Professional Documents
Culture Documents
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
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
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
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
Antonio Gens
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
Flow problems
Page 8
Advanced Computational Geotechnics
Equilibrium (3 equations)
Unknowns: 15
Compatibility (6 equations)
(6 stresses, 6 strains, 3 displacements)
Constitutive equation (6 equations)
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
Page 10
Advanced Computational Geotechnics
Page 11
Advanced Computational Geotechnics
the FEM is a computational procedure that may be used to obtain an approximate solution
to a boundary value problem
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
examples: embankment
Page 13
Advanced Computational Geotechnics
Node
Page 14
Advanced Computational Geotechnics
Ku F
For soil we dont have a direct relation between load and displacement,
we have a relation between stress and strain.
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
Page 15
Advanced Computational Geotechnics
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
displacement interpolation
Page 16
Advanced Computational Geotechnics
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
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
Page 17
Advanced Computational Geotechnics
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
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
D
with D depending on the current and past stress history
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
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)
The stiffness matrix for the complete mesh is evaluated by combining the individual
element stiffness matrixes (assembly)
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
assembly
schemes for storage
solution
number of dofs
KU P
These have to be solved to give values for the nodal displacements
Page 21
Advanced Computational Geotechnics
BU e
= D
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
Antonio Gens
Contents
Page 24
Advanced Computational Geotechnics
Constitutive models
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
d1
Two parameters:
- d1
- Youngs modulus E
- Poissons ratio
d3
- 1
d1 E
E
d1 1
- 1
d
3
d1 1
E 0 ; -1 0 .5 3
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
Page 28
Advanced Computational Geotechnics
e p
General three-dimensional stress state e p
Y0 = yield stress
YF = failure stress
e p
General three-dimensional stress state e p
Page 29
Advanced Computational Geotechnics
One-dimensional
Y0 = yield stress
YF = failure stress
Page 30
Advanced Computational Geotechnics
f f 1 , 2 , 3
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
Page 31
Advanced Computational Geotechnics
frictional
resistance
independent of
normal stress
Page 32
Advanced Computational Geotechnics
t* = (3 - 1)
s* = (3+1)
1
2 '3 '1 c' cos ' 12 '3 '1 sin '
f 1
2
'1 '3 12 '1 '3 sin ' c ' cos '
-1
f < 0 Elasticity
-3
-2
Page 33
Advanced Computational Geotechnics
Summing up:
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
Page 34
Advanced Computational Geotechnics
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?
dilatancy angle
Page 35
Advanced Computational Geotechnics
plastic dilatancy
= + 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)
consistency condition
Page 36
Advanced Computational Geotechnics
Parameters of MC model
yy
tan
xy
Page 37
Advanced Computational Geotechnics
Page 38
Advanced Computational Geotechnics
Page 39
Advanced Computational Geotechnics
Page 40
Advanced Computational Geotechnics
Possibilities and limitations of the Linear Elastic- Perfectly Plastic (LEPP) Mohr-
Coulomb model
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
Computational Geotechnics 1
Page 43
Advanced Computational Geotechnics
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.
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
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
Calculation of vertical and horizontal load representing building weight and wind force
A) Geometry input
General settings
Mesh generation
B) Calculations
Construct footing
C) Inspect output
Computational Geotechnics 3
Page 45
Advanced Computational Geotechnics
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).
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.
In order to enter the proper settings for the footing project, follow these steps:
4 Computational Geotechnics
Page 46
Advanced Computational Geotechnics
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.
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
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.
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
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.
Computational Geotechnics 7
Page 49
Advanced Computational Geotechnics
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).
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.
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.
8 Computational Geotechnics
Page 50
Advanced Computational Geotechnics
& 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.
To create a material set for the clay layer, follow these steps:
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
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
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
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
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
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).
14 Computational Geotechnics
Page 56
Advanced Computational Geotechnics
Make sure the Calculation type is set to K0 -procedure on the General tabsheet. This is
the default setting.
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.
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.
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.
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
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.
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
The point load A is now active (blue) and has a load value of 50 kN/m.
Press Update.
Computational Geotechnics 17
Page 59
Advanced Computational Geotechnics
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.
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.
18 Computational Geotechnics
Page 60
Advanced Computational Geotechnics
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 ( ).
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.
Computational Geotechnics 19
Page 61
Advanced Computational Geotechnics
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).
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
Figure 19 shows the Curve generation window after applying the steps mentioned.
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
Table 2: Results for the maximum load reached on a strip footing on the drained sub-soil for
different 2D and 3D meshes
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
ADDITIONAL EXERCISE:
UNDRAINED FOOTING
Computational Geotechnics 23
Page 65
Advanced Computational Geotechnics
24 Computational Geotechnics
Page 66
Advanced Computational Geotechnics
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
A) Geometry input
B) Calculations
Construct footing
C) Inspect output
Computational Geotechnics 25
Page 67
Advanced Computational Geotechnics
GEOMETRY INPUT
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
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
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
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
Computational Geotechnics 29
Page 71
Advanced Computational Geotechnics
30 Computational Geotechnics
Page 72
Advanced Computational Geotechnics
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 )
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
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
Page 75
Advanced Computational Geotechnics
Geometry: Meshing
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)
Page 76
Calculations: Phases
Advanced Computational Geotechnics
Phase 2
Phase 3
Phase 4
Page 77
Advanced Computational Geotechnics
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
CONTENTS
Page 81
1
Advanced Computational Geotechnics
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
v v
Mayne, 2006
CC
Void Ratio, e
NC NC
CSL CSL
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
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
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
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
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
Mayne, 2006
CC
Void Ratio, e
e0
OC
CS NC NC
CSL CSL
tan'
Undrained OC Soil: u
e0, vo, and OCR
Page 84
4
Advanced Computational Geotechnics
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
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
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
1+e
Page 86
6
Advanced Computational Geotechnics
Page 87
7
Advanced Computational Geotechnics
UNDRAINED PLANES
Page 88
8
Advanced Computational Geotechnics
DRAINED PLANES
Page 89
9
Advanced Computational Geotechnics
OVERCONSOLIDATION
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
d vp 0, d sp 0 (contraction, hardening)
d vp 0, d sp 0 (dilatancy, softening)
CSL f=0
PLAXIS Hong Kong Course, 6 - 9 November 2012
Page 91
11
Advanced Computational Geotechnics
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
heavily overconsolidated
drained compression
normally consolidated
undrained compression
Page 93
13
Advanced Computational Geotechnics
due
duetotoplastic
plasticsoil
soilbehaviour
behaviour
due
duetotochange
changeofofstress
stress
lightly overconsolidated
undrained compression
heavily overconsolidated
undrained compression
Page 94
14
Advanced Computational Geotechnics
e = void ratio
= swelling index
= compression index
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
Page 95
15
Advanced Computational Geotechnics
M
q 1 MC-line
K0NC
p
pp
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
Page 96
16
Advanced Computational Geotechnics
MC
-0.1
-0.2
-0.3
-0.4
-0.5
0 -100 -200 -300 -400 -500 -600
vertical stress [kN/m2]
MC
-500
-400
-300
-200
-100
0
0 -50 -100 -150 -200 -250 -300 -350
horizontal stress [kN/m2]
Page 97
17
Advanced Computational Geotechnics
elastic
Stiffness: primary loading
region
Page 98
18
Advanced Computational Geotechnics
Page 99
19
Advanced Computational Geotechnics
H
HKCG 5
H
HARDENING
G SOIL
L AND
D HS-small M
MODEL
L
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
Page 101
1
Advanced Computational Geotechnics
1-
HS MC
Page 102
2
Advanced Computational Geotechnics
q [kN/m2]
250
HS-Model
150
100
50
0
0 0.01 0.02 0.03 0.04 0.05
eps_axial
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
Page 103
3
Advanced Computational Geotechnics
Mohr Coulomb
100
vertical displacements [mm]
80
Hard. Soil
60
40
20
0
0 5 10 15 20
-20
-40
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
Page 104
4
Advanced Computational Geotechnics
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
q 1 3
Asymptote
qf
qa R f 0.9
Hyperbola qf
Rf
2 sin
qf 3 a a c cot MC failure criterion
1 sin
Page 105
5
Advanced Computational Geotechnics
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
c 0
q q
0.05
0.01
p p
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
Ref. : Ishihara, Tatsuoka and Yasuda (1975). Undrained deformation and liquefaction of sand under
cyclic stresses. Soils and Foundations, Vol. 15, No. 1.
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 '
Page 107
7
Advanced Computational Geotechnics
3 600 kPa
1500
300
500
100
Triaxial tests:
Unloading is
purely elastic in
HS model
Page 108
8
Advanced Computational Geotechnics
Biarez, J. & Hicher, P.-Y. (1994), Elementary Mechanics of Soil Behaviour, Balkema - Publishers.
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 17
q [MN/m]
vol 0.25
vol 0.20
0.14
0.07
p [MN/m]
Page 109
9
Advanced Computational Geotechnics
q
MC failure line
pp Cap
pp p
1 m
c ' cot ' p p is determined by K0nc
p
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
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
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
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
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
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
Page 113
13
Advanced Computational Geotechnics
Volumetric behaviour
"stress dilatancy theory" (Rowe, 1962)
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
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
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
plastic potential Q
m > cv = positive > dilation
q
x cv
x m = cv = 0
x
m < cv = negative > contraction
p'
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
Page 115
15
Advanced Computational Geotechnics
Rf = qf / qa (default Rf = 0.9)
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
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
m for clays:
approx. 0.9 1.0
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 34
Page 117
17
Advanced Computational Geotechnics
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
drained
Ohde / Janbu:
m
1
Eoed E ref
oed ref
p
Page 118
18
Advanced Computational Geotechnics
Parameter limitations
E50
pc
Eur E oed
e.g. E50 / Eoed > 2 difficult to input
p
double hardening
|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
Page 119
19
Advanced Computational Geotechnics
1200
1000
800
200
0
0 5 10 15
-1 [%]
3
Hardening soil model
2
test data
-1
-2
0 5 10 15
-1 [%]
Page 120
20
Advanced Computational Geotechnics
test data
300
200
100
0
0 0.005 0.01 0.015 0.02 0.025
-yy
200
100
0
0 0.005 0.01 0.015 0.02 0.025
-yy
Page 121
21
Advanced Computational Geotechnics
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
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
Page 122
22
Advanced Computational Geotechnics
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]
-600
-550
-500
-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]
Page 123
23
Advanced Computational Geotechnics
'0yy , p p
'0xx , p '0zz , p K0nc p pp p
q* | '0xx , p '0yy , p |
2
q*
p p ( p*)2
PLAXIS HONG KONG COURSE / 6 - 9 November 2012 47
p0
OCR POP
'yy0
'yy0 p0 'yy0 p0
Page 124
24
Advanced Computational Geotechnics
-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
elastic-plastic
elastic
double hardening
Page 125
25
Advanced Computational Geotechnics
Mohr- Hardening
Coulomb Soil Model
Model
failure according to
Mohr-Coulomb YES YES
Page 126
26
Advanced Computational Geotechnics
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[]
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.
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.
Page 127
27
Advanced Computational Geotechnics
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
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.
Page 128
28
Advanced Computational Geotechnics
Typical values for G0 (in MPa) for reference stress of 100 kPa
(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:
Page 129
29
Advanced Computational Geotechnics
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
Computational Geotechnics 1
Page 132
Advanced Computational Geotechnics
2 Computational Geotechnics
Page 133
Advanced Computational Geotechnics
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
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.
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
4 Computational Geotechnics
Page 135
Advanced Computational Geotechnics
Computational Geotechnics 5
Page 136
Advanced Computational Geotechnics
6 Computational Geotechnics
Page 137
Advanced Computational Geotechnics
Figure 6: Undrained triaxial (CU) tests at cell pressures of 100 kPa and 400 kPa
Computational Geotechnics 7
Page 138
Advanced Computational Geotechnics
8 Computational Geotechnics
Page 139
Advanced Computational Geotechnics
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
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
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
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
10 Computational Geotechnics
Page 141
Advanced Computational Geotechnics
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
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
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
12 Computational Geotechnics
Page 143
Advanced Computational Geotechnics
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).
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.
Computational Geotechnics 13
Page 144
Advanced Computational Geotechnics
CLAY
We start with the determination of the strength parameters based on the CU triaxial tests.
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
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.
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
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
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
0
m
ref x
Eur = Eur pref
ref
Follows, in a similar way as for the Eoed , that
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.
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:
Poissons ratio
The Poissons ratio for unloading and reloading is again estimated as ur = 0.2
16 Computational Geotechnics
Page 147
Advanced Computational Geotechnics
Computational Geotechnics 17
Page 148
Advanced Computational Geotechnics
18 Computational Geotechnics
Page 149
Advanced Computational Geotechnics
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
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 .
In the main window select the Oedometer tabsheet and set the parameters as indicated in Figure .
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
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
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
Figure 20: Unloading/reloading cycle in a triaxial test using the General option
Computational Geotechnics 23
Page 154
Advanced Computational Geotechnics
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
SECTION 1
Page 156
Advanced Computational Geotechnics
Fa/A = q (deviatoric
stress)
Typical sample size 38 mm x 76 mm a = q + r
450
400
350
Deviatorstress(kPa)
300
250
100 Testdata
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Axialstrain
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
Settlementdialgauge
OedometerCell
Sample:dia.=75mm
Protruded
leverarm Height=20mm
Heavydeadweights
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)
400
300
Verticalpressure(kPa)
200
100
Testdata
0
0 50 100 150 200
Lateralstress(kPa)
SECTION 2
Page 162
Part 1: Strength parameters
Advanced Computational Geotechnics
450
400
Since c = 0 for sand, it can be
350
simplified to:
Deviatorstress(kPa)
300
250
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
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
Page 164
pc p
Part 2: Stiffness parameters
Advanced Computational Geotechnics
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
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%
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
Jakys formula:
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
SECTION 3
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)
Gradient:
350
Testdata
300
200
195
150
100 = 25
50
0 Intercept:
0 50 100 150 200 250 300 350 400 450
p'(kPa)
c = 0
Testdata
0.1
Verticalstrain(%)
0.2
0.3
0.4
0.5
1 10 100 1000
Verticalpressure(kPa)
d y '
Testdata
Eoed
0.1
d y
Verticalstrain(%)
0.2
d (log y ' )
Gradient _ k
d y
0.3
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 '
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
0
Eur refers to when 3 = 100kPa
Testdata
0.1 During oedometer loading, when y
=100kPa, x<100kPa;
Verticalstrain(%)
0.2
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
SECTION 4
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
Antonio Gens
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
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
Excavations: general
II Pre-stressing anchor
Page 182
Advanced Computational Geotechnics
Excavations: general
III Final excavation
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
Dry excavation
Click on the cluster or clusters to be excavated
Use cluster dry option or cluster phreatic line
Set new hydraulic conditions
Page 184
Advanced Computational Geotechnics
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
Page 185
Advanced Computational Geotechnics
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
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
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
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
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]
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
Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions
DENSE SAND
2 2
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
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
Page 193
Advanced Computational Geotechnics
Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions
3 3
5 5
6 6
7 7
8 8
9 9
10 10
Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions
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
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
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
3 3
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
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
Page 197
Advanced Computational Geotechnics
Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions
MARL
Introduction | Benchmark 1 | Benchmark 2 | Comparison Measurements | Safety Analysis | Summary and Conclusions
MARL
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
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
Page 199
Advanced Computational Geotechnics
Material behaviour
NOTE ON / c - REDUCTION
Material parameters soil layer:
sheet pile
type AZ-18 = 35
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
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
Dr Johnny Cheuk
Aecom Asia
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
4. Interfaceelement
Section 3.44 & 14.1
Overview:
1. 3or5nodedlineelements(for6nodedor15nodedelementmesh)
2. 3degreesoffreedompernode
3. Plateshave:
o Axialforces
o Shearforces
o Bendingmoments
o Hoopforces(axisymmetry)
4. Elasticorelastoplasticbehaviour
5. Formodellingwalls,floors,tunnels
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
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
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
Plates weight,insoil
dreal
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
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
(Illustration: Beam.P2D): d
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)
Probableactualdistributionofaxial
forcesingroundanchor
axialforcesingeotextileelement
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)
23
References
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 3
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 3
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)
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
CONTENTS
A. INTODUCTION
C. RESULTS
D. DISCUSSIONS
Page 239
Advanced Computational Geotechnics
PART 1
EXERCISE 3A
INTRODUCTION
Page 240
INTRODUCTION
Advanced Computational Geotechnics
Page 241
STRENGTH AND
Advanced MODULUS
Computational GeotechnicsVALUES
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
Page 243
STAGES
Advanced Computational Geotechnics
RESULTS
Page 244
Inspect Output 1:Advanced
Deformed mesh
Computational for MC-E50 (Stage 5)
Geotechnics
Page 245
Inspect Output 2: Displacements (Shading)
Advanced Computational Geotechnics for MC E50 Stage 5
MC-model (E50)
MC-model (Eur)
MC-model (Eur)
Page 246
Advanced Computational Geotechnics
EXERCISE 3A
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]
Page 247
Advanced Computational Geotechnics
-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]
PART 2
EXERCISE 3B
Page 248
Advanced Computational Geotechnics
SAME MODEL, BUT CONSTITUTIVE MODEL IS HS & HS-small
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?
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]
Page 251
Advanced Computational Geotechnics
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]
-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]
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
Antonio Gens
outline
Page 254
Advanced Computational Geotechnics
drained undrained
Page 255
Advanced Computational Geotechnics
drained undrained
1 3
t
2
3
s' 1
2
3
s 1
2
Page 256
Advanced Computational Geotechnics
Undrained behaviour
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
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!
Cu
1 3 1 3
2
F
2
F cu ,
-Cu
Effective stresses
OC
OC
NC
NC
Loading
Unloading
s, s
OC
OC
NC
NC
Loading
Unloading
s, s
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
We need D D
' f D D ' D f ( D ' D f )
D D' D f
Page 260
Advanced Computational Geotechnics
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
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
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
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
Page 262
Advanced Computational Geotechnics
E' 1 u
K total assuming u = 0.495
31 2 u 1 '
Note: this procedure gives reasonable results only for ' < 0.35 !
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', '
uf
u
TSP
cu ESP
s, s
Page 265
Advanced Computational Geotechnics
TSP=ESP
cu
s, s
Page 266
Advanced Computational Geotechnics
ESP TSP
cu
s, s
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
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
cuMC
cu real
s, s
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)
cu/v=0.279 cu/v=0.214
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)
cu/v=0.279 cu/v=0.277
Page 270
Advanced Computational Geotechnics
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
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
300
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
Antonio Gens
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)
EXAMPLE
Excess pore
water pressure
Consolidation
Page 275
Advanced Computational Geotechnics
OTHER EXAMPLES
1973 1984
Ekofisk tank
OTHER EXAMPLES
Page 276
Advanced Computational Geotechnics
CONSOLIDATION
q x q y n
div q
x y t
n
n q x q y t
0
t x y
Theory of consolidation
Considering:
Page 277
Advanced Computational Geotechnics
Theory of consolidation
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
Dependence on void ratio
Page 280
Advanced Computational Geotechnics
PERMEABILITY
k e
log
k0 ck
Default value for ck is 1015
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)
Node
Element
Stress
point
Mesh:
Elements: Interpolation of primary variables
Nodes: Primary variables (displacements, pore pressures)
Stress points: Derived variables (strains, stresses, Darcy velocities)
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
* *
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
Sand
Clay pw
Sand pw0
Page 284
Advanced Computational Geotechnics
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
Calculations:
Consolidation Staged construction > Time interval t
Consolidation Minimum pore pressure > |p-stop|
Consolidation Incremental multipliers > Time increment
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)
Page 286
Advanced Computational Geotechnics
FEM for consolidation analysis
Page 287
Advanced Computational Geotechnics
Conclusions
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
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
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
2. Advanced
3. Flow
Page 292
Advanced Computational Geotechnics
2. Steady-state pre pressures are generated due to water conditions (hydraulic boundaries)
assigned to soil clusters (layers)
Page 293
Advanced Computational Geotechnics
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
B2.ADVANCED MODE
A. Consolidation analysis
B. Transient groundwater flow analysis
KERNEL
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
Page 296
Advanced Computational Geotechnics
2. Cluster Dry
Page 297
Advanced Computational Geotechnics
Page 298
Advanced Computational Geotechnics
Cluster: Dry
Page 299
Advanced Computational Geotechnics
Page 300
Advanced Computational Geotechnics
Cluster Dry
Interpolated
Case Histories
Page 301
Advanced Computational Geotechnics
Page 302
Advanced Computational Geotechnics
1. Boundary conditions:
2. Soil permeabilities
Page 303
Advanced Computational Geotechnics
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
Page 304
Advanced Computational Geotechnics
1. Qualitative evaluation:
Flow field
Location of phreatic line
2. Quantitative evaluation:
Heads, pore pressures compared to hydrostatic,
Compare with measurements or field experience
Page 305
Advanced Computational Geotechnics
krel
1
hp = -
hp = 0
m
0 hp
Page 306
Advanced Computational Geotechnics
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
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
Page 308
Advanced Computational Geotechnics
Upper soils:
< 1m below soil surface
Lower soils:
all deeper soils
Hydraulic Properties of
European Soils
Particle distribution:
< 2m
2m - 50m
50m 2mm
Page 309
Advanced Computational Geotechnics
Particle distribution:
< 2m
2m - 50m
50m 2mm
12 soils data sets
No difference between
upper and lower soils
Relative permeability
Degree of saturation
Page 310
Advanced Computational Geotechnics
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
Computational Geotechnics 1
Page 317
Advanced Computational Geotechnics
2 Computational Geotechnics
Page 318
Advanced Computational Geotechnics
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.
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
(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
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.
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
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.
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
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.
6 Computational Geotechnics
Page 322
Advanced Computational Geotechnics
In Staged construction mode activate the wall and the interface extensions below the wall. The interfaces
along the wall are activated automatically.
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 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.
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).
Computational Geotechnics 7
Page 323
Advanced Computational Geotechnics
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.
Check that both the axis of symmetry and the bottom of the model are closed boundaries.
Draw a new general phreatic level from (x,y) = (-2,-10) to (20,-10), (30,-3) and (47,-3).
8 Computational Geotechnics
Page 324
Advanced Computational Geotechnics
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
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.
Computational Geotechnics 9
Page 325
Advanced Computational Geotechnics
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.
Also on the Parameters tabsheet, set the number of Additional steps to 500.
Dene the staged construction phase and switch to Water conditions mode.
10 Computational Geotechnics
Page 326
Advanced Computational Geotechnics
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
Figure 5: Bending moments in the wall for the steady state ow analysis (left) and the transient ow analyis
(right)
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
Antonio Gens
Page 330
Advanced Computational Geotechnics
Unsaturated soils
Solid
Liquid Gas
Fig. 4
Collapse in Via Luigi Settembrini, Naples (15-09-2001)
Page 332
Advanced Computational Geotechnics
c o g z
Matric Osmotic Gas Gravitational
Gravitational
potential
Gas pressure
potential
Matric potential
SEMIPERMEABLE
MEMBRANE
ns
os RT
V
SOLUTE
PURE
WATER
Osmotic potential
Page 334
Advanced Computational Geotechnics
Page 335
Advanced Computational Geotechnics
n q x q y n
0
t x y t
(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
(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
Page 336
Advanced Computational Geotechnics
UNSATURATED
SATURATED
SATURATED
Page 337
Advanced Computational Geotechnics
UNIFORM
1 g n
gn
gn
S ( p ) S residu ( S sat S residu ) 1 g a p
p s ( pa pw ) in m
h pw
q k k h y
dy dy w
Relative permeability
k k rel ( S r ) k sat
Fredlund &
Rahardjo (1993)
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
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
Page 341
Advanced Computational Geotechnics
s ua u w
Page 342
Advanced Computational Geotechnics
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
Moderate suctions
Variation of
apparent cohesion
and friction with
suction
Fredlund & Rahardjo (1985)
s
s Large suctions
s
Page 346
Advanced Computational Geotechnics
Sr
' ( 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
Page 347
Advanced Computational Geotechnics
S r pw
S e pw
Page 348
Advanced Computational Geotechnics
(Oedometer
tests on a
Brazilian
residual soil;
Fig. 5 Lemos, 1998)
Page 349
Advanced Computational Geotechnics
Behaviour of unsaturated soils: volume change
Collapse
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
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
Page 352
Advanced Computational Geotechnics
domain
SUCTION, s
S3
S3
S2 S2
S1
S1
S=0
MEAN NET STRESS, p MEAN NET STRESS, p
p
o 1
L
S1
C
Collapse
p p
*
o 1
*
o 2 MEAN NET STRESS, p
Isotropic plane
Page 353
Advanced Computational Geotechnics
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
q
LC
s CSL (s)
Page 354
Advanced Computational Geotechnics
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
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.
53
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
Page 356 54
Advanced Computational Geotechnics
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
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
Page 357 56
Advanced Computational Geotechnics
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
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
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
Helmut F. Schweiger
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
K0-procedure
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
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
INITIAL STRESSES
Page 363
Advanced Computational Geotechnics
INITIAL STRESSES
Special cases
Gravity loading needed due to geometry, but K0
1
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
reset displacements to 0
Page 364
Advanced Computational Geotechnics
PHI-C-REDUCTION
FACTOR OF SAFETY
Many possible definitions
PHI-C-REDUCTION
Page 365
Advanced Computational Geotechnics
PHI-C-REDUCTION
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
PHI-C-REDUCTION
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
PHI-C-REDUCTION
1.2
1
0 0.025 0.05 0.075 0.1
Total displacement node X
Page 368
Advanced Computational Geotechnics
PHI-C-REDUCTION
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
Plaxis:
F = 1,34
cu = 50 kPa
cu = 60 kPa
Page 369
Advanced Computational Geotechnics
PHI-C-REDUCTION
Plaxis:
F = 1,19
cu = 50 kPa
cu = 50 kPa
PHI-C-REDUCTION
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
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
Mesh: Refine and redo the phi-c analysis until the factor of
safety remains constant upon further refinement of the mesh.
Page 371
Advanced Computational Geotechnics
CG15-B
SLOPE STABILITY
(considering rainfall infiltration)
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION
Page 372
1
Advanced Computational Geotechnics
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION
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.
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
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
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION
= ( - S uw)
Page 374
3
Advanced Computational Geotechnics
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION
10m
15m
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION
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
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
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
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
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
saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
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
saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
After 72 hours rain infiltration: 65,00
60.00
Page 378
7
Advanced Computational Geotechnics
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION
saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
After 6 hours rain infiltration: 65,00
60.00
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION
saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
After 72 hours rain infiltration: 65,00
60.00
Page 379
8
Advanced Computational Geotechnics
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION
saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION
saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
After 72 hours rain infiltration: 65,00
60.00
Page 380
9
Advanced Computational Geotechnics
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION
saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION
saturation
(%)
100.00
95.00
90.00
85,00
80.00
75,00
70.00
After 72 hours rain infiltration: 65,00
60.00
Page 381
10
Advanced Computational Geotechnics
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION - SUMMARY
Clay
Sandy
Clay
Silt
Loamy
Sand
SLOPE STABILITY
FOS OF SLOPE CONSIDERING RAINFALL INFILTRATION - SUMMARY
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)
Page 382
11
Advanced Computational Geotechnics
SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST
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.
SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST
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)
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
SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST
Soil Parameters:
Description Symbol Unit Value
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.
Page 384
13
Advanced Computational Geotechnics
SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST
Period 1 Period 2
90
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)
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
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
Page 386
15
Advanced Computational Geotechnics
SLOPE STABILITY
COMPARISON OF FE-SIMULATION WITH FIELD TEST - RESULTS
R2--1.2
R2--1.6
R2-TC-1.6
SLOPE STABILITY
CONCLUSIONS
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.
Page 387
16
Advanced Computational Geotechnics
SLOPE STABILITY
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.
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
Computational Geotechnics 1
Page 390
Advanced Computational Geotechnics
2 Computational Geotechnics
Page 391
Advanced Computational Geotechnics
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.
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.
Construct the new road under dry (summer) 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
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.
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
Weath
ered s
iltston
e
A A A
Intact siltstone
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.
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
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.
6 Computational Geotechnics
Page 395
Advanced Computational Geotechnics
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).
One the General tabsheet make sure this calculation phase is Safety.
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
Computational Geotechnics 7
Page 396
Advanced Computational Geotechnics
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.
Switch on the plate representing the road by clicking on it. Make sure the distributed load representing the
trac load remains switched o.
8 Computational Geotechnics
Page 397
Advanced Computational Geotechnics
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.
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.
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.
Computational Geotechnics 9
Page 398
Advanced Computational Geotechnics
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 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).
10 Computational Geotechnics
Page 399
Advanced Computational Geotechnics
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
1.30
1.10
FoS 1.15 (winter conditions, no nails)
1.00
0.00 0.50 1.00 1.50 2.00
|U| [m]
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
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
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
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
CONTENTS
Page 404
1
Advanced Computational Geotechnics
for n
n tan (T i)
for n
n tan G c G
Page 405
2
Advanced Computational Geotechnics
JCS
n tan JRC log10 r
n
JOINTED ROCK
Page 406
3
Advanced Computational Geotechnics
JOINTED ROCK
Assumption:
approximately parallel, continuous, unfilled joint sets
m
d rock d int act rock d joj int set
j 1
F j , n 0
Page 407
4
Advanced Computational Geotechnics
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
Parameters required
Page 408
5
Advanced Computational Geotechnics
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
Page 409
6
Advanced Computational Geotechnics
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
Page 410
7
Advanced Computational Geotechnics
Dip Direction = 0
Dip Direction = 90
Page 411
8
Advanced Computational Geotechnics
Dip Direction = 45
full excavation
MStage (until failure)
unsupported
overburden: 60 m
width of tunnel: approx. 14 m
height of tunnel: approx. 12 m
Page 412
9
Advanced Computational Geotechnics
Page 413
10
Advanced Computational Geotechnics
INFLUENCE OF K0
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)
INFLUENCE OF DILATANCY
joints 1 = 45; = 0
MStage = 0.13
Page 414
11
Advanced Computational Geotechnics
INFLUENCE OF DILATANCY
joints 1 = 45; = 5
MStage = 0.18
INFLUENCE OF DILATANCY
joints 1 = 45; = 10
MStage = 0.23
Page 415
12
Advanced Computational Geotechnics
INFLUENCE OF DILATANCY
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
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
Page 416
13
Advanced Computational Geotechnics
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
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)
Page 417
14
Advanced Computational Geotechnics
Page 418
15
Advanced Computational Geotechnics
25.0
15.0
Stress [MPa]
10.0
Page 419
16
Advanced Computational Geotechnics
30.0
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]
Longitudinal section
1. Project description
Page 420
17
Advanced Computational Geotechnics
PLAXIS MODEL
132m
120m
PLAXIS Hong Kong Course, 6 - 9 November 2012
Page 421
18
Advanced Computational Geotechnics
activated pore
tunnel interior
pressure
Queenston
Q10
Page 422
19
Advanced Computational Geotechnics
Calculation phases:
RESULTS
Analytical PLAXIS
Solution
Normal force 10334 kN/m 10220 kN/m 1%
Page 423
20
Advanced Computational Geotechnics
2 1
1
4
3
D =
6
5
8
8 7
Unsymmetrical pressure
application in Plaxis
Page 424
21
Advanced Computational Geotechnics
RESULTS
D [mm] D D
1-8 3-6 4-5 2-7 [mm] [-]
D [mm] D D
1-8 3-6 4-5 2-7 [mm] [-]
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)
Page 425
22
Advanced Computational Geotechnics
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
Acknowledgement: W. Schubert
Page 426
23
Advanced Computational Geotechnics
HKCG 18
1
MO
ODELLING OF TU
UNNEL
LS IN 2
2D
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
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
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
Plane Strain
2D Modelling of Tunnels
4
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
0 2 4 6
1 3 5 7
Page 429
2
Advanced Computational Geotechnics
2D Modelling of Tunnels
5
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
Chainage
settlements ahead
of tunnel face
Settlements
settlements of
unsupported zone
settlements after
installation of
monitoring
section
2D Modelling of Tunnels
6
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
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-
Page 430
3
Advanced Computational Geotechnics
2D Modelling of Tunnels
7
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
E0
= 0.3 0.5
tunnel lining
PRE-RELAXATION
EXCAVATION
2D Modelling of Tunnels
8
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
PRE-RELAXATION
Page 431
4
Advanced Computational Geotechnics
2D Modelling of Tunnels
9
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
2D Modelling of Tunnels
10
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
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)
...........
2D Modelling of Tunnels
12
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
WHICH METHOD ?
Page 433
6
Advanced Computational Geotechnics
2D Modelling of Tunnels
13
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
y = 60 m
Z = 60 m
y = 142 m
x = 70 m
Finite Element Meshes x = 70 m
2D Modelling of Tunnels
14
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
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
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
Page 435
8
Advanced Computational Geotechnics
2D Modelling of Tunnels
17
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
-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 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
2D Modelling of Tunnels
18
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
-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
Page 436
9
Advanced Computational Geotechnics
2D Modelling of Tunnels
19
Introduction | NATM Excavation Sequence | 3D Effects in 2D | Calibration of pre-relaxation | Conclusion
-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
2D 9) HSS
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
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
Computational Geotechnics 1
Page 439
Advanced Computational Geotechnics
2 Computational Geotechnics
Page 440
Advanced Computational Geotechnics
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:
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
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
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
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
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:
Computational Geotechnics 7
Page 445
Advanced Computational Geotechnics
8 Computational Geotechnics
Page 446
Advanced Computational Geotechnics
Computational Geotechnics 9
Page 447
Advanced Computational Geotechnics
FURTHER EXERCISE
Repeat the calculation for dip angles 1 = 0 o and 1 = 90 o and compare results
10 Computational Geotechnics
Page 448
Advanced Computational Geotechnics
Figure 9: Bending moments in the lining for 1 = 0o (left) and 1 = 90o (right)
Computational Geotechnics 11
Page 449