November 2012

Page 433
6
Advanced Computational Geotechnics

2D Modelling of Tunnels
13
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

EXAMPLE NATM TUNNEL


Weathered Silt-, Claystone
Overburden 25.0 m
Top Heading Excavation

y = 60 m

Z = 60 m

y = 142 m

x = 70 m
Finite Element Meshes x = 70 m

PLAXIS Hong Kong Course, 6 - 9 November 2012

2D Modelling of Tunnels
14
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

Parameters for different constitutive models


Eoed,ref E50,ref Eur,ref c m K0 POP K0,nc ur G0,ref 0,7
MODEL [MN/m] [MN/m] [MN/m] [kN/m] [] [] [] [kN/m] [] [] [MN/m] []

1)MC, E135 E=135MN/m 35 27 0,54


2)HS, EMC=Eoed 45 45 135 35 27 0,8 0,54 0,2
A 0.7 500
B auto 500
C 0,7 0
3)HS, EMC=Eur 20 20 60 35 27 0,8 0,7 500 0,54 0,2
7)HSS, EMC=Eoed 45 45 135 35 27 0,8 0,7 500 0,54 0,2 225 2*104
9)HSS, EMC=Eur 20 20 60 35 27 0,8 0,7 500 0,54 0,2 100 2*104

PLAXIS Hong Kong Course, 6 - 9 November 2012 14

Page 434
7
Advanced Computational Geotechnics

2D Modelling of Tunnels
15
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

3D staged construction
Tunnel excavation in step i
Activation of lining with parameters shotcrete young in step i+1
Increase of cohesion in anchored region in step i+1
Increase of lining stiffness to shotcrete old in step i-2

PLAXIS Hong Kong Course, 6 - 9 November 2012 15

2D Modelling of Tunnels
16
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

2D staged construction
Pre-relaxation (MStage < 1,0)
Activation of lining with parameters shotcrete young
Increase of cohesion in anchored region
(MStage < 1,0)
Increase of lining stiffness to shotcrete old (MStage < 1,0)





MStage<1.0 MStage<1.0

Prerelaxation Shotcreteyoung+anchors Shotcreteold


PLAXIS Hong Kong Course, 6 - 9 November 2012 16

Page 435
8
Advanced Computational Geotechnics

2D Modelling of Tunnels
17
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

Calibration of pre-relaxation factors by means of crown settlements


Chainage [m]
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140

-5
Crown settlements [mm]

-15
Good agreement between
2D and 3D for final settlement
-25
Significant differences in
relaxation factor depending
-35 on constitutive model applied

-45

-55
3D 1) MC drained 2D 1) MC drained

3D 2) HS E45 (POP500,K0=0.7) drained 2D 2) HS E45 (POP500, K0=0.7) drained

3D 3) HS E20 (POP500,K0=0.7) drained 2D 3) HS E20 (POP500, K0=0.7) drained

3D 7) HSS E45 (POP500, K0=0.7) drained 2D 7) HSS E45 (POP500, K0=0.7) drained

3D 9) HSS E20 (POP500, K0=0.7) drained 2D 9) HSS E20 (POP500, K0=0.7) drained

PLAXIS Hong Kong Course, 6 - 9 November 2012 17

2D Modelling of Tunnels
18
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

Calibration of pre-relaxation factors by means of surface settlements


Chainage [m]
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
0
-2
Surface settlements [mm]

-4
-6 Poor agreement between
-8 2D and 3D for final settlement
-10
-12 Significant differences in
-14 relaxation factor depending
-16 on constitutive model applied
-18
-20
-22
-24
3D 1) MC drained
2D 1) MC drained
3D 2) HS E45 (POP500,K0=0.7) drained
2D 2) HS E45 (POP500, K0=0.7) drained
3D 3) HS20 (POP500,K0=0.7) drained
2D 3) HS E20 (POP500,K0=0.7) drained
3D 7) HSS E45 (POP500, K0=0.7) drained
2D 7) HSS E45 (POP500, K0=0.7) drained
3D 9) HSS E20 (POP500, K0=0.7) drained
2D 9) HSS E20 (POP500, K0=0.7) drained

PLAXIS Hong Kong Course, 6 - 9 November 2012 18

Page 436
9
Advanced Computational Geotechnics

2D Modelling of Tunnels
19
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

Calibration of pre-relaxation factors by means of lining forces ?


Tunnel circumference [m]
0 2 4 6 8 10 12
-300

-400

-500
N [kN/m]

-600

-700

-800

-900

3D 1) MC 2D 1) MC 3D 2) HS

2D 2) HS 3D 3) HS 2D 3) HS

3D 7) HSS 2D 7) HSS 3D 9) HSS

2D 9) HSS

PLAXIS Hong Kong Course, 6 - 9 November 2012 19

2D Modelling of Tunnels
20
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion

Conclusion
Pre-relaxation is a way of considering 3D
effects in a 2D analysis but the amount of
pre-relaxation to be taken into account in
order to match 3D results depends on a
Mstage

number of factors, such as


reference displacement
constitutive model
drained/undrained conditions

PLAXIS Hong Kong Course, 6 - 9 November 2012 20

Page 437
10
Advanced Computational Geotechnics

HKCG 19
1

EXERCIS
SE 6
LLING IN RO
TUNNEL OCK

Dr Wiilliam Ch
heang

Page 438
Advanced Computational Geotechnics

Excavation of a tunnel in rock

EXCAVATION OF A TUNNEL IN ROCK

This excercise has been created by

Prof. Helmut F. Schweiger


Technical University Graz, Austria

Computational Geotechnics 1

Page 439
Advanced Computational Geotechnics

Excavation of a tunnel in rock

2 Computational Geotechnics

Page 440
Advanced Computational Geotechnics

Excavation of a tunnel in rock

INTRODUCTION
In this example, a circular deep tunnel within a 1-layer homogeneous rock mass under a
hydrostatic pressure is modelled. As the hydrostatic pressure is the equivalent of 1500m
water it is not so practical to model this with the use of a phreatic level. In order to take into
account the hydrostatic pressure there are two possible method:

1. Apply the hydrostatic pressure as a distributed load

2. Apply the hydrostatic pressure by means of a thin heavy soil layer

Method 1) works fine if the hydrostatic pressures is applied during the calculation. However,
in our case the pressure is there from the initial situation. As distributed loads are not taken
into account during the K0 -procedure using a distributed load would require gravity loading for
applying the distributed load. This would cause additional problems to solve as gravity loading
may give plasticity in the rock joints that should not occur.
Method 2) therefore is applied in this exercise. It is chosen to use a thin layer with the same
properties as the rock (that is, a Jointed Rock material set) but with high weight. Alternatively
one can use a Linear Elastic material set with high weight, but in this case its necessary that
the Linear Elastic material has low stiffness so that it cannot prevent sliding along the joints to
reach the soil surface.

INPUT
The geometry and mesh of the model are illustrated in figure 1.
The refinement area around the tunnel can be modelled by inserting an additional circular
tunnel without lining or interface at the same location as the tunnel.

Note: Though the geometry is symmetric it is not possible to only model half the
geometry. The rock itself has an inclined stratification under an angle of
450 , what makes the model asymmetric!

Material properties
The material data sets of the rock and the thin heavy layer representing the hydrostatic
pressure (referred to as Equivalent pressure) are shown in table 1.
After excavation of the tunnel it is finished by applying a 0.2m thick concrete lining with a uni-
axial cylindrical compression strength of 55 N/mm2 and 3% reinforcement with FeB500 steel
bars (yield stress 435 N/mm2). Table 2 shows the material properties for the tunnel lining.

Mesh generation
For the generation of the mesh set the Global coarseness parameter to Medium. In addition,
select the 3 clusters that form the area around and in the tunnel, and perform a cluster

Computational Geotechnics 3

Page 441
Advanced Computational Geotechnics

Excavation of a tunnel in rock

Figure 1: Geometry of the tunnel

refinement by choosing the menu option Mesh -> Refine cluster. After the mesh has been
generated, repeat the cluster refinement for just the inner two (circular) areas.

CALCULATION
The tunnels is excavated and at a certain distance behind the tunnel face the lining is applied.
The tunnel is supposed to be self supporting and so the main purpose of the lining is to protect
the tunnel from rock fall. However, it has to be taken into account that the lining will deliver
limited support for the tunnel due to continuous stress redistribution and deformations after the
tunnel lining is installed. This is done by calculating 80% of the effect of the excavation with
an unsupported tunnel and 20% of the effect of the excavation with lining installed. Hence, the
-method is used with = 0.2.
In PLAXIS, at the beginning of the staged construction calculation phase, the changes made
cause a certain unbalance between the externally applied forces on the mesh and the internal
stresses in both soil and structural elements. This unbalance physically cannot exists and has
to be solved in the calculation phase. This is done with the use of a multiplier, -Mstage, that
indicates the amount of the unbalance that has already been solved. This means that at the
beginning of the calculation phase -Mstage=0 as all unbalance still has to be solved and at
the end of a succesfully calculated phase -Mstage=1 indicating the full unbalance could be
solved. However, in our case we only want to solve 80% of the unbalance with an unsupported
excavation before applying the tunnel lining which implies that we have to tell the calculation
kernel that not the full unbalance should be solved but that the calculation must stop when
-Mstage=0.8 is reached.

4 Computational Geotechnics

Page 442
Advanced Computational Geotechnics

Excavation of a tunnel in rock

Table 1: Rock properties

Parameter Symbol Rock Eq. pressure Units


Material model Material model Jointed Rock Jointed Rock -
Type of behaviour Drainage type Drained Drained -
Weight sat , unsat 24.0 15000 kN/m3
Youngs modulus for intact rock E1 13.0105 13.0105 kN/m2
Poissons ratio for intact rock 1 0.25 0.25 -
Youngs modulus for jointed rock E2 4.0105 4.0105 kN/m2
Poissons ratio for jointed rock 2 0.25 0.25 -
Shear modulus for jointed rock G2 1.33105 1.33105 kN/m2
Number of joint planes Number of planes 1 1 -
Cohesion of plane 1 c 50.0 50.0 kN/m2
0
Friction angle of plane 1 22.0 22.0
0
Dilatancy angle of plane 1 0.0 0.0
0
Dip angle of plane 1 1 45.0 45.0
Coefficient for initial lateral stress K0 1.0 1.0 -

Table 2: Tunnel lining properties

Parameter Symbol Lining Units


Type of behaviour Type Elastoplastic -
Axial stiffness EA 7.7106 kN/m
Flexural stiffness EI 25667 kNm2 /m
Weight w 2.4 kN/m2
Poissons ratio 0.0 -
Full plastic bending moment Mp 350.0 kNm/m
Full plastic normal force Np 10.0103 kN/m

Note: The results of a calculation phase with -Mstage < 1 have no physical
meaning as there is no equilibrium in external forces and internal stresses.
The calculation consists of the initial phase and 2 additional calculation phases

Initial phase
It is assumed there are no pore pressures in the rock, therefore the phreatic level can remain
below the geometry and no pore pressures have to be generated.
Initial stresses are generated using the K0 procedure. Make sure the thin layer representing
the top load is switched on.

Phase 1
In this stage construction phase the tunnel is being excavated while the tunnel lining remains
switched off.

Computational Geotechnics 5

Page 443
Advanced Computational Geotechnics

Excavation of a tunnel in rock

In PLAXIS Calculation go to the Parameters tabsheet.

In the Loading Input box make sure the type of loading input is Staged construction and
then press the button Advanced. A new window will appear.

In this window one can specify that the end value for the -Mstage multiplier as described
above. Fill in a value of 0.8 and press OK to close the window again

Now press the Define button in order to define the staged construction changes, i.e. the
excavation of the tunnel by switching off the cluster representing the tunnel.

Phase 2
In this phase the tunnel lining is applied. Define the staged construction phase and switch
on the lining. No other changes have to be made in this phase. PLAXIS will apply the full
unbalance that exists at the beginning of this phase, which is the remaining 20% unbalance
from excavating the tunnel and the full additional unbalance caused by switching on the tunnel
lining.

Note: The Jointed Rock model is a rather sensitive model for both mesh
refinement and local failure. With this model it may occur more often than
for the other standard consitutive models in Plaxis that the calculation
stops reporting failure, while in fact the failure is only very local and not
important from engineering point of view. In some cases this situation can
be overcome by swichting off the Arc-length control in the Manual settings
of the Iterative procedure.
Arc-length control is a method to obtain reliable load values in case of
failure, but may also lead to detection of local failure mechanisms hat are
tof little interest as stated. As long as a calculation phase ends succesfully
(that is, it was possible to apply all changes requested by the user) the
answers should be the same wether arc-length control is switched on or
not. Only when failure is reached and arc-length control is switched off,
the user will be confronted with the fact that no automatic failure detection
will be done by PLAXIS (hence the user has to decide on whether he
considers the situation failure or not) and the failure load reached is not as
accurate.
In this exercise it may be necessary to switch of arc-length control in the
first calculation phase. However, the user is encouraged to always first try
with arc-length control switched on.

One can select some points for curves, for instance at the crown and the side of the tunnel
and somewhere halfway those two points.
Press the Calculate button to start the calculation.

6 Computational Geotechnics

Page 444
Advanced Computational Geotechnics

Excavation of a tunnel in rock

OUTPUT
Check the influence of the stratification on the stresses and deformations of the rock around
the tunnel. Also, check the bending moments of the tunnel lining:

Figure 2: Plastic points at the end of phase 2

Computational Geotechnics 7

Page 445
Advanced Computational Geotechnics

Excavation of a tunnel in rock

Figure 3: Volume strains at the end of phase 2

Figure 4: Total displacements at the end of phase 2

8 Computational Geotechnics

Page 446
Advanced Computational Geotechnics

Excavation of a tunnel in rock

Figure 5: Bending moments in the lining

Computational Geotechnics 9

Page 447
Advanced Computational Geotechnics

Excavation of a tunnel in rock

FURTHER EXERCISE
Repeat the calculation for dip angles 1 = 0 o and 1 = 90 o and compare results

Figure 6: Plastic points for 1 = 0o (left) and 1 = 90o (right)

Figure 7: Volume strains for 1 = 0o (left) and 1 = 90o (right)

10 Computational Geotechnics

Page 448
Advanced Computational Geotechnics

Excavation of a tunnel in rock

Figure 8: Total displacements for 1 = 0o (left) and 1 = 90o (right)

Figure 9: Bending moments in the lining for 1 = 0o (left) and 1 = 90o (right)

Computational Geotechnics 11

Page 449

You might also like