Plaxis Introductory Course, Danang, Vietnam
PLAXIS DANANG 2013
10:30 12:00 Exercise 5 Stability of A Slope Reinforced with Soil Nails Dr Cheang
12:00 1:30 Lunch
4 of 375
LECTURERS
Dr. William Cheang Wai Lum Dr. Phung Duc Long
Plaxis AsiaPac, Singapore VSSMGE, Hanoi, Vietnam
William obtained his PhD from the Dr. Phung got PhD degree at Chalmers
National University of Singapore. His University of Technology, Sweden. He
interest is in Computational has more than 30 years of international
Geotechnics. He has worked as a experience, including more than 20 years
Geotechnical Engineer in Malaysia, with Plaxis. His expertise areas are: deep
Singapore and Thailand. He is involved foundations, deep excavations, soil
with many seminars and workshops improvement, pile dynamics, tunnelling,
around Asia for the promotion of good and numerical analysis. He has worked
and effective usage of Plaxis Finite with projects in many countries, among
Element Codes. other, Sweden, Norway, Denmark, USA,
England, Russia, Germany, India, Hong
Kong, China and Vietnam, etc.
ORGANIZERS
Construction Informatics and Consultancy
Plaxis AsiaPac, Singapore
JSC (CIC)
ADD: 37 Le Dai Hanh, Hai Ba Trung
16 Jalan Kilang Timor,
Dist., Hanoi, Vietnam,
0508 Redhill Forum, TEL#: (84.4) 39746798
Singapore FAX#: (844) 38216793
Contact: Mr. Luong Thanh Hung
Mobile: (84) 988 922 884
Email: Hunglt@cic.com.vn
Skype : Hunglt07
University of Transport and
Communications, Fecon
Contact: Mr. Le Quang Hanh
Mobile: (84) 948 171 135
Email: hanhquangle@fecon.com.vn
CONTENTS
LECTURES&EXERCISES PAGE
Lecture1 GeotechnicalFiniteElementAnalysis 5
Lecture2 MohrCoulombModel 21
Exercise Exercise1:SimpleFoundationwithMohrCoulombModel 43
Lecture3 NonlinearCalculations 74
Lecture4 HardeningSoil&HSsmallModel 88
Antonio Gens
Outline
Introduction
Finite Elements displacement analysis
Elements for twodimensional 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
Twodimensional elastic analysis
5 of 375
design requirements in geotechnical engineering
Plaxis Introductory Course, Danang, Vietnam
Equilibrium (3 equations)
Unknowns: 15
Compatibility (6 equations)
((6 stresses,, 6 strains,, 3 displacements)
p )
Constitutive
C tit ti equation
ti (6 equations)
ti )
6 of 375
geotechnical numerical analysis
Plaxis Introductory Course, Danang, Vietnam
while the FEM has been used in many fields of engineering practice for
over 30 years, it is only recently that it has begun to be widely used for
analyzing geotechnical problems.
when properly used, this method can produce realistic results which are
of value to practical soil engineering problems
7 of 375
geotechnical finite element analysis
Plaxis Introductory Course, Danang, Vietnam
(Nea l ) unavoidable
(Nearly) na oidable uncertainties
nce tainties
Ground profile
Initial conditions (initial stresses
stresses, pore water pressure
pressure))
Boundary conditions (mechanical, hydraulic)
pp op ate model
Appropriate ode for
o soil
so behaviour
be a ou
Model parameters
8 of 375
three final remarks
Plaxis Introductory Course, Danang, Vietnam
1. ggeotechnical engineering
g g is complex.
p y
It is not because youre
using the FEM that it becomes simpler
9 of 375
introduction: the Finite Element Method
Plaxis Introductory Course, Danang, Vietnam
Footing
width = B
Node
Gauss point
examples: embankment
10 of 375
examples: multianchored diaphragm wall
Plaxis Introductory Course, Danang, Vietnam
11 of 375
displacement interpolation
Plaxis Introductory Course, Danang, Vietnam
displacement interpolation
shape functions
= nodal
d l displacements
di l t
12 of 375
illustration for thePlaxis
sixnoded triangular element
Introductory Course, Danang, Vietnam
y quadratic interpolation
3
v x
6 5
strains may be derived within the element using the standard definitions
B matrix
Bmatrix
1 v v 0
D
E v 1 v 0
(1 2v)(1 v) 1 2v
0 0
2
in this case the coefficients of the matrix are constants,, which means
that (for linear kinematics) the resulting F.E. equations are linear
13 of 375
what happens with inelastic constitutive relations?
Plaxis Introductory Course, Danang, Vietnam
D
with D depending on the current and past stress history
it is necessaryy to apply
pp y the external load in separate
p increments
and to adopt a suitable nonlinear solution scheme
P1x
nodal forces may be related
to the nodal displacements by: 3 1y
P
P2 x
K e Ue Pe 6 5 P2 y
P e
P1x 1
Ke element stiffness matrix 2
4
P1y
P6 x
Ke
B T DBdv P
6y
14 of 375
Gauss points Plaxis Introductory Course, Danang, Vietnam
Ke
B T DBdv
the stiffness matrix for the complete mesh is evaluated by combining the
individual
i di id l element
l t stiffness
tiff matrixes
ti (assembly)
( bl )
the assembled stiffness matrix and force vector are related by:
KUe P
where vector Ue contains the displacements at all the nodes in the mesh
15 of 375
global stiffness matrix (2)
Plaxis Introductory Course, Danang, Vietnam
the global stiffness matrix generally contains many terms that are zero
if the node numbering scheme is efficient then all of the non
nonzero
zero
terms are clustered in a band along the leading diagonal
assembly
schemes for storage
solution
number of dofs
KU e P
these have to be solved to give values for the nodal displacements
16 of 375
Computation of other variables
Plaxis Introductory Course, Danang, Vietnam
once the nodal displacements have been obtained from the inversion
off the
h matrix
i K
KU P
e
= D
a xx 1
b yy
xx xx E 0
E
xx 2b xx
yy
v 1 v 0.5
xx
2a
1
xx ( xx v yy v zz )
stressstrain equations E
in three dimensions
1
yy ( yy v xx v zz )
(linear isotropic elasticity) E
1
zz ( zz v xx v yy )
E
17 of 375
elasticity as applied to soils
Plaxis Introductory Course, Danang, Vietnam
yx
shear stress shear strain
xy
xy xy xy
xy
G
yx
E E
set of relations, e.g.: G K
2(1 v) 3(1 2v)
Iff v 0.5 then
h K
xx
E
(1 v) xx v yy
(1 2v)(1 v)
yy
E
(1 v) yy v xx
(1 2v)(1 v)
xy G
G xy
18 of 375
twodimensional elastic analysis
Plaxis Introductory Course, Danang, Vietnam
Matrix D
E E
In terms of K and G
3(1 2 ) 2(1 )
4 2 2
K G K G K G 0
xx
3 3 3
xx
e
D
yy
K 2 G K 4 G K 2 G 0 e
3 3 3 eyy
zz zz
K G K G K G 0 e
2 2 4
xy 3 3 3 xy
0 G
0 0
19 of 375
elastic effective stress analysis
Plaxis Introductory Course, Danang, Vietnam
in practice,
practice elastic soil parameters obtained from laboratory or
field testing may be specified in terms of either drained or
undrained values. If undrained values are specified, then it
i necessary to
is t convertt these
th t drained
to d i d values
l i order
in d tot
carry out undrained calculations using effective stress analysis
Gu G G
drained and undrained Young's moduli are related by
20 of 375
Plaxis Introductory Course, Danang, Vietnam
MOHRCOULOMB MODEL
Elasticity, Plasticity & Yielding of Soils
William Cheang Wai Lum
PhD MSc PGDip BEng (Hons)
MohrCoulomb
Mohr Coulomb model and soil stiffness
Objectives:
21 of 375
Plaxis Introductory Course, Danang, Vietnam
F
13
strength
t th
P
stiffness
tiff
1
1
v
1 dilatancy
1
3 v 3
P
Preconsolidation
lid ti stress
t
1 reloading 1
primary
i lloading
di
1
1
unloading
22 of 375
Plaxis Introductory Course, Danang, Vietnam
1 time
1 creep
23 of 375
Plaxis Introductory Course, Danang, Vietnam
24 of 375
Plaxis Introductory Course, Danang, Vietnam
Types of stress
stressstrain
strain behaviour
Lin. elast. perfectlyplast. EP strainhardening EP strainsoftening
Plaxis Introductory Course, Da Nang 10  12 July 2013
Hookes
Hooke s law
xx 1 0 0 0 xx
1 0 0 0
yy yy
zz E 1 0 0 0 zz
(1 )(1 2 ) 0 0 0 1 0 0
xy 2 xy
yz 0 0 0 0 1
2
0 yz
1
zx 0 0 0 0 0 2 zx
Inverse:
xx 1 0 0 0 xx
1 0 0 0
yy yy
zz 1 1 0 0 0 zz
E 0 0 0 2 2 0 0
xy xy
yz 0 0 0 0 2 2 0 yz
zx 0 0 0 0 0 2 2 zx
25 of 375
Plaxis Introductory Course, Danang, Vietnam
Hookes
Hooke s law
Two parameters:
 d1
 Youngs modulus E
 Poissons ratio
d3
 1
Meaning (axial compr.):
d1 E
E
d1 1
d  1
3
d1 1
3
Plaxis Introductory Course, Da Nang 10  12 July 2013
26 of 375
Plaxis Introductory Course, Danang, Vietnam
Bulk modulus:
dpp
dp E
K dv
d v 31 2
Oedometer modulus:
 d1
d E 1
 d1
Eoed 1
d1 1 1 2
Stress definitions
1. In general
1 general, soil cannot sustain tension
tension, only compression
2. PLAXIS adopts the general mechanics definition of stress and
strain: Tension/extension is positive; Pressure/compression is
negative
yy yy
xx xx xx xx
yy yy
27 of 375
Plaxis Introductory Course, Danang, Vietnam
Hookes
Hooke s law for effective stress rates
The modeling
Th d li off nonlinear
li soilil behaviour
b h i requires
i a relationship
l ti hi
between effective stress rates (d ) and strain rates (d)
Symbolic: d ' D d
e
d D e 1
d '
Plasticity
Basic principle of elastoplasticity:
28 of 375
Plaxis Introductory Course, Danang, Vietnam
Plasticity
29 of 375
Plaxis Introductory Course, Danang, Vietnam
f<0
Within the yield contour: f<0 f>0
On the yield contour: f=0
Outside the yield contour: f>0 (impossible stress state)
A c
F n
Plaxis Introductory Course, Da Nang 10  12 July 2013
30 of 375
Plaxis Introductory Course, Danang, Vietnam
The Mohr
MohrCoulomb
Coulomb failure criterion
c cos MC criterion:
c
3
 1
 n

s*
31 of 375
Plaxis Introductory Course, Danang, Vietnam
The Mohr
MohrCoulomb
Coulomb failure criterion
( 3  1)
t* = (
s* = (3+1)
1
2 '3 '1 c' cos ' 12 '3 '1 sin '
c
n
1
2c' cos '
a
b 1 sin '
1
1 sin '
a b
1 sin '
3

Plaxis Introductory Course, Da Nang 10  12 July 2013
32 of 375
Plaxis Introductory Course, Danang, Vietnam
1 1
2 '3 '2 c' cos ' 12 '3 '2 sin '
2 '2 '3 c ' cos ' 2 '2 '3 sin '
1 1
3
2 '1 '2 c ' cos ' 2 '1 '2 sin '
2 1 1
1
2 '3 '1 c' cos ' 12 '3 '1 sin '
33 of 375
Plaxis Introductory Course, Danang, Vietnam
34 of 375
Plaxis Introductory Course, Danang, Vietnam
Failure in a simple
shear test: xx
t*
yy
g
d ije D e ijkl d 'kl 0
1
d ijp d d xx 0
'ij
g ' ' yy 1
d xxp d d xx 2 sin 0
' xx 4 t*
g ' ' xx 1
d yyp d d yy 2 sin d sin
' yy 4 t*
g '
d xyp d d xy d cos
' xy t*
d yy d yyp
t
tan
Plaxis Introductory Course, Da Nang 10  12 July 2013
d xy d xyp
d yy d yyp
tan xy
d xy d xyp
xy
yy
dilatancy xy
35 of 375
Plaxis Introductory Course, Danang, Vietnam
Model parameters:
Model
M d l parameters must beb determined
d i d such h that
h
real soil behaviour is approximated in the best possible way
36 of 375
Plaxis Introductory Course, Danang, Vietnam
Parameter determination
1
v
2 sin
1 sin
1
12
1 2
Plaxis Introductory Course, Da Nang 10  12 July 2013
37 of 375
Plaxis Introductory Course, Danang, Vietnam
Eoed
1
(1 )(1 2 )
Eoed E
(1 )
E50 3'
 Sand: 150..500 ref
p ref p
Loose Dense
15000 cu 5000 cu
 Clay:
u
E50 or G50
I p [%] I p [%]
Ip = plasticity index
38 of 375
Plaxis Introductory Course, Danang, Vietnam
Eoedd 1'
150..500 ref
p ref p
500
Eoed '1 (correlation)
Ip
pref Eoed
Eoed 3..5 qc (correlation)
1
d 1 1 E (1 )(1 2 )
Eoedd E Eoedd
d1 1 1 2 (1 )
Thiss Evalue
a ue applies
app es to
o primary
p a y compression
co p ess o
Plaxis Introductory Course, Da Nang 10  12 July 2013
39 of 375
Plaxis Introductory Course, Danang, Vietnam
1
Eoed
1
1(1)
Eoed
1 (0)
1
G
cu
40 of 375
Plaxis Introductory Course, Danang, Vietnam
41 of 375
Plaxis Introductory Course, Danang, Vietnam
42 of 375
Plaxis Introductory Course, Danang, Vietnam
ELASTOPLASTIC ANALYSIS OF A
FOOTING
Computational Geotechnics 1
43 of 375
Plaxis Introductory Course, Danang, Vietnam
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
44 of 375
Plaxis Introductory Course, Danang, Vietnam
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 perfectlyplastic 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
45 of 375
Plaxis Introductory Course, Danang, Vietnam
GEOMETRY INPUT
Start PLAXIS by doubleclicking 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
46 of 375
Plaxis Introductory Course, Danang, Vietnam
In the Project tabsheet, enter Exercise 1 in the Title box and type Elastoplastic
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 15node 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
47 of 375
Plaxis Introductory Course, Danang, Vietnam
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 subsoil
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 xaxis 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
48 of 375
Plaxis Introductory Course, Danang, Vietnam
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
49 of 375
Plaxis Introductory Course, Danang, Vietnam
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 crosshatched 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 xvalue and 0.0 as yvalue. These values are the
input load of point force B. Click OK to close the window.
8 Computational Geotechnics
50 of 375
Plaxis Introductory Course, Danang, Vietnam
& 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 MohrCoulomb Linear elastic
Type of behaviour Type Drained Nonporous
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 MohrCoulomb from the Material model combo box and Drained from the Material
type combo box.
Computational Geotechnics 9
51 of 375
Plaxis Introductory Course, Danang, Vietnam
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 MohrCoulomb 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
52 of 375
Plaxis Introductory Course, Danang, Vietnam
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
53 of 375
Plaxis Introductory Course, Danang, Vietnam
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
54 of 375
Plaxis Introductory Course, Danang, Vietnam
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
55 of 375
Plaxis Introductory Course, Danang, Vietnam
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
56 of 375
Plaxis Introductory Course, Danang, Vietnam
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 xaxis 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 = 1sin()).
> 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
57 of 375
Plaxis Introductory Course, Danang, Vietnam
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 yvalue to 50 kN/m and press the Ok
button.
16 Computational Geotechnics
58 of 375
Plaxis Introductory Course, Danang, Vietnam
The point load A is now active (blue) and has a load value of 50 kN/m.
Press Update.
Computational Geotechnics 17
59 of 375
Plaxis Introductory Course, Danang, Vietnam
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
60 of 375
Plaxis Introductory Course, Danang, Vietnam
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
61 of 375
Plaxis Introductory Course, Danang, Vietnam
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 xaxis we want to plot the settlement of our chosen point in the middle of the
footing. In the xaxis box choose point A from the dropdown list and then below in
Deformations and then Total displacements choose u.
On the yaxis 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 yaxis 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
62 of 375
Plaxis Introductory Course, Danang, Vietnam
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
63 of 375
Plaxis Introductory Course, Danang, Vietnam
Table 2: Results for the maximum load reached on a strip footing on the drained subsoil 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 15noded elements is superior over the performance of the lower
order 6noded elements. Needless to say that computation times are also influenced by the
number and type of elements.
22 Computational Geotechnics
64 of 375
Plaxis Introductory Course, Danang, Vietnam
ADDITIONAL EXERCISE:
UNDRAINED FOOTING
Computational Geotechnics 23
65 of 375
Plaxis Introductory Course, Danang, Vietnam
24 Computational Geotechnics
66 of 375
Plaxis Introductory Course, Danang, Vietnam
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 FEcodes 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
67 of 375
Plaxis Introductory Course, Danang, Vietnam
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 regenerate 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
68 of 375
Plaxis Introductory Course, Danang, Vietnam
Hint: After regeneration 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
69 of 375
Plaxis Introductory Course, Danang, Vietnam
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
70 of 375
Plaxis Introductory Course, Danang, Vietnam
Computational Geotechnics 29
71 of 375
Plaxis Introductory Course, Danang, Vietnam
30 Computational Geotechnics
72 of 375
Plaxis Introductory Course, Danang, Vietnam
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
73 of 375
Plaxis Introductory Course, Danang, Vietnam
Content
1. Learning
g objectives
j
2. Introduction
p
3. Multipliers
4. Iteration process
5. Plastic points
p
6. Recommendations
L
Learningobjectives
i bj i
I
Introduction
d i
1 Load multipliers
1.
I
Introduction
d i
1 Load multipliers
1.
2. Miscellaneous parms
I
Introduction
d i
1 Load multipliers
1.
2. Miscellaneous parms
3. Loaddispl. curve
I
Introduction
d i
1 Load multipliers
1.
2. Miscellaneous parms
3. Loaddispl. curve
4. Iteration process
I
Introduction
d i
1 Load multipliers
1.
2. Miscellaneous parms
3. Loaddispl. curve
4. Iteration process
5. Plastic points
L d
Loadmultipliers
l i li
Applied load = Load multiplier x Input load
Defaults:
1. Load multiplier = 1
2 Input
2. I t load
l d = 1 unit
it
Loading input:
1. Staged construction: Change Input load
2. Total multipliers: Change Load multiplier (M)
3. Incremental multipliers: Change Load multiplier (M)
T t l multiplier
Total lti li (phase)
( h ) = Sum
S off incremental
i t l multipliers
lti li (step)
( t )
L d
Loadmultipliers
l i li
MdispX
p : Tot. mult. p
prescribed xdisplacements
p
MdispY : Tot. mult. prescribed ydisplacements
MloadA : y
Tot. mult. loads system A
MloadB : Tot. mult. loads system B
Mweight
g : Tot. mult. soil & structural weights
g
Maccel : Tot. mult. pseudostatic acceleration
Msf : Tot. mult. Phic reduction process
Mstage : Tot. mult. stagedconstruction process
L d
Loadmultipliers
l i li I
Incrementalmultipliersinput
l l i li i
L d
Loadmultipliers
l i li T l
Totalmultipliersinput
l i li i
Mi ll
Miscellaneousparameters
PMax : Maximum ((excess)) p
pore pressure
p in the model
Marea : Relative part of the mesh area currently active
ForceX : Reaction force due to horizontal p
prescr. displ.
p
ForceY : Reaction force due to vertical prescribed displ.
Stiffness : Current ((relative)) Stiffness Parameter
Time : Elapsed model time (usually in days)
y
Dynamic time : Elapsed model time for dynamics
y ((s))
L d di l
Loaddisplacementcurve
1. Evaluation of calculation progress:
p g
1. Multipliers
2. Stiffness (CSP)
( )
3. Pmax
p
4. Loaddisplacement curve
5. Iterations
6. Global error
7. Plastic points
I
Iterationprocess
i
Calculation phase
Load steps (q)
Equilibrium iterations
constitutive model
q displacement strain stress reaction
Equilibrium?
I
Iterationprocess
i
Load q
Elastic stiffness (K)
qex Nonlinear
iterations behaviour
Unbalance
Load step q
qin
Settlement of Node A
I
Iterationprocess
i
Current step
p Max. step
p Additional steps
p
Iteration Max. iterations Maximum iterat.
Unbalance Global error Tolerance Tolerated error
Control parameters
I
Iterationprocess
i C
Controlparameters
l
I
Iterationprocess
i O
Overrelaxation
l i
A B
Overrelaxation
A
uo u
I
Iterationprocess
i A l
Arclengthcontrol
h l
I
Iterationprocess
i A l
Arclengthcontrol
h l
qP qP
0 0+ +
1 P = I Pe Pc I = const.
const
K
Pq0
P0 0 Arc length control
uo u
Plaxis Introductory Course, Danang 1012 July 2013 21 / 27
I
Iterationprocess
i D i d i i
Desiredminimum/maximum
/ i
I
Iterationprocess
i D i d i i
Desiredminimum/maximum
/ i
q
Converged within desired minimum Scaling up
number of iterations:
Scaling up load step by a
factor 2 Scaling down
Not converged within desired
maximum number of iterations:
Scaling down load step by a
Scaling up
factor 2
Pl i
Plasticpoints
i
1 Cap (HS, SS and SSCreep model)
Cap point
MohrCoulomb
failure surface
MohrCoulomb point
f<0
Shear h
Sh hardening
d i Cap & Hardening point
yield surface (HS model)
Hardening point
3
Apex point 1
Tension point
3
Tension cutoff: Principal tensile stress is set to zero
Plaxis Introductory Course,85 of 375 1012 July 2013
Danang 24 / 27
Plaxis Introductory Course, Danang, Vietnam
Pl i
Plasticpoints
i
Local error criterion: Constitutive stress c:
R
Recommendations
d i
1. Use mostly
y defaults
2. Monitor and evaluate calculation progress
g
3. In case of bad convergence or numerical failure, check input
4. Use output facilities to trace input errors
5. In case input is right, consider control parameters
6. Dont change control parameters without understanding
consequences!
7. Dont increase tolerated error to speed up convergence!
END
87 of 375
Plaxis Introductory Course, Danang, Vietnam
88 of 375
Plaxis Introductory Course, Danang, Vietnam
INTRODUCTION
INTRODUCTION
89 of 375
Plaxis Introductory Course, Danang, Vietnam
90 of 375
Plaxis Introductory Course, Danang, Vietnam
91 of 375
Plaxis Introductory Course, Danang, Vietnam
92 of 375
Plaxis Introductory Course, Danang, Vietnam
93 of 375
Plaxis Introductory Course, Danang, Vietnam
94 of 375
Plaxis Introductory Course, Danang, Vietnam
95 of 375
Plaxis Introductory Course, Danang, Vietnam
Do you need plasticity when unloading (back into the yield locus)?
Yes..if the accumulation plastic volumetric strains are important in cyclically loaded
soils..dynamic liquefaction related boundary value problems
96 of 375
Plaxis Introductory Course, Danang, Vietnam
97 of 375
Plaxis Introductory Course, Danang, Vietnam
98 of 375
Plaxis Introductory Course, Danang, Vietnam
99 of 375
Plaxis Introductory Course, Danang, Vietnam
100 of 375
Plaxis Introductory Course, Danang, Vietnam
101 of 375
Plaxis Introductory Course, Danang, Vietnam
102 of 375
Plaxis Introductory Course, Danang, Vietnam
103 of 375
Plaxis Introductory Course, Danang, Vietnam
104 of 375
Plaxis Introductory Course, Danang, Vietnam
105 of 375
Plaxis Introductory Course, Danang, Vietnam
106 of 375
Plaxis Introductory Course, Danang, Vietnam
107 of 375
Plaxis Introductory Course, Danang, Vietnam
108 of 375
Plaxis Introductory Course, Danang, Vietnam
109 of 375
Plaxis Introductory Course, Danang, Vietnam
110 of 375
Plaxis Introductory Course, Danang, Vietnam
111 of 375
Plaxis Introductory Course, Danang, Vietnam
112 of 375
Plaxis Introductory Course, Danang, Vietnam
113 of 375
Plaxis Introductory Course, Danang, Vietnam
114 of 375
Plaxis Introductory Course, Danang, Vietnam
115 of 375
Plaxis Introductory Course, Danang, Vietnam
116 of 375
Plaxis Introductory Course, Danang, Vietnam
20kPa
5m
3m
7m
3 analyses with Mohr Coulomb, Hardening Soil & Hardening SoilSmall models using equivalent
soil input parameters
117 of 375
Plaxis Introductory Course, Danang, Vietnam
0.004 Heave
0.002
Setttlement (m)
0.000
0.002
0.004
Settlement
0.006
MC
0.008 HS
HSsmall
0.010
118 of 375
Plaxis Introductory Course, Danang, Vietnam
0.015
eave (m)
0.010
He
0.005
0 000
0.000
0.005
MC HS HSsmall
Ux=6mm
Ux=11mm Ux=10mm
3 FOS=2.8
2.5
2
HS
1.5
Phic'
Phi c reduction
reduction for predicting FOS
3 FOS=2.8
FSP III sheetpile properties:
2.5 EI=34440 kNm2/m; EA=3
EA=3.9210
92106kN/m
2 HSsmall
Mp=369 kNm/m; Np=3575 kN/m
1.5
Summary of Predictions
Analyses Surface settlement Heave at Wall horizontal FOS for wall
behind wall excavation level displacement
p stability
y
MC Heave 4 mm Heave 20 mm 6 mm 2.8
(not OK)
HS Settle 9 mm Heave 11 mm 11 mm 2.8
HS
HSsmall
ll S l 9 mm
Settle H
Heave 8 mm 10 mm 28
2.8
120 of 375
Plaxis Introductory Course, Danang, Vietnam
A: unloading compression;
B: unloading extension
Rotation of stress paths at121Aof&375B
Plaxis Introductory Course, Danang, Vietnam
Exc.
t (kPa)
7m K0 20kPa B
5
0
5m
5
10
Failure line
A: unloading compression
15
B: unloading extension 0 10 20 30 40 50 60
s' (kPa)
Stress path rotation,
t
=0
=180
K0
=90
s'
Atkinson et al
al. (1990)
Triaxial tests on
London Clay
Shear strain (%)
1 0.1 0.01 0.01 0.1 1
=0,, no change
g in stress path
p direction
=180, full reversal of stress path direction
122 of 375
Plaxis Introductory Course, Danang, Vietnam
Extension
Compress
Compression
Extension
=90
123 of 375
Plaxis Introductory Course, Danang, Vietnam
Computational Geotechnics 1
124 of 375
Plaxis Introductory Course, Danang, Vietnam
2 Computational Geotechnics
125 of 375
Plaxis Introductory Course, Danang, Vietnam
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
126 of 375
Plaxis Introductory Course, Danang, Vietnam
4 Computational Geotechnics
127 of 375
Plaxis Introductory Course, Danang, Vietnam
Computational Geotechnics 5
128 of 375
Plaxis Introductory Course, Danang, Vietnam
6 Computational Geotechnics
129 of 375
Plaxis Introductory Course, Danang, Vietnam
Figure 6: Undrained triaxial (CU) tests at cell pressures of 100 kPa and 400 kPa
Computational Geotechnics 7
130 of 375
Plaxis Introductory Course, Danang, Vietnam
8 Computational Geotechnics
131 of 375
Plaxis Introductory Course, Danang, Vietnam
For a cell pressure 30 = 100 kPa a maximum value of approximately 10 30  = 400 kPa is reached at failure.
The MohrCoulomb 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 MohrCoulomb 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
132 of 375
Plaxis Introductory Course, Danang, Vietnam
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 unloadingreloading stiffness
can be determined. In the triaxial test results an unloadingreloading cycle is done for this. The Hardening Soil
model does not have unloadingreloading behaviour with hysteresis but simple nonlinear elastic unloading
reloading behaviour. Therefore a secant value is taken for the unloadingreloading 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
133 of 375
Plaxis Introductory Course, Danang, Vietnam
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
134 of 375
Plaxis Introductory Course, Danang, Vietnam
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
135 of 375
Plaxis Introductory Course, Danang, Vietnam
If no triaxial test with unloadingreloading is available the unloadingreloading stiffness can also be determined
from an oedometer test with unloading. However, the unloadingreloading 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
136 of 375
Plaxis Introductory Course, Danang, Vietnam
CLAY
We start with the determination of the strength parameters based on the CU triaxial tests.
The black dotted lines is the MohrCoulomb failure criterium in the pq 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 MohrCoulomb line in pq plane is defined as:
6sin 195
M= 3sin = 200 = 250
From the intersection between MohrCoulomb line and the vertical axis where p=0 the cohesion can be determined:
6c0 cos
q= 3sin = 0 c = 0 kPa
14 Computational Geotechnics
137 of 375
Plaxis Introductory Course, Danang, Vietnam
From the results of the oedometer test the oedometer stiffness as well as the unloadingreloading 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
138 of 375
Plaxis Introductory Course, Danang, Vietnam
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 25 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 K0value 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
139 of 375
Plaxis Introductory Course, Danang, Vietnam
Computational Geotechnics 17
140 of 375
Plaxis Introductory Course, Danang, Vietnam
18 Computational Geotechnics
141 of 375
Plaxis Introductory Course, Danang, Vietnam
In the following paragraphs a stepbystep 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
142 of 375
Plaxis Introductory Course, Danang, Vietnam
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 doubleclick these graphs to view them in separate windows. Furthermore, custom charts can be
added, see figure 4.
20 Computational Geotechnics
143 of 375
Plaxis Introductory Course, Danang, Vietnam
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 doubleclick this graphs to view them in separate windows
as well as add custom charts.
Computational Geotechnics 21
144 of 375
Plaxis Introductory Course, Danang, Vietnam
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
145 of 375
Plaxis Introductory Course, Danang, Vietnam
Figure 20: Unloading/reloading cycle in a triaxial test using the General option
Computational Geotechnics 23
146 of 375
Plaxis Introductory Course, Danang, Vietnam
DerivationofSoilParametersfrom
LabTestResults&Verificationin
PlaxisSoilTest
by
RFShen
23Nov2011
Singapore 2011
147 of 375
Plaxis Introductory Course, Danang, Vietnam
Part 1: Sand
Singapore 2011
148 of 375
Plaxis Introductory Course, Danang, Vietnam
Fa/A = q (deviatoric
stress)
Typical sample size 38 mm x 76 mm a = q + r
Singapore 2011
149 of 375
Plaxis Introductory Course, Danang, Vietnam
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
0.06
0.05
0.04
Volumetricstrain
0.03
0.02
0.01
Testdata
0
0 0.02 0.04 0.06 0.08 0.1
0.01
Axialstrain
Singapore 2011
150 of 375
Plaxis Introductory Course, Danang, Vietnam
Settlementdialgauge
OedometerCell
Sample:dia.=75mm
Protruded
leverarm Height=20mm
Heavydeadweights
Singapore 2011
151 of 375
Plaxis Introductory Course, Danang, Vietnam
Boundary
conditions
0
0.1
0.2
0.3
0.4
Verticalstrain(%)
0.5
0.6
0.7 Testdata
0.8
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 400
Verticalpressure(kPa)
Singapore 2011
152 of 375
Plaxis Introductory Course, Danang, Vietnam
400
300
Verticalpressure(kPa)
200
100
Testdata
0
0 50 100 150 200
Lateralstress(kPa)
Singapore 2011
153 of 375
Plaxis Introductory Course, Danang, Vietnam
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
Singapore 2011
154 of 375
Plaxis Introductory Course, Danang, Vietnam
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
pc p
Singapore 2011
155 of 375
Plaxis Introductory Course, Danang, Vietnam
450
400 400
350
Deviatorstress(kPa)
300
250
3 = 100 kPa
200
150
100 Testdata
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
0.013
Axialstrain
400
E50ref 30800 kPa 30000 kPa
0.013
Singapore 2011
156 of 375
Plaxis Introductory Course, Danang, Vietnam
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
Singapore 2011
157 of 375
Plaxis Introductory Course, Danang, Vietnam
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 ref
c cot ' p
Singapore 2011
158 of 375
Plaxis Introductory Course, Danang, Vietnam
0
0.1
320
29900kPa 30000kPa
0.2 ref
Eoed
0.3
0.4
1.4% 0.33%
0.47
Verticalstrain(%)
0.5
400
0.6 200 kPa
Eoed 43000kPa
0.7
0.8
Testdata
1.4% 0.47%
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 400
Verticalpressure(kPa) 400
m
c cot ' '1 200
200 kPa m
Eoed 43000
m = 0.5
c cot ' p
ref ref
Eoed 100 30000
Jakis formula:
300
Verticalpressure(kPa)
200
100
Testdata
0
0 50 100 150 200
Lateralstress(kPa)
x ' 100
K 0NC 0.33
y ' 300
Singapore 2011
159 of 375
Plaxis Introductory Course, Danang, Vietnam
Singapore 2011
160 of 375
Plaxis Introductory Course, Danang, Vietnam
Part 2: Clay
Singapore 2011
161 of 375
Plaxis Introductory Course, Danang, Vietnam
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)
Singapore 2011
162 of 375
Plaxis Introductory Course, Danang, Vietnam
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
Singapore 2011
163 of 375
Plaxis Introductory Course, Danang, Vietnam
Testdata
0.1
Verticalstrain(%)
0.2
0.3
0.4
0.5
1 10 100 1000
Verticalpressure(kPa)
Singapore 2011
164 of 375
Plaxis Introductory Course, Danang, Vietnam
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 ref
0.5
c cot ' p
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
Singapore 2011
165 of 375
Plaxis Introductory Course, Danang, Vietnam
100kPa
Testdata
0.1
when y =100kPa,
0.2
0.3 x<100kPa;
When y loaded to about
0.418
0.4
300kPa and unload to
0.427
0.5 100kPa, x is expected to be
30
closer to 100kPa. As such, we
1 10 100 1000
Verticalpressure(kPa) 120
can approximately accept the
derived Eur.
log(120) log(30)
gradient _ k 66.9 Eur 2.3 100 66.9 15000 kPa
0.427 0.418
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
Singapore 2011
166 of 375
Plaxis Introductory Course, Danang, Vietnam
Singapore 2011
167 of 375
Plaxis Introductory Course, Danang, Vietnam
Thank you!
Singapore 2011
168 of 375
Plaxis Introductory Course, Danang, Vietnam
1/35
Contents
Plane strain, Axisymmetry, 3D
Model boundaries
General considerations
Excavations
Shallow foundations
Embankments
Tunnels
Conclusions
References
169 of 375
Plaxis Introductory Course, Danang, Vietnam
Plane strain
Considerations:
One
O dimension
di i iis relatively
l ti l llong
Similar geometry and stress or loading conditions in any cross
section long dimension
Consequences:
No strain long dimension (stress can change!)
No shear stress and arching long dimension
Model represents 1 length unit long dimension
y
Plane strain
Examples:
170 of 375
Plaxis Introductory Course, Danang, Vietnam
Plane strain
NOT a planestrain situation:
45 m
30 m
45 m
8m
Axisymmetry
Considerations:
Geometry
G t isi circular
i l
Similar geometry and stress or loading conditions in any cross
section that includes the central axis
Consequences:
Stress and strain central axis are radial
Model represents 1 radian around central axis
171 of 375
Plaxis Introductory Course, Danang, Vietnam
Axisymmetry
Examples:
Axisymmetry
NOT an axisymmetric situation:
Gravity!
172 of 375
Plaxis Introductory Course, Danang, Vietnam
3D models
Considerations:
Do I really need a 3D model?
If yes, but still I use a 2D model:
q
What are the consequences?
Would this give conservative or optimistic results?
How large is the error?
3D models
173 of 375
Plaxis Introductory Course, Danang, Vietnam
3D models
3D models
174 of 375
Plaxis Introductory Course, Danang, Vietnam
3D models
Model boundaries
General considerations
175 of 375
Plaxis Introductory Course, Danang, Vietnam
Model boundaries
Stability analysis:
Model boundaries
Deformation analysis:
drained undrained
176 of 375
Plaxis Introductory Course, Danang, Vietnam
Model boundaries
Dynamic analysis:
~
~ ~ ~ ~
Model boundaries
Stability analysis
Drained
deformation analysis
Undrained
deformation analysis
~
~ ~ ~ ~
Dynamic analysis
177 of 375
Plaxis Introductory Course, Danang, Vietnam
limit
w depth
(0.1 to 0.2) accepted
178 of 375
Plaxis Introductory Course, Danang, Vietnam
179 of 375
Plaxis Introductory Course, Danang, Vietnam
l d
U HS
Use HSsmallll or b
bottom
tt llayer with
ith smallstrain
ll t i stiffness
tiff f Eurref (height
for (h i ht a))
180 of 375
Plaxis Introductory Course, Danang, Vietnam
TBM or
NATM excavation
w
D D
a a
a
Suggestions: Face stability: a D ; w 2D
Structural analysis: a D ; w 2D
Deformation analysis: a D ; w 3D
Use HSsmall or bottom layer with smallstrain stiffness for Eurref (height a)
181 of 375
Plaxis Introductory Course, Danang, Vietnam
Meshing
Type of element:
Two types of volume elements are available in Plaxis 2D:
node (ux, uy)
stress point (, )
x
x
x x
x
yaxis x x
x x x x
x x x x
i
xaxis
182 of 375
Plaxis Introductory Course, Danang, Vietnam
Meshing
Type of element (2D):
Which type of element in which situation?
Meshing
Type of element 3D: 3
12
1
9
8
6
1
4 7
10
2
15 2
3
14
5
4 6 11
13
5
3DT, 3DF: 15node wedge New Plaxis 3D: 10node tetrahedral
((quadratic interpolation)) ((quadratic interpolation))
Do not confuse 15node wedge in 3D (quadratic) with 15node triangle in 2D (4th order)!
183 of 375
Plaxis Introductory Course, Danang, Vietnam
Meshing
General considerations:
Hint:
Use local element size factors to make meshes fine near loads and
structures and coarse at model boundaries (local element size
factor may be larger than 1.0!).
Meshing
184 of 375
Plaxis Introductory Course, Danang, Vietnam
Conclusions
Conclusions:
Model size and boundaries depend, a.o., on type of analysis and type of
behaviour (stability analysis, drained deformation undrained deformation,
dynamic analysis).
References
Potts D.M., Zdravkovic L. (2001). Finite element analysis in geotechnical engineering
Application. Thomas Telford, London.
S h i
Schweiger H.F.
H F (2002).
(2002) Musterlsung
M t l und
dPParameterstudie
t t di f fr dreifach
d if h verankerte
k t B Baugrube,
b
Geotechnik 25, 101109.
Ruse N.M. (2003). Rumliche Betrachtung der Standsicherheit der Ortsbrust beim
Tunnelvortrieb. PhD thesis. Institut fr Geotechnik. Universitt Stuttgart.
Vermeer P.A., Wehnert M. (2005). Beispiele von FEAnwendungen Man lernt nie aus. In:
FEM in der Geotechnik (ed. Grabe et.al.). Technische Universitt HamburgHarburg.
Brinkgreve R.B.J, Bakker K.J., Bonnier P.G. (2006). The relevance of smallstrain stiffness
in numerical simulation of excavation and tunnelling projects. In: NUMGE 2006 (ed.
Schweiger). Taylor & Francis, London. 133139.
185 of 375
Plaxis Introductory Course, Danang, Vietnam
Presentation by
Dr William Cheang
Principal Geotechnical Consultant
Plaxis AsiaPac Pte Ltd
Contents
1. StructuralelementsavailableinPlaxis
2. UsageofstructuralelementsinFEmodelling
3. Plate elements(BeamandShellelement)
4. Anchor elements(Springelement)
5
5. Geotextile elements
elements (Membraneelement)
(Membrane element)
6. Interface elements(Zerothicknesselement)
186 of 375
2
1.Structural elements in Plaxis
Plaxis Introductory Course, Danang, Vietnam
4 Interfaceelement
4. Interface element
Section 3.44 & 14.1
187 of 375
strut anchored wall 4
3.1 Plate Element Plaxis Introductory Course, Danang, Vietnam
Overview:
1. 3or5nodedlineelements(for6nodedor15nodedelementmesh)
2. 3degreesoffreedompernode
3. Plateshave:
o Axialforces
o Shearforces
Shear forces
o Bendingmoments
o Hoopforces(axisymmetry)
4. Elasticorelastoplasticbehaviour
5. Formodellingwalls,floors,tunnels
Plates elasticparameters
h3 b
EI E (b = 1 m)
12
EA E h b (b = 1 m)
EI (Equivalent rectangular
d h 12
EA plate thickness)
h h
b
b = 1 m in plane strain
b = 1 meter in axisymmetry
188 of 375 b
6
3.3 Plate Element Plaxis Introductory Course, Danang, Vietnam
Plates elastoplasticbehaviour
Np
M
Mp
10 Elasticplate
600
N
400 15
200 20
189 of 375
0
25 8
200 0 200 400 600 800 1000 1200 M
M
3.5 Plate Element Plaxis Introductory Course, Danang, Vietnam
EffectonGlobalFOSbyc/phiReduction
ff l b l b / hi d i
1. Elastic wall excludes possibility of wall plastic hinge; and overestimate FOS=1.75
2. Allowing for wall plastic hinge (Elastoplastic
(Elasto plastic wall) gave lower FOS
FOS=1.40
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)
191 of 375
12
Illustration: Connection.P2D
3.9 Plate Element Plaxis Introductory Course, Danang, Vietnam
Walls thinwallvs.thickwall
Thinwall
h ll
Wallthickness<<walllength
Nomuchendbearing,onlyfriction
Plateelementsuffices
Thickwall
Wallthicknesssignificant
Endbearingcapacityneeded
Usesoilelementswithmaterialsetrepresen ngwallmaterial
Inordertoobtainstructuralforcesaplatewithfictitiousproperties
maybeinserted
b d
13
(Illustration: Beam.P2D): d
192 of 375
14
2.PlaxisAnchor Element
Introductory Course, Danang, Vietnam
Anchors fixedend
a) Tomodelsupports,anchorsandstruts
To model supports, anchors and struts
a) Elastoplasticspringelement
b) Oneendfixedtopointinthegeometry,otherendisfully
fixedfordisplacement
c) Positioningatanyangle
d) Prestressingoption
P i i
Anchors nodetonode
a) To
Tomodelanchors,columns,strutsandrods
model anchors columns struts and rods
a) Elastoplasticspringelement
b) Connectstwogeometrypointsinthegeometry
c)) Nointeractionwiththemeshalongtheanchorrod
h h h l h h d
d) Prestressingoption
15
4.1Anchor Element
Anchors materialproperties
Axialstiffness,EA
Axial stiffness EA (foroneanchor)
(for one anchor) [kN]
Spacing,Ls (outofplanedistancebetweenanchors) [m]
Maximumanchorforceforcompressionandtension,
Fmax,compandFmax,tens [kN]
Ls
193 of 375
16
4.2 Anchor Element
Plaxis Introductory Course, Danang, Vietnam
Anchors prestressing
DefinedinStagedconstructionphase
Bothtension(groutanchor)orcompression(strut)
possible
Tension = positive
17
5.1.Geogrid Element
Geogrids
1.
1 3or5nodedlineelement
3 5 d d li l t
2. Elasticorelastoplasticbehaviour
3. Noflexuralrigidity(EI),onlyaxialstiffness(EA)
4. Onlyallowsfortension,notforcompression
194 of 375
18
+ Geogrid Element
Plaxis Introductory Course, Danang, Vietnam
5.2 Anchor Element
Groundanchors
1. Combinationofnodetonodeanchorandgeogrid
2. Nodetonodeanchorrepresentsanchorrod(freelength)
(nointeractionwithsurroundingsoil)
3
3. G
Geogridrepresentsgroutedpart(fullinteractionwithsurroundingsoil)
id t t d t (f ll i t ti ith di il)
4. Nointerfacearoundgroutedpart;interfacewouldcreateunrealisticfailuresurface
5
5. Working load conditions onl nopullout
Workingloadconditionsonly no p llo t
6. Ifpulloutforceisknownthiscanbeusedbylimitinganchorrodforce
19
5.3Groundanchors
Axial force distribution along fixed length (modelled using geogrid)
Probableactualdistributionofaxial
forcesingroundanchor
axialforcesingeotextileelement
195 of 375
Input geometry Generated mesh
20
5.4GroundAnchors:Influenceofnodenumbers
Plaxis Introductory Course, Danang, Vietnam
21
6.1InterfaceElement
Interfaces materialproperties
1. Soilstructureinteraction
1. Wallfriction
2. Slipandgappingbetweensoilandstructure
2. Soilmaterialproperties
Soil material properties
A. TakenfromsoilusingreductionfactorRinter
3. Individualmaterialsetforinterfacepossible
196 of 375
22
6.2 Interface Element
Plaxis Introductory Course, Danang, Vietnam
Interfaces reductionfactor
SuggestionsforR
gg inter:
Interactionsand/steel =Rinter 0.6 0.7
Interactionclay/steel =Rinter 0.5
I t
Interactionsand/concrete
ti d/ t =R
Rinter 1.0
1 0 0.8
08
Interactionclay/concrete =Rinter 1.0 0.7
Interactionsoil/geogrid(groutedbody) =Rinter1.0
(interfacemaynotberequired)
Interactionsoil/geotextile =Rinter0.9 0.5(foil,textile)
23
References
1 Brinkgreve
1. Brinkgreve, R.,
R Engin,
Engin E,
E & Swolf,
Swolf WW. (2010)
(2010), Plaxis 2d 2010
197 of 375
24
Plaxis Introductory Course, Danang, Vietnam
25
198 of 375
Plaxis Introductory Course, Danang, Vietnam
TIEDBACK EXCAVATION
Using the HSsmall model
Computational Geotechnics 1
199 of 375
Plaxis Introductory Course, Danang, Vietnam
2 Computational Geotechnics
200 of 375
Plaxis Introductory Course, Danang, Vietnam
INTRODUCTION
A building pit was constructed in the south of the Netherlands. The pit is 15 m deep and 30 m
wide. A diaphragm wall is constructed using 100 cm diameter bored piles; the wall is anchored
by two rows of prestressed ground anchors. In this exercise the construction of this building
pit is simulated and the deformation and bending moments of the wall are evaluated.
The upper 40 m of the subsoil consists of a more or less homogeneous layer of medium dense
fine sand with a unit weight of 18 kN/m3 . Triaxial test data of a representative soil sample is
given in figure 2. Underneath this layer there is very stiff layer of gravel, which is not to be
included in the model. The groundwater table is very deep and does not play a role in this
analysis.
AIMS
Using interface elements
Prestressing of anchors
0 x 4 1
Stage 1
Secant wall
11 12
Stage 2
13 14 Anchor rods
Stage 3
7 8 15
Grout bodies
16 17
9 5 18 10
6
3 2
Computational Geotechnics 3
201 of 375
Plaxis Introductory Course, Danang, Vietnam
4 Computational Geotechnics
202 of 375
Plaxis Introductory Course, Danang, Vietnam
MATERIAL PARAMETERS
In this exercise the HSsmall model is used and the model parameters for the sand layer
have been extracted from the triaxial test data (see figure 2). The HSsmall model takes into
account the stressdependency of soil stiffness, elastoplastic behaviour under both compres
sion loading and shear loading and increased stiffness in areas with very low strain levels.
The soil parameters can be found in table 1, while the determination of the soil parameters
can be found in appendix A.
Secant wall
The secant wall consists of 100cm diameter bored piles with an intermediate distance of 80cm,
hence there is a 20cm overlap of the piles. This configuration is taken this into account for the
determination of the cross sectional area (A) and moment of inertia (I) per meter outofplane
(see Appendix B). The concrete stiffness is Ec =2.7107 kN/m2 with a specific weight =16
kN/m3, which leads to the material parameters as given in Table 2. The determination of the
stiffness parameters can be found in Appendix A.
Computational Geotechnics 5
203 of 375
Plaxis Introductory Course, Danang, Vietnam
Ground anchors
The anchors are made of 32mm diameter steel bars at an intermediate distance of 1m. The
steel bars have a stiffness of Es =2.1*108 kN/m2 . The anchors have an ultimate strength
of 605 kN per anchor. In combination with a secant wall the anchors may be prestressed
to a maximum level of 60% of the ultimate strength, hence up to 363 kN per anchor. The
maximum compression force of the anchor is not important as the anchors will not be loaded
under compression. The grout body that forms the bonded length of the anchor behaves
relatively weak under tension compared to the steel bar inside. Therefore it is assumed that
both stiffness and strength of the bonded part of the anchor are fully determined by the steel
bar. This leads to the material properties for both the anchor rod (free length) and grout body
(bonded length) as given in tables 3 and 4.
6 Computational Geotechnics
204 of 375
Plaxis Introductory Course, Danang, Vietnam
Computational Geotechnics 7
205 of 375
Plaxis Introductory Course, Danang, Vietnam
8 Computational Geotechnics
206 of 375
Plaxis Introductory Course, Danang, Vietnam
GEOMETRY INPUT
Start a new project
Project properties
Accept the default values in the Project tab sheet of the Project properties (15node
elements). For the dimensions see figure 3.
Geometry
(15,0)
(0,0) (70,0)
0 x 4 1
(0,5) 11 12
(0,10) 13 14
(30,15)
(0,15) 7 8 15
(37.5,20)
(30,20) 16 17
(0,25) (70,25)
9 5 18 10
6 (37.5,25)
(15,27)
(0,60) (70,60)
3 2
Click the Geometry line button and draw the geometry contour and soil layers as
specified in figure 4.
Computational Geotechnics 9
207 of 375
Plaxis Introductory Course, Danang, Vietnam
Click the Plate button and draw the secant wall from (15, 0) to (15, 25).
Click the Interface button and draw the interface from (15,0) to (15, 27) and back to
(15,0). This creates an interface on both sides of the secant wall.
Click the Nodetonode anchor button and insert both anchor rods. These anchors
connect the beginning of the grout bodies to the wall.
Finally, click the Geometry line button again to introduce the two levels of excavation.
Fixities
Material properties
Enter the material properties for the four soil data sets, as determined in table 1of this
exercise.
After entering all properties for the three soil types, drag and drop the properties to the
appropriate clusters.
Enter material properties for the plates, anchors and geogrids as indicated in tables 2,
3and 4.
10 Computational Geotechnics
208 of 375
Plaxis Introductory Course, Danang, Vietnam
Mesh generation
From the Mesh menu, set the Global coarseness to Medium and press the Generate
button. This will result in a mesh as shown in figure 5.
Select the geogrid and plate elements and press Refine line from the Mesh menu. This
will result in a refinement around the selected lines as shown in Figure 6.
Computational Geotechnics 11
209 of 375
Plaxis Introductory Course, Danang, Vietnam
12 Computational Geotechnics
210 of 375
Plaxis Introductory Course, Danang, Vietnam
CALCULATION
When starting the calculation program choose the Classical mode.
The entire construction process consists of five phases. Define the phases, as shown graph
ically below. For each phase, use the Plastic calculation, Staged construction.
Initial phase
For the initial phase choose the K0 procedure for calculating the initial stresses. As the phreatic
line is located below the geometry the generation of initial pore pressures can be skipped and
since its not necessary to switch off any soil for the initial situation it is not needed to define
the initial phase.
Phase 1
In the first phase, the diaphragm wall is activated and the first excavation takes place.
Note that though the the interfaces along the wall are activated automatically with the
activation of the wall, the extensions below the diaphragm wall have to be activated
manually.
Phase 2
In the second phase, a new option is used, namely the prestressing of anchors.
First the groutbody (the geogrid) is switched on by clicking on the geogrid element.
The element will appear in yellow as soon as it is switched on. The light grey colour
indicates nonactive elements.
Computational Geotechnics 13
211 of 375
Plaxis Introductory Course, Danang, Vietnam
Now that the groutbody is active, the anchor element needs to be prestressed. By
double clicking on a nodetonode anchor a window will appear as shown in figure 9.
Select the option Adjust prestress, fill in a prestress force of 300 kN/m (tension) and
press OK.
In the geometry a black nodetonode anchor indicates that the anchor is activated. The
letter P indicates that a prestress force will be active in the anchor.
Phase 3, 4 and 5
Now define the remaining phases according to figures 10, 11 and 12.
14 Computational Geotechnics
212 of 375
Plaxis Introductory Course, Danang, Vietnam
Figure 10: Phase 3: Second excavation Figure 11: Phase 4: Activation and
prestressing of 2nd anchor
Computational Geotechnics 15
213 of 375
Plaxis Introductory Course, Danang, Vietnam
16 Computational Geotechnics
214 of 375
Plaxis Introductory Course, Danang, Vietnam
INSPECT OUTPUT
The results of fase 5 is presented in Figure 13. After this final stage the excavation bottom
heave calculated is about 5 cm.
By double clicking on the nodetonode anchors, Plaxis will present a table, in which the
stress in all anchors may be inspected. Anchor forces are approximately 340 kN where
the lower anchor has a slightly higher anchor force than the upper anchor.
When doubleclicking on one of the geogrids the change of axial forces within the grout body
can be investigated. What is immediately noticeable is that the axial force at the connection
with the anchor rod is significantly lower than the force in the anchor rod itself. This is due the
fact that the end of the anchor rod is not only connected to the grout body, but also to several
soil elements surrounding the end of the anchor rod. Therefore part of the anchor force is
transferred directly to those soil elements while part of the anchor force is transferred to the
geotextile representing the grout body. The amount of force transferred to the soil depends on
the stiffness of the soil; in this exercise it is 2535% of the anchor force. However, this effect
has very little influence on other calculation results. That is, it is not so important for other
calculation results how the anchor rod transfers its force; directly to the soil or by means of the
grout body.
By doubleclicking on the wall the structural forces in the wall can be inspected. The
maximum bending moment should be in the order of 350 kNm/m (figure 14)
When doubleclicking on an interface only the results of part of the interface can be seen.
In order to see the results for the whole interface chain, keep Ctrl + Shift pressed on the
keyboard while doubleclicking on the interface. In figure 15the left side are the passive
earth pressures and the right side are the active earth pressures. It can be seen that
only a small part of the maximum passive earth pressures has been mobilized at this
stage.
Computational Geotechnics 17
215 of 375
Plaxis Introductory Course, Danang, Vietnam
Figure 14: Bending moments in the sec Figure 15: Effective normal stresses in the
ant wall interface
Geometry size
For any project the geometry has to be made sufficiently large so that the boudary conditions
have no influence on the calculation results. This means in practice that close to the boundar
ies (with exception of a axis of symmetry) displacements should be small and stresses should
be undisturbed. When using the HSsmall model there is an interesting plot that can be used
to check this.
From the Stresses menu choose the option State parameters and then G/Gur .
This plot shows the actual shear stiffness divided by the unloading/reloading shear stiffnes
at engineering strain level. For areas with very small deformations the stiffness will be high
(small strain stiffness) and so the value of G/Gur > 1. Hence, the geometry is sufficiently large
if next to the boundaries, with exception of the axis of symmetry, G/Gur > 1, which indeed is
the case.
Hint: State parameters are additional quantities that relate to the state of the
material in the current calculation step, taking into account the stress
history. Examples of state parameters are the isotropic overconsolidation
pressure (pp ) and the hardening parameter p that specifies the maximum
shear strain level reach in the stress history.
18 Computational Geotechnics
216 of 375
Plaxis Introductory Course, Danang, Vietnam
Surface settlements
In Plaxis Output it is possible to see calculation results in a userdefined cross section. This
feature will be used to check the surface settlements behind the secant wall.
Click the Cross section button . The Cross section points window appears, see figure
16.
It is possible to draw a cross section by hand and check in the Cross section points window
what the coordinates are of the start and end point of the cross section. However, it is also
possible to position the cross section at a specific location by defining the coordinates of the
start and end point manually.
Move the mouse to the Cross section points window and fill in the coordinates (15, 0.1)
for the first point and (70, 0.1) for the second point and press OK. This will create a
cross section from the secant wall until the right boundary of the model just below the
soil surface. The cross section will open in a new window.
From the Deformations menu select Total displacements and then u y to see the vertical
displacements of the soil surface. The maximum settlement is 1213 mm, see figure 17.
Computational Geotechnics 19
217 of 375
Plaxis Introductory Course, Danang, Vietnam
20 Computational Geotechnics
218 of 375
Plaxis Introductory Course, Danang, Vietnam
Strength parameters
1 3 = (1 + 3 ) sin + 2c cos
1 3
1 +3
= sin
370100
370+100
= sin
= 35o
= 30 = 5o
Computational Geotechnics 21
219 of 375
Plaxis Introductory Course, Danang, Vietnam
Stiffness parameters
Since excavation is considered in this exercise, the input of Youngs modulus E should be
based on unloading, rather than on primary loading. For the same reason, Poissons ratio
should also be based on unloading, which results in a somewhat lower value.
The triaxial test has a cell pressure 3 = 100 kPa. This corresponds with reference pressure,
so E50 = Eref
50 .
ref v 135
E50 = v
= 0.675%
= 2.0 104 kP a
22 Computational Geotechnics
220 of 375
Plaxis Introductory Course, Danang, Vietnam
Compared to the original bored piles the repetitive sections have a reduced cross sectional
area. Though it can be analytically derived how much the reduction is, the fastest way to
determine this is to draw the repetitive section on paper with a fine grid based on the original
bored piles with a diameter of 1000mm and an overlap of 200mm and count squares. Using
this method the cross sectional area of the repetitive section is determined as As = 0.74 m2 .
Since the sections are at a distance D apart where D is given as 800mm, the cross sectional
area of the wall per meter is given as:
Awall = ADs = 0.74
0.8
= 0.93 m2 /m
For the moment of inertia is assumed that the influence of the reduced cross sectional area
is negligble as the reduction is close to the axis of bending and symmetric. Therefore the
moment of inertia per meter wall is determined as:
Ipile 4 (0.5)4
Iwall = D
= r
4D
= 40.8
= 61.3 103 m4 /m
w = A = 16 0.93 = 15 kN/m/m
Computational Geotechnics 23
221 of 375
Plaxis Introductory Course, Danang, Vietnam
Contents
Drained / undrained (conditions and analysis)
Drained / undrained soil behaviour
Typical results from drained and undrained triaxial tests
Strength parameters
What is the critical case: drained or undrained?
Modeling undrained behaviour with Plaxis
Three methods
Effective stress analysis:
y how does it actually
y work
Undrained shear strength
Undrained behaviour with MohrCoulomb Model
Undrained behaviour with Hardening Soil Model
Influence of dilatancy
Summary
2
222 of 375
Plaxis Introductory Course, Danang, Vietnam
k = Permeability
k E oed Eoed = Oedometer modulus
T 2
t w = Unit weight of water
w D
D = Drainage length
t = Construction time
T = Dimensionless time factor
U = Degree of consolidation
4
223 of 375
Plaxis Introductory Course, Danang, Vietnam
Contents
Drained / undrained (conditions and analysis)
Drained / undrained soil behaviour
Typical results from drained and undrained triaxial tests
Strength parameters
What is the critical case: drained or undrained?
Modeling undrained behaviour with Plaxis
Three methods
Effective stress analysis:
y how does it actually
y work
Undrained shear strength
Undrained behaviour with MohrCoulomb Model
Undrained behaviour with Hardening Soil Model
Influence of dilatancy
Summary
6
224 of 375
Plaxis Introductory Course, Danang, Vietnam
8
225 of 375
Plaxis Introductory Course, Danang, Vietnam
Strength parameters
MohrCoulomb parameters in terms of effective stress (real soil behaviour)
c tan
c
3 ' s 1 '
1 3 3 c c
1 sin ; t s sin
2 2 tan tan 9
Strength parameters
MC parameters in terms of total stresses (only undrained conditions!)
Cu
1 3 1 3
2
F
2
F cu ,
Cu
Effective
ff stresses
t
1.
short
short long
2.
ESP
long
s, s
1
1.
ESP
s, s
For very soft NC soil, factor of safety against failure may be lower for
short term (undrained) conditions for unloading problems (e.g.
excavations)
For very stiff OC soil, factor of safety against
g failure may be lower for
short term (undrained) conditions for loading problems (e.g.
embankment) 12
227 of 375
Plaxis Introductory Course, Danang, Vietnam
Contents
Drained / undrained (conditions and analysis)
Drained / undrained soil behaviour
Typical results from drained and undrained triaxial tests
Strength parameters
What is the critical case: drained or undrained?
Modeling undrained behaviour with Plaxis
Three methods
Effective stress analysis:
y how does it actually
y work
Undrained shear strength
Undrained behaviour with MohrCoulomb Model
Undrained behaviour with Hardening Soil Model
Influence of dilatancy
Summary
13
Kw Eu 2 G 1 u
K total K'
n 31 2 u 31 2 u
E ' 1 u 3 ' B (1 2 ')
K total u
3 1 2 u 1 ' 3 B (1 2 ')
Notes:
Skempton Bvalue can be entered explicitely for undrained materials in order to
simulate effect of partially saturated soil on the effective and excess pore
pressures.
This procedure gives reasonable relation between u and B only for f < 0.35 !
Real value of Kw/n ~ 1.10 kPa (for n = 0.5)
6
15
uf
u
TSP
ESP
cu
s s
s,
single
g set of parameters
p in terms of effective stress (consistent)
( )
realistic prediction of pore pressures (if model is appropriate)
the undrained analysis can be followed by a consolidation analysis
Cu is a consequence of the model, not an input parameter!!
16
229 of 375
Plaxis Introductory Course, Danang, Vietnam
TSP=ESP
cu
s s
s, s
pa
parameters
a ete s in te
terms
soof tota
total st
stress
ess
no prediction of pore pressures (only total stresses are obtained)
the undrained analysis can not be followed by a consolidation analysis
Cu is an input parameter!!
17
ESP TSP
cu
s s
s, s
parameters
p in terms of total stress and effective stress
prediction of pore pressures (generally unrealistic)
the undrained analysis should not be followed by a consolidation
analysis (pore pressures unrealistic)
Cu is an input parameter!! 18
230 of 375
Plaxis Introductory Course, Danang, Vietnam
Contents
Drained / undrained (conditions and analysis)
Drained / undrained soil behaviour
Typical results from drained and undrained triaxial tests
Skempton's parameters A and B
Strength
g parameters
p
What is the critical case: drained or undrained?
Modeling undrained behaviour with Plaxis
Three methods
Effective stress analysis: how does it actually work
Undrained shear strength
Undrained behaviour with Mohr
MohrCoulomb
Coulomb Model
Undrained behaviour with Hardening Soil Model
Influence of dilatancy
Summary
20
231 of 375
Plaxis Introductory Course, Danang, Vietnam
cu,MC tan()
cu,real
MohrCoulomb
Real soil
1
cu c ' cos ' s 'sin ' c ' cos ' vo ho sin '
2
1
cu c ' cos ' 'v 0 1 K 0 sin
i '
2 21
HS_1 30 000 90 000 30 000 35 0 / 10 0.0 0.2 100 0.75 0.426 0.9
HS_2 50 000 150 000 50 000 35 0 0.0 0.2 100 0.75 0.426 0.9
HS_3
_ 15 000 45 000 15 000 35 0 0.0 0.2 100 0.75 0.426 0.9
HS_4 30 000 90 000 40 000 35 0 0.0 0.2 100 0.75 0.426 0.9
HS_5 30 000 90 000 15 000 35 0 0.0 0.2 100 0.75 0.426 0.9
HS_6 50 000 150 000 30 000 35 0 0.0 0.2 100 0.75 0.426 0.9
P
Parameters
t for
f MC M
Model
d l
see also Schweiger (2002)
2
E = 30 000 kN/m , = 0.2, = 35, = 0 and 10 23
HS_1
HS_2
125 HS_3
HS_4
HS 5
HS_5
100 HS_6
total stress path
q [kN/m ]
2
75
50
25
0
0.00 25.00 50.00 75.00 100.00 125.00 150.00
2
24
p' [kN/m ]
233 of 375
Plaxis Introductory Course, Danang, Vietnam
125
100
N/m ]
2
75
q [kN
50 HS_1
HS_2
HS_3
25 HS_4
HS_5
HS_6
0
0.00 1.00 2.00 3.00 4.00 5.00 6.00
25
1 [%]
70
excess pore pressure [kN//m ]
2
60
50
40
30
HS_1
HS_2
20
HS_3
HS_4
10 HS_5
HS_6
0
0.00 1.00 2.00 3.00 4.00 5.00 6.00
26
1 [%]
234 of 375
Plaxis Introductory Course, Danang, Vietnam
Influence of dilatancy
if we set 0 then, negative volumetric plastic deformations
occur at failure:
v ve vp (elasticplastic
(elastic plastic behavior)
v 0 (undrained conditions)
vp 0 ve 0 p ' K ve 0
At failure: q M p ' q 0
Influence of dilatancy
Simulation of undrained triaxial compression test MC / HS model  q vs 1
300
275
250
225
200
175
N/m ]
2
150
q [kN
125
100
75
MC non dil
50 MC dil
HS_1 non dil
25 HS_1 dil
0
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00
28
1 [%]
235 of 375
Plaxis Introductory Course, Danang, Vietnam
Influence of dilatancy
Simulation of undrained triaxial compression test MC / HS model  q vs p
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
29
p' [kN/m2]
Influence of dilatancy
Simulation of undrained triaxial compression test MC / HS model  pw vs 1
100
90 MC non dil
MC dil
80
HS_1 non dil
excess pore pressure [kN//m ]
2
70 HS_1 dil
60
50
40
30
20
10
10
20
20
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00
30
1 [%]
236 of 375
Plaxis Introductory Course, Danang, Vietnam
Summary
Undrained analysis should be performed in effective stresses and
with
ith effective
ff ti stiffness
tiff and
d strength
t th parameters
t (Method
(M th d A)
Undrained shear strength is result of the constitutive model
The MohrCoulomb model generally overestimates the undrained
shear strength in a Method A calculation. This can be solved by
doing a Method B analysis, but this is a trick that generally
generates incorrect excess pore pressures
One should not use dilatancy in an undrained analysis
31
References
Atkinson, J.H., Bransby, P.L. (1978)
The Mechanics of Soils,
Soils An Introduction to Critical State Soil Mechanics
Mechanics. McGraw Hill
Ortigao, J.A.R. (1995)
Soil Mechanics in the Light of Critical State Theories An Introduction. Balkema
Schweiger, H.F. (2002)
Some remarksk on pore pressure parameters A andd B ini undrained
d i d analyses
l with
i h the
h Hardening
d i Soil il
Model. Plaxis Bulletin No.12
Skempton, A.W. (1954)
The PorePressure Coefficients A and B. Geotechnique, 4, 143147
Vermeer, P.A., Meier, C.P. (1998)
Proceedings Int. Conf. on SoilStructure Interaction in Urban Civil Engineering, Darmstadt, 177191
32
237 of 375
Plaxis Introductory Course, Danang, Vietnam
CONTENTS
A. Introduction
1. Groundwater in Geotechnical Engineering
2
2. Plaxis
B. Definitions Porewater Pressures in Plaxis
1. Active
2
2. Steadystate
3. Excess
C. Generation of Porewater Pressures in Plaxis
1.
1 Porewater Pressure due to Hydrostatic Condition
2. Pore Pressures due to Groundwater Flow (Steady or Transient States)
D. Hydraulic models
1.
1 Fully Saturated Soils
2. Partially Saturated Soils
E. Case Histories
1. Excavations
2. Embankments and Dams
3. Slopes
F. References
Groundwater Analysis
A.
A Geotechnical problems are related to groundwater
B. Two extreme conditions of porewater response are normally considered,
they are:
1.
1 Drained
D i d
2. Undrained (Method A, B & C)
C. Real soil behaviour is related to time , i.e. transient, with the porewater
pressure being dependent on imposed:
1. Permeability
2. Rate of loading
3. Hydraulic boundary
D. The interstitial voids of the soil skeleton can be fully or partially filled with
pore fluid and therefore effective stresses are influenced by this action
E. This lecture will look into the following issues:
1. The setup of pore pressures in Plaxis
2. Input parameters
3. Some examples of groundwater regimes
2. Advanced
3
3. Fl
Flow
B
B. Active (Total) = Steady
Steadystate
state + Excess
1. ACTIVE porewater pressures is combination of STEADYSTATE and EXCESS porewater
pressures (see Reference Manual 5.9).
2. Steadystate pre pressures are generated due to water conditions (hydraulic boundaries)
assigned to soil clusters (layers)
B. CALCULATION MODES
B1.CLASSICAL MODE
A. Steadystate
Steady state pore pressures
1. Phreatic lines
2. Steadystate groundwater flow analysis
3. Transientstate g
groundwater flow analysis
y =
Steadystate background pore pressure
B. Excess pore pressures
1. Undrained material type in combination with
Pl ti calculation
Plastic l l ti
2. Consolidation analysis
INPUT KERNEL
St d State
Steady St t E
Excess Porewater
P t D f
Deformation
ti
Note:
Note:
1. Undrained Analysis
1 Hydrostatic
1.
2. Consolidation Analysis
2. GWF calculation
3. Excess (Soil Model)
B2.ADVANCED MODE
A. Consolidation analysis
B
B. Transient groundwater flow analysis
KERNEL
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
2. Cluster Dry
Cluster: Dry
Cl ster Dry
Cluster Dr
Interpolated
Case Histories
1. y conditions:
Boundary
2. Soil permeabilities
Steadystate flow
3 28 29 6 9 30 31 2
4 8 11 5
General General
General
16 17
21 26
19
9 18
8
20 23 24 27
22 25
13 14 12
15 7 10
0 1
1. Qualitative evaluation:
Flow field
Location of phreatic line
2. Quantitative evaluation:
Heads, pore pressures compared to hydrostatic,
Compare with measurements or field experience
krel
1
hp = 
hp = 0
m
0 hp
1 g n
S ( h p ) S res ( Ssat Sres ) 1 g a h p
gn ( g )
n
2
g n 1
g n
g n
krel S Se l 1 1 Se n
g g 1
with S Sres
Se
Ssat Sres
Ssat,S
Sres: saturated and residual saturation
ga, gn and gl: curve fitting parameters
Plaxis Introductory Course, July 1012,
36
Danang.
255 of 375
Plaxis Introductory Course, Danang, Vietnam
1 if hp 0
Linear in Saturation
hp
S hp 1 iff hps h p 0
h ps
0 if hp hps
1 if hp 0
4h p Loglinear in Permeability
h
k rel h p 10 ppk if h pk h p 0
4
10 if h p h pk
Upper soils:
U il
< 1m below soil surface
Lower soils:
all deeper
p soils
Hydraulic Properties of
European Soils
Particle distribution:
< 2m
2m  50m
50m 2mm
Particle distribution:
< 2m
2m  50m
50m 2mm
12 soils data sets
No difference between
upper and lower soils
Relative permeability
Degree of saturation
REFERENCES
A. Galavi, V. (2010), Groundwater flow, fully coupled flow deformation and undrained analyses in
Plaxis 2D and 3D. Technical Report, Plaxis B.V.
263 of 375
Plaxis Introductory Course, Danang, Vietnam
Computational Geotechnics 1
264 of 375
Plaxis Introductory Course, Danang, Vietnam
2 Computational Geotechnics
265 of 375
Plaxis Introductory Course, Danang, Vietnam
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 nodetonode anchor
and a geogrid (yellow line). The geogrid simulates the grout body whereas the nodetonode anchor simulates
the anchor rod. The diaphragm wall is modelled as a plate. The interfaces around the plate are used to model
soilstructure 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
266 of 375
Plaxis Introductory Course, Danang, Vietnam
(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 smallstrain 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
267 of 375
Plaxis Introductory Course, Danang, Vietnam
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
268 of 375
Plaxis Introductory Course, Danang, Vietnam
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
269 of 375
Plaxis Introductory Course, Danang, Vietnam
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.
Doubleclick on the top nodetonode anchor, the properties window for the nodetonode 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
270 of 375
Plaxis Introductory Course, Danang, Vietnam
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 steadystate 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
271 of 375
Plaxis Introductory Course, Danang, Vietnam
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 steadystate 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
272 of 375
Plaxis Introductory Course, Danang, Vietnam
Though the phreatic level in the excavation doesn't change, the ow eld is not steadystate 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
273 of 375
Plaxis Introductory Course, Danang, Vietnam
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 steadystate 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 steadystate
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 steadystate ow and transient ow.
Computational Geotechnics 11
274 of 375
Plaxis Introductory Course, Danang, Vietnam
Figure 5: Bending moments in the wall for the steady state ow analysis (left) and the transient ow analyis
(right)
12 Computational Geotechnics
275 of 375
Plaxis Introductory Course, Danang, Vietnam
CONTENTS
1
1. Ko Procedure
KoProcedure
2. Gravity Switch On
276 of 375
2
Plaxis Introductory Course, Danang, Vietnam
INITIAL STRESSES
1. Initial stresses represent the equilibrium sttate of the undisturbed soil and consist of:
a) Soil weight
b) Loading history
a) K0 procedure
b) Gravity loading
277 of 375
Plaxis Introductory Course, Danang, Vietnam
K0PROCEDURE
1. Generation of initial stresses during Initial Phasse Material Set
'h 'v K 0
GRAVITY LOADING
1. Calculation of initial stresses by weight
loading.
2. Disadvantage: Nonphysical displacements
are created.
3. Advantage: Equilibrium satisfied in all cases.
For 1D compression:
'n 'v Nonphysical displacements
1 reset in subsequent phase)
so
K0
1
278 of 375
Plaxis Introductory Course, Danang, Vietnam
GRAVITY LOADING
1. Procedure
a. Initial phase
a Skip K0 procedure,
a. procedure Mweight remains zero
b. Phase 1
a. Choose Plastic calculation, Total multip
pliers
b Set
b. S t weight
i ht multiplier
lti li Mweight
M i ht = 1
c. Phase 2
a. Select Reset displacements to zero to discard all displacements from raising the gravity
GRAVITY LOADING
Points
1. Undrained material
279 of 375
Plaxis Introductory Course, Danang, Vietnam
GRAVITY LOADING
Cases where gravity loading should be ussed instead of K0procedure:
280 of 375
Plaxis Introductory Course, Danang, Vietnam
SPECIAL CASES
Initial stresses
SPECIAL CASES
1. For complex initial situations like inn
ner city building projects it may be
needed to use several calculation ph hases to model the current situation
b f
before starting
t ti theth actual
t l project.
j t
existing buildings
our
project
our project
initial p ase 1
pha pphase 2 our p
project
j
reset displacements
281 of 375
Plaxis Introductory Course, Danang, Vietnam
PlaxisIntroductoryCourse,Danang1012July2013 1
Main advantages:
1. Requires no apriori assumptions on the failure mechanism
2. Critical surface is found automatically as slope failure occurs naturally through the zones due to
insufficient shear strength to resist shear stresses.
3. No requirement of assumptions on the interslice shear force distribution
4. Applicable to complex conditions
5. Information such as stresses, movements and porepressures and numerical tool as for
deformation analysis
6
6. Powerful alternative approach
PlaxisIntroductoryCourse,Danang1012July2013 2
282 of 375
Plaxis Introductory Course, Danang, Vietnam
PRELIMINARIES
PHICREDUCTION
PlaxisIntroductoryCourse,Danang1012
3
July2013
Safetyfactor
Many possible definitions
Manypossibledefinitions
a v a ila b le s o il r e s is ta n c e
1 .8
m o b iliz e d s o il r e s is ta n c e
fa ilu r e lo a d
5 .9
w o r k in g lo a d
PLAXIS:safetyfactoronsoilresistance
f f
PlaxisIntroductoryCourse,Danang1012
4
July2013
283 of 375
Plaxis Introductory Course, Danang, Vietnam
FactorofSafety(PhiCReduction
Phi/creduction:
a. Reductionofstrengthparameterscandtan()until
failure is reached.
failureisreached.
b. Thefactorofsafetyistheratioofinitialandreduced
strength
c tan
Msf
creduced tan reduced
c
cm
FS
tan
t m
tan
FS
PlaxisIntroductoryCourse,Danang1012 5
July2013
PlaxisIntroductoryCourse,Danang1012July2013 6
284 of 375
Plaxis Introductory Course, Danang, Vietnam
1.12
1.12
SumMsf
umMsf
1.08
1 08
1.08
S
Su
1.04
1.04
11.00
1.0
0.0 0.3 0.6 0.9 1.2 1.5 0.3 0.6 0.9 1.2
0.0
displacement
displacement
PlaxisIntroductoryCourse,Danang1012
7
July2013
PlaxisIntroductoryCourse,Danang1012July2013 8
285 of 375
Plaxis Introductory Course, Danang, Vietnam
PHICREDUCTION
PlaxisIntroductoryCourse,Danang1012
9
July2013
(13.737,5) (25,5)
Clay
((10
10,0)
0) t =16kN/m3
(0 0)
(0,0) 20
20
c=su =20kN/m2
=0
Eu/s
/ u =500
500
PlaxisIntroductoryCourse,Danang1012July2013 10
286 of 375
Plaxis Introductory Course, Danang, Vietnam
Su/(H)
/( H)
PlaxisIntroductoryCourse,Danang1012July2013 11
Su/(H)
/( )
5m Clay
20
10
10m t =16kN/m
k / 3
c=su =20t/m2
0.165
D/H=2;slopeangle=20
Chart:1/Nf =0.165,Nf =6.061
Nfield =H/s
H/ u =(16)(5)/20=4
(16)(5)/20 4
FS=6.061/4=1.515
PlaxisIntroductoryCourse,Danang1012July2013 12
287 of 375
Plaxis Introductory Course, Danang, Vietnam
PHICREDUCTION
PlaxisIntroductoryCourse,Danang1012
13
July2013
AutoSLOPE: LEM
SpencersMethod:
Classicalmethod: FS=1.547
Taylorsstabilitychart
FS=1.515
PlaxisIntroductoryCourse,Danang1012July2013 14
288 of 375
Plaxis Introductory Course, Danang, Vietnam
AutoSLOPE: Comparisons
AutoSLOPE:Comparisons
PlaxisIntroductoryCourse,Danang1012
15
July2013
PlaxisIntroductoryCourse,Danang1012July2013 16
289 of 375
Plaxis Introductory Course, Danang, Vietnam
Factor of safety
Plaxis:FS=1.537
Classicalmethod:
Taylorsstabilitychart
y y
FS=1.515
SpencersMethod:
FS=1.547
PlaxisIntroductoryCourse,Danang1012July2013 17
PlaxisIntroductoryCourse,Danang1012July2013 18
290 of 375
Plaxis Introductory Course, Danang, Vietnam
Plaxis:FS=1.537
PlaxisIntroductoryCourse,Danang1012July2013 19
Comparisons
Plaxis:FS=1.537
SpencersMethod:
FS=1.547
PlaxisIntroductoryCourse,Danang1012July2013 20
291 of 375
Plaxis Introductory Course, Danang, Vietnam
(20,10) (30,10)
Soil(drained)
Soil (drained)
t =16kN/m3
(0,5) (10,5)
E=10000kN/m2
c =5kN/m2
=20
(0,0) (30,0)
PlaxisIntroductoryCourse,Danang1012July2013 21
SpencersMethod:
FS=1.409
PlaxisIntroductoryCourse,Danang1012July2013 22
292 of 375
Plaxis Introductory Course, Danang, Vietnam
AutoSLOPE: Comparisons
AutoSLOPE:Comparisons
PlaxisIntroductoryCourse,Danang1012
23
July2013
Modeling
Excavate1
Excavate 2
Excavate2 Excavation
Excavation
problem
Fill2
Fillproblem
Fill 1
Fill1
PlaxisIntroductoryCourse,Danang1012July2013 24
293 of 375
Plaxis Introductory Course, Danang, Vietnam
PlaxisIntroductoryCourse,Danang1012July2013 25
Defaultvalue:
K0=1sin v =z
(Jakyseqn.) v=
vu
u
h=K0v
PlaxisIntroductoryCourse,Danang1012July2013 26
294 of 375
Plaxis Introductory Course, Danang, Vietnam
Analysis Stage
stage1
excavation1
stage2
excavation2
PlaxisIntroductoryCourse,Danang1012July2013 27
Excavation1
Deformed
mesh:h
excavation 1,2
,
Excavation2
PlaxisIntroductoryCourse,Danang1012July2013 28
295 of 375
Plaxis Introductory Course, Danang, Vietnam
Factor of safety:
e ca ation
excavation
Plaxis:FS=1.373
SpencersMethod:FS=1.409
Failure
mechanism
PlaxisIntroductoryCourse,Danang1012July2013 29
Plastic
p
points
PlaxisIntroductoryCourse,Danang1012July2013 30
296 of 375
Plaxis Introductory Course, Danang, Vietnam
Comparisons
Plaxis: FS = 1 373
Plaxis:FS=1.373
Spencer sMethod:
Spencers Method:
FS=1.409
PlaxisIntroductoryCourse,Danang1012July2013 31
Fill2
Fill1
PlaxisIntroductoryCourse,Danang1012July2013 32
297 of 375
Plaxis Introductory Course, Danang, Vietnam
Elementsnotusedintheinitialgeometry
Elementsnotused
intheinitialgeometry
PlaxisIntroductoryCourse,Danang1012July2013 33
PlaxisIntroductoryCourse,Danang1012July2013 34
298 of 375
Plaxis Introductory Course, Danang, Vietnam
Stage1:Fill1
Stage
construction
Stage 2 Fill2
Stage2:Fill2
PlaxisIntroductoryCourse,Danang1012July2013 35
Fill1
Deformed
mesh: h
Fill 1,2
12
Fill2
PlaxisIntroductoryCourse,Danang1012July2013 36
299 of 375
Plaxis Introductory Course, Danang, Vietnam
Factor of safety:
Fill problem
bl
Plaxis:FS=1.376
Plaxis FS 1 376
SpencersMethod:FS=1.409
Failure
mechanism
PlaxisIntroductoryCourse,Danang1012July2013 37
Plastic
points
PlaxisIntroductoryCourse,Danang1012July2013 38
300 of 375
Plaxis Introductory Course, Danang, Vietnam
Comparisons
Plaxis:FS=1.376
SpencersMethod:
FS=1.409
PlaxisIntroductoryCourse,Danang1012July2013 39
Excavation 2
Excavation2
Fill 2
Fill2
PlaxisIntroductoryCourse,Danang1012
40
July2013
301 of 375
Plaxis Introductory Course, Danang, Vietnam
Excavation:FS
Fill:FS
PlaxisIntroductoryCourse,Danang1012
41
July2013
Excavation:FS
Fill:FS
PlaxisIntroductoryCourse,Danang1012
42
July2013
302 of 375
Plaxis Introductory Course, Danang, Vietnam
PHICREDUCTION
PlaxisIntroductoryCourse,Danang1012
43
July2013
Note:
1 Griffiths & Lane (1999)
Model Slope G2 (757 elements15n) : 2 Bishop & Morgenstern (1960)
FOS= 1.323 (1.4001, 1.7522,1.2794,1.3795, 1,3756) 3.Taylor (1937)
4.Janbu
5.Bishop
6 MorgensternPrice
6.Morgenstern Price
PlaxisIntroductoryCourse,Danang1012July2013 44
303 of 375
Plaxis Introductory Course, Danang, Vietnam
Cu2
Cu2/Cu1 = 2
2.0
1.8
1.6
1.4
Factor of safety)
1.2
Model Slope G3A cu2/cu1=1 (3436 elements Model Slope G3D cu2/cu1=0.8 (3436 elements
1.0
15n) 15n)
0.8
FOS 1.428
FOS= 1 428 (1.47
(1 471, 1.47
1 473) FOS= 1.384 (1.451) 0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
Cu2 /Cu1)
Note:
Model Slope G3B cu2/cu1=0.6 (3436 elements Model Slope G3F cu2/cu1=0.5 (3436 elements 1 Griffiths & Lane (1999)
15n) 15n) 2 Bishop & Morgenstern (1960)
FOS= 1.319(1.401, 1.404) FOS= 1.112(1.251) 3 T l (1937) ((green liline))
3.Taylor
4.Janbu
Model Slope G3E cu2/cu1=0.4 (3436 elements Model Slope G3c cu2/cu1=0.2 (3436 elements
15n) 15n) :
FOS 0.903(1.05
FOS= 0 903(1 051) FOS= 0.470 (0.591, 1.304)
PlaxisIntroductoryCourse,Danang1012July2013 46
304 of 375
Plaxis Introductory Course, Danang, Vietnam
EffectonGlobalFOSbyc/phiReduction
1. Elastic wall excludes possibility of wall plastic hinge; and overestimate FOS=1.75
2
2. All i for
Allowing f wallll plastic
l ti hinge
hi (El
(Elastoplastic
t l ti wall) ll) gave llower FOS
FOS=1.40
1 40 and
d smaller
ll soilil yielded
i ld d zone
behind the wall
PlaxisIntroductoryCourse,Danang1012July2013 47
Some points
1. Always inspect the incremental displacements or strains as computed in the last
load step to make sure that failure is reached.
3. Mesh: Refine and redo the phic analysis until the factor of safety remains constant
upon further refinement of the mesh.
4 Al
4. Always use th
the arclength
l th titime stepping
t i procedure
d within
ithi th
the Phi C reduction
PhiC d ti
(default)
PlaxisIntroductoryCourse,Danang1012July2013 48
305 of 375
Plaxis Introductory Course, Danang, Vietnam
Appendix:
Appendix:2
2Dvs.3
Dvs.3D(Benz)
1 3Danalysismayyieldsubstantiallydifferentresultsfrom2Danalysis
1. 3D analysis may yield substantially different results from 2D analysis
2. TheadvantageofFEMoverclassicaldesigntoolsisobvious.I
3. ntheexample:stabilityofabentoniteslurrytrench.
PlaxisIntroductoryCourse,Danang1012
49
July2013
TheCaseofGriffithandMarquez2007(Case1)
PlaxisIntroductoryCourse,Danang1012
50
July2013
306 of 375
Plaxis Introductory Course, Danang, Vietnam
Incremental disp
PlaxisIntroductoryCourse,Danang1012
51
July2013
SlipSurface:Spoonunderisosurfaces
p p
L/H=1 L/H=2 L/H=4
L/H=8 L/H=12
PlaxisIntroductoryCourse,Danang1012
52
July2013
307 of 375
Plaxis Introductory Course, Danang, Vietnam
Case:InfluenceofSlopeLength/HeightratioonFOSin3D
PlaxisIntroductoryCourse,Danang1012
53
July2013
Case2:InclinedSideFace(tobecontinued)
PlaxisIntroductoryCourse,Danang1012
54
July2013
308 of 375
Plaxis Introductory Course, Danang, Vietnam
References
1
1. M
Matsui, i T
T. & San,
S K.C.
K C (1992) Finite
Fi i element
l slope
l stability
bili analysis
l i b by shear
h strength
h reduction
d i technique.
h i S il and
Soils d
Foundation, Vol.32 (1),pp.5970
2. Zienkiewicz, O.C., Humpheson, C. & Lewis, R.(1975) Associated and nonassociated viscoplasticity and plasticity in
soil mechanics. Geotechnique 25(4).pp. 671689.
3
3. Ugai K.(1989).
Ugai, K (1989) A method of calculation of total safety factor of slope by elastoplastic FEM.
FEM Soils and Foundation 29(2).
29(2)
pp.190195.
4. Farias, M.M., Naylor, D.J.(1998). Safety analysis using finite elements. Computer and Geotechnics, Vol 22(2) p.p. 165
181.
5. Griffiths, D.V., Lane, P.A. (1999). Slope stability analysis by finite elements. Geotechnique 49 (3), pp.387403.
pp.387 403.
6. Griffith, D.V. & Marquez, R,M (2007). Threedimensional slope stability analysis by elastoplastic finite elements,
Geotechnique 57, No. 6, 537546.
PlaxisIntroductoryCourse,Danang1012July2013 55
309 of 375
Plaxis Introductory Course, Danang, Vietnam
Computational Geotechnics 1
310 of 375
Plaxis Introductory Course, Danang, Vietnam
2 Computational Geotechnics
311 of 375
Plaxis Introductory Course, Danang, Vietnam
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 socalled 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 15node elements as basic element type since in this exercise
we will deal with failure behaviour.
Computational Geotechnics 3
312 of 375
Plaxis Introductory Course, Danang, Vietnam
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
313 of 375
Plaxis Introductory Course, Danang, Vietnam
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
314 of 375
Plaxis Introductory Course, Danang, Vietnam
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
315 of 375
Plaxis Introductory Course, Danang, Vietnam
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 nonhorizontal 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
316 of 375
Plaxis Introductory Course, Danang, Vietnam
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
317 of 375
Plaxis Introductory Course, Danang, Vietnam
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
318 of 375
Plaxis Introductory Course, Danang, Vietnam
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
Loaddisplacement curves
Before starting the calculation choose some points for nodedisplacement 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
319 of 375
Plaxis Introductory Course, Danang, Vietnam
INSPECT OUTPUT
Factors of safety
Check the factors of safety in the Curves program. Create a new curve of displacements vs. SumMsf 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 913. 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
320 of 375
Plaxis Introductory Course, Danang, Vietnam
Figure 10: Most critical failure mechanisme after construction, summer conditions.
Figure 11: Most critical failure mechanisme after construction, winter conditions.
12 Computational Geotechnics
321 of 375
Plaxis Introductory Course, Danang, Vietnam
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
322 of 375
Plaxis Introductory Course, Danang, Vietnam
Content
Overview of soil models,
models parameters,
parameters possibilities & limitations
in PLAXIS 3D
MohrCoulomb model
Hardening Soil model (HS + HSsmall)
Soft Soil model
Soft
S f S
Soil C
Creep model
HoekBrown model
Standard soil tests with different models
Which model in which situation?
323 of 375
Plaxis Introductory Course, Danang, Vietnam
MohrCoulomb model
Li
Linearelastic
l ti perfectlyplastic
f tl l ti b behaviour
h i
Hookes law:
d xx 1 0 0 0 d xx
d 1 0 0 0 d yy
yy
d zz E 1 0 0 0 d zz
d xy (1 )(1 2 ) 0 0 0 1
2 0 0 d xy
d yz 0 0 0 0 1
0 d yz
2
d zx 0 0 0 0 0 1
2 d zx
324 of 375
Plaxis Introductory Course, Danang, Vietnam
MohrCoulomb model
Linear elastic perfectlyplastic
Linearelastic perfectly plastic behaviour
Yield function:
f 12 '3 '1 12 '3 '1 sin ' c 'cos
cos '
MohrCoulomb model
Parameters:
E Youngs modulus
Poissons
Poisson s ratio
c Cohesion
Friction angle
Dilatancy angle
325 of 375
Plaxis Introductory Course, Danang, Vietnam
MohrCoulomb model
Possibilities:
First order approach of soil behaviour in general
(Drained) failure behaviour quite well described
Limitations:
Linear elastic behaviour until failure (no strain or stress or stress path
dependent stiffness behaviour)
Be careful with efffective strength in undrained behaviour
326 of 375
Plaxis Introductory Course, Danang, Vietnam
2p,fric
p fric
1p,fric
q
Hardening Soil model MC failure line
Compaction hardening:
Elastoplastic formulation of
p  v relationship in Cap
primary compression pc
fc = 0
pc p
c 1
v
327 of 375
Plaxis Introductory Course, Danang, Vietnam
Straindependent stiffness
Hysteresis
Energy dissipation
Damping
329 of 375
Plaxis Introductory Course, Danang, Vietnam
330 of 375
Plaxis Introductory Course, Danang, Vietnam
Limitations:
Less suitable for overconsolidated clay and in certain unloading stress
paths; not suitable for sand
No timedependent behaviour (secondary compression)
331 of 375
Plaxis Introductory Course, Danang, Vietnam
332 of 375
Plaxis Introductory Course, Danang, Vietnam
Limitations:
Side role of overconsolidation ratio OCR
Influence of K0ncparameter (M)
No softening
333 of 375
Plaxis Introductory Course, Danang, Vietnam
HoekBrown model 
Hoek Brown failure criterion (2002):
HoekBrown
a
'3
'1 '3 ci mb s 
ci 
GSI 100
mb mi exp
28 14 D ci = Intact uniaxial compressive
1 1 GSI 20 strength
a exp exp
2 6 15 3 GSI = Geological Strength Index
GSI 100 mi = Intact rock parameter
s exp D = Disturbance Factor
9 3D
HoekBrown model 
1
Uni axial compressive strength:
Uniaxial
c ci s a
 
Tensile strength:
s ci
t
mb
c
3

t
334 of 375
Plaxis Introductory Course, Danang, Vietnam
HoekBrown model 
Parameters:
HoekBrown model 
Possibilities:
Continuum approach of rock strength
 
Limitations:
Only applicable to rock
No anisotropy
335 of 375
Plaxis Introductory Course, Danang, Vietnam
336 of 375
Plaxis Introductory Course, Danang, Vietnam
337 of 375
Plaxis Introductory Course, Danang, Vietnam
338 of 375
Plaxis Introductory Course, Danang, Vietnam
340 of 375
Plaxis Introductory Course, Danang, Vietnam
Content
Theory of consolidation
FEM for consolidation analysis
Validation: Onedimensional consolidation
New features in PLAXIS 2D 2010
Conclusions
341 of 375
Plaxis Introductory Course, Danang, Vietnam
Theory of consolidation
C
Considering:
id i
Theory of consolidation
P
Pore water
t flow:
fl
k k
Darcys law: q pw v 2 pw t
w w
Total change of volumetric strain in time, k = permeability
considering homogeneous permeability: w = unit weight of water
v ( v1 v 2 ) n pw k 2
pw
t t K w t w
342 of 375
Plaxis Introductory Course, Danang, Vietnam
Theory of consolidation
G
General
l 3D case:
v 1 p ' 1 ( p pw ) 1 p 1 pw
t K ' t K ' t K ' t K ' t
1 p 1 n pw k 2
pw
K ' t K ' K w t w
E'
where K' = bulk stiffness of soil skeleton and p = mean total stress
3 1 2 '
kK' pw p
Considering incompressible water: 2 pw
w t t
Theory of consolidation
1D consolidation:
lid ti
v 1 ' 1 ( pw )
2H
t Eoed t Eoed t
1 1 n pw k 2
pw
Eoed t Eoed K w t w
(1 ') E '
where Eoed = constrained modulus of soil skeleton
(1 ')(1
)(1 2 '))
k Eoed pw
Considering incompressible water: 2 pw
w t t
343 of 375
Plaxis Introductory Course, Danang, Vietnam
Theory of consolidation
1D consolidation,
lid ti considering
id i a constant
t t ttotal
t l stress
t : 0
t
pw k Eoed
cv 2 pw where cv = consolidation coefficient =
w
t
cv t
T
H2
Assumptions:
Steady state pore pressure is constant in time (horizontal phreatic level or
steady state pore pressure from groundwater flow calculation)
Excess pore pressure can change in time
Fully saturated soil (above and below phreatic level)
Limitation:
Time dependent hydraulic boundary is not possible (variable phreatic level)
344 of 375
Plaxis Introductory Course, Danang, Vietnam
345 of 375
Plaxis Introductory Course, Danang, Vietnam
A C B D
C
axisymmetry
A
B E
D F E
346 of 375
Plaxis Introductory Course, Danang, Vietnam
347 of 375
Plaxis Introductory Course, Danang, Vietnam
348 of 375
Plaxis Introductory Course, Danang, Vietnam
no creep
creep
time (logscale)
349 of 375
Plaxis Introductory Course, Danang, Vietnam
Mesh:
M h Stress
Elements: Interpolation of primary variables point
Nodes: Primary y variables ((displacements,
p p
pore p
pressures))
Stress points: Derived variables (strains, stresses, Darcy velocities)
350 of 375
Plaxis Introductory Course, Danang, Vietnam
l2
tcritical
Cv
Note: smaller steps
p may yggive
stress oscillations
l = element length
= 80 ffor 15
15node
d ttriangles
i l
= 40 for 6node triangles
351 of 375
Plaxis Introductory Course, Danang, Vietnam
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
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
* *
352 of 375
Plaxis Introductory Course, Danang, Vietnam
353 of 375
Plaxis Introductory Course, Danang, Vietnam
354 of 375
Plaxis Introductory Course, Danang, Vietnam
Bishop stress
Suction (a new variable)
Retention curves (MualemVan Genuchten + user defined models)
Existing Plaxis soil models (Bishop stress)
User defined soil models (Bishop stress and suction)
independent of deformation)
Advanced mode
Bishop stress
Flow mode
Steady state groundwater flow
355 of 375
Plaxis Introductory Course, Danang, Vietnam
Conclusions
FEM is quite suitable for 2D and 3D consolidation analysis
2D or 3D coupled consolidation is different from 1D or uncoupled consolidation
PLAXIS has several options for consolidation based on excess pore pressure
Adding creep gives more realistic timedependent behavour and leads to
delayed consolidation
R
Recent t development:
d l t F
Fully
ll coupled
l d flflowdeformation
d f ti analysis
l i andd unsaturated
t t d
soil behaviour
356 of 375
Plaxis Introductory Course, Danang, Vietnam
GEOTEXTILE REINFORCED
EMBANKMENT WITH CONSOLIDATION
Computational Geotechnics 1
357 of 375
Plaxis Introductory Course, Danang, Vietnam
2 Computational Geotecnics
358 of 375
Plaxis Introductory Course, Danang, Vietnam
INTRODUCTION
In 1979 a test embankment was constructed in the Netherlands near the town of Almere. The
objective of this test was to measure the influence of geotextile reinforcement on the short
term stability of an embankment on soft soil. Two test embankments were constructed on top
of a layer, one with and one without geotextile. The construction procedure was such that a
ditch was excavated in the clay layer while at the same time a retaining bank was made with
the excavated clay. A crosssection of the reinforced test embankment is given in figure 1.
line of symmetry
geotextile
retaining bank
1
sand fill 2
2
soft clay
1,5
strong sand layer
1 3.5 3.5 1 3 14 7
Cone penetration tests gave an average cone resistance of qc = 150 kPa for the clay. The clay
is considered to be normally consolidated. The behaviour is assumed to be undrained (the
retaining bank should be drained, however). The saturated weight of the clay is 13.5 kN/m3.
A plasticity index of Ip = 50% is assumed. Due to the limited soil data, parameters should
be selected using engineering judgement and by using the correlations given in the lecture
"Evaluation of soil stiffness parameters". To obtain an undrained shear strength for the clay
layer it is suggested to use the correlation su qc /15. Having no data for the effective cohesion
and the effective friction angle, they have to be estimated from the undrained shear strength in
order to do a consolidation analysis. For the determination of a stiffness parameter for the clay
layer it is suggested to use the correlation Eu 15000 su /Ip (%). The shear modulus G is one
third of the undrained Youngs modulus Eu . The effective Poissons ratio should be chosen
such that a realistic K0nc is obtained in onedimensional compression (K0nc = 0 /(1 0 ) 0.5).
The effective Youngs modulus is calculated from the shear modulus E 0 = 2G(1 + 0 ). The
fill was reported to be fully saturated loose sand with a saturated weight of 18 kN/m3 . The
behaviour is considered to be drained. The effective strength properties are estimated at 0 =
30 and c = 3 kPa. K0nc is assumed at 0.5. For the stiffness one should take E = 4000 kPa
and 0 =0.33.
Computational Geotechnics 3
359 of 375
Plaxis Introductory Course, Danang, Vietnam
AIMS
Calculation of two alternatives within one project.
SCHEME OF OPERATIONS
1. Determination of stiffness & strength properties (clay)
2. Geometry input
3. Calculation
(a) Initial conditions (Pore pressure generation, Initial geometry configuration, Genera
tion of initial stresses)
(b) Switch on geotextile, excavate ditch + raise retaining embankment
(c) Apply first hydraulic fill
(d) Apply second hydraulic fill
(e) Determine factor of safety
(f) Repeat this using consolidation phases instead of plastic phases.
4. Inspect output
4 Computational Geotecnics
360 of 375
Plaxis Introductory Course, Danang, Vietnam
Note: The main purpose of the exercise is to assess the failure mechanism and
the factor of safety, which has the following consequences for the model:
The geometry size is chosen such that the failure mechanism fits
within the model boundaries. This means the geometry can be fairly
small.
Computational Geotechnics 5
361 of 375
Plaxis Introductory Course, Danang, Vietnam
GEOMETRY INPUT
General settings
Start a new project and select appropriate General settings. Use 15node elements as basic
element type since in this exercise we will deal with failure behaviour.
(9.5,7.5) (12,8.5)
10 11
(33,8.5)
(8,7.5)
7 9 8 (33,7.5)
(4.5,5.5) (12,5.5) (26,5.5)
(0,5.5) 1 6 12 13 2
(33,5.5)
(0,3.5) 4 5 (1,3.5)
(0,2) 14 15
(33,2)
y
(0,0) 0 x 3
(33,0)
Enter the geometry as indicated in the previous graph. The order in which geometry
points are created is arbitrary.
Click the Geogrid button to introduce the geotextile (from (4.5, 5.5) to (26.0, 5.5)).
Click the Standard fixities button for the standard boundary conditions.
After entering all properties for the three soil types, drag and drop the properties to the
appropriate clusters, as indicated in figure 3.
6 Computational Geotecnics
362 of 375
Plaxis Introductory Course, Danang, Vietnam
3
2 3
1
1
Geotextile
In the project database select the data type Geogrids and create a new material set. In
this material set, enter 2500 kN/m as stiffness. Note that this is the stiffness in extension.
In compression no stiffness is used.
Drag the geogrid data set to the geotextile in the geometry and drop it there. The
geotextile should flash red once, indicating the properties have been set.
Mesh generation
From the Mesh menu select the option Global coarseness. In the window that appears,
set the mesh coarseness to Medium and click on the Generate button, which will present
the following FE mesh composed of 15node elements.
Computational Geotechnics 7
363 of 375
Plaxis Introductory Course, Danang, Vietnam
Select the clay layer (this consists of two clusters, see also hint) and press Refine cluster
from the Mesh menu. This will result in a refinement in the clay layer that will be needed
for the consolidation analysis. See figure 5.
Close the window showing the generated mesh and continue to the Calculations program.
8 Computational Geotecnics
364 of 375
Plaxis Introductory Course, Danang, Vietnam
CALCULATION
The calculation consists of two alternatives for the construction of the embankment: without
and with consolidation taken into account. After both alternatives the factor of safety is
determined. In the calculations list 8 phases are needed, 4 phases for each alternative. First
start with the fully undrained construction, that is without taking consolidation into account.
When starting Plaxis Calculations, choose Classical mode.
Initial conditions
Select the initial phase in the phase list and then press the Define button on the Para
meters tabsheet in order to define the initial phase. The input window now opens in
Staged Construction mode.
Deselect all material clusters and geotextile elements that are not present at the start of
the analysis. As we want to model the entire construction sequence from the beginning,
switch off:
Geotextile elements
Material clusters for the fill
Material cluster for retaining bank
Now continue to the Water conditions mode by clicking the equally named button.
Enter a phreatic level at ground level by two coordinates (0, 5.5) and (33, 5.5). Click on
the Water pressures button to generate the pore pressures.
Computational Geotechnics 9
365 of 375
Plaxis Introductory Course, Danang, Vietnam
After this, we will construct the embankment taking into account consolidation:
10 Computational Geotecnics
366 of 375
Plaxis Introductory Course, Danang, Vietnam
Computational Geotechnics 11
367 of 375
Plaxis Introductory Course, Danang, Vietnam
INSPECT OUTPUT
In order to get a good idea of the displacement mechanism, one can view the contours
of incremental displacements. Figure 6 shows this plot of the final calculation step for the
undrained construction. It clearly shows the effect of the geotextile reinforcement. Figure 7
shows the incremental displacement for the consolidated construction. Here the embankment
has a more gradual settlement without showing an upcoming failure mechanism.
The axial forces of the geotextile can be visualised by double clicking on the geotextile. This
will first present the displacement of the geotextile. On using the menu item Forces, one can
select Axial forces N.
At the ends of the geotextile the axial force must be zero, but due to the discretisation and
some numerical inaccuracy this is not completely achieved. The maximum axial forces is
approx. 8 kN/m. figure 9 shows the axial forces for the consolidated construction. The
maximimum axial force here is only 56 kN/m.
Finally, the factors of safety are checked. In order to do so follow these steps:
Start the curves manager by selecting the Curves manager option from the Tools menu.
12 Computational Geotecnics
368 of 375
Plaxis Introductory Course, Danang, Vietnam
In the curves manager (see figure 10) select New in the Charts tabsheet. This presents
the Curve Generation window as shown in figure 11.
On the xaxis we want to show the displacements of the point at the toe of the embank
ment, hence choose Point A and Deformations Total displacements u.
On the yaxis we want to show the strength reduction factor, hence select Project and
Multiplier M sf on the yaxis.
The created curve indicates a safety factor around 1.4 for the undrained construction and a a
safety factor of 2.1 for the consolidated construction of the embankment, as can be seen in
figure 12.
From the graph above, the factor of safety can be determined. Always look for a steady state
solution (slight variations in the load multipliers, increasing displacements). In most case, the
phi/c reduction calculation shows some variation at the beginning of the calculation. Note
that the displacements resulting from a Safety analysis are nonphysical. Hence the total
displacements are not relevant. An incremental displacement plot of the last step, however,
shows the failure mechanism that corresponds the calculated value for M sf .
Addicionally, figures 13 and 14 show the failure mechanisms with the lowest factor or safety
for both the undrained and consolidated case.
Computational Geotechnics 13
369 of 375
Plaxis Introductory Course, Danang, Vietnam
Consolidated: Msf=2.1
Undrained: Msf=1.4
14 Computational Geotecnics
370 of 375
Plaxis Introductory Course, Danang, Vietnam
Computational Geotechnics 15
371 of 375
Plaxis Introductory Course, Danang, Vietnam
FACTORS OF SAFETY
The factors of safety are checked with the Curves program, see figure 19.
16 Computational Geotecnics
372 of 375
Plaxis Introductory Course, Danang, Vietnam
Consolidated: Msf=1.4
Undrained: Msf=1.1
Computational Geotechnics 17
373 of 375
Plaxis Introductory Course, Danang, Vietnam
18 Computational Geotecnics
374 of 375
Plaxis Introductory Course, Danang, Vietnam
Computational Geotechnics 19
375 of 375