FLAC Slope Stability Example

Example Applications
PRECIS
This volume contains documentation on a series of example application problems that have been
solved using FLAC. These example applications demonstrate the various classes of problems to
which FLAC may be applied.*
Table 1 presents a summary of the example applications described in this volume. The table also
identifies the specific FLAC feature that is examined in each problem.
The problems in this volume represent a brief sampling of potential applications for FLAC. We
plan to update this volume on a regular basis and will send new examples as they are prepared. We
also invite users to submit their own examples for inclusion or inform us of any type of problem
that they would like to see in this volume.
* The data files in this volume are all created in a text editor. The files are stored in the directory
ITASCA\FLAC500\ExampleApplications with the extension .DAT. A project file is also provided for each example. In order to run an example and compare the results to plots in this volume,
open a project file in the GIIC by clicking on the File / Open Project menu item and selecting
the project file name (with extension .PRJ). Click on the Project Options icon at the top of the
Project Tree Record, select Rebuild unsaved states and the example data file will be run and plots
created. All problems in this volume were run on a 2.4 GHz computer.
Example applications for the optional features thermal analysis, creep material models and
dynamic analysis are provided in Section 1, Section 2 and Section 3 in Optional Features,
respectively. There are also additional example applications for interface elements, structural elements and groundwater analysis in Section 4 in Theory and Background, Section 1 in Structural
Elements, and Section 1 in Fluid-Mechanical Interaction, respectively.
FLAC Version 5.0
Summary of Example Applications
1 Slope Stability for a Cohesive
1-1
SLOPE.DAT
2 Axisymmetric Modeling of
Post-Pillar Mining
2-1
POSTP.DAT
3 Cemented Backfill Pillar
3-1
SAND.DAT
4-1
CAV.DAT
5-1
PILL.DAT
rockbolts
piles
liners
beams
Interface Elements
plane stress
groundwater
extra
dynamic
cppudm
axisymmetry
ats
ubiquitous joint
cables
STRUCTURAL
ELEMENTS
CONFIG
strain-softening
Mohr-Coulomb
Finn
elastic, isotropic
CONSTITUTIVE MODEL
elastic, anisotropic
INPUT FILE
Drucker-Prager
PAGE
Cam-clay
DESCRIPTION
double-yield
Table 1
and Frictional Soil

x
Performance
4 Undrained Cylindrical Cavity
Expansion in a Cam-Clay
Medium
5 Post-Peak Pillar Behavior
and the Effects of Backfill

Confinement
6 Full-Scale Test Wall in Sand
6-1
WALL.DAT
7 Stresses around a Pressurized
7-1
PRESSTUNNEL.DAT
x
x
x
x
Concrete Tunnel
8 Displacements near the Face
8-1
SHAX.DAT
9-1
PULL.DAT
10-1
ROCKSL.DAT
11-1
DIAP.DAT
12-1
MSTUNNEL.DAT
of an Advancing Tunnel
9 Simulation of Pull-Tests for
Grouted Cable Anchors

10 Slope Stability Analysis of a
Rock Slope
11 Analysis of a Concrete
Diaphragm Wall
12 Multi-Stage Tunnel Excavation
and Support
13 Embankment Loading on a
Cam-Clay Foundation
13-1
EMC.DAT
14 Dewatered Construction of a
Braced Excavation
14-1
EXCAVATE.DAT
15 Earthquake Loading of a
Pile-Supported Wharf
15-1
WHARF.DAT
16 Pile-Supported Highway
Embankment
16-1
PEMBANK.DAT
17 Lined Tunnel Construction in

Saturated Ground
17-1
LINER.DAT
18 Seismic Analysis of an
18-1
EARTHDAM.DAT
Embankment Dam
FLAC Version 5.0
x
x
x
x
x
x
x
x
x
Slope Stability for a Cohesive and Frictional Soil
1-1
1 Slope Stability for a Cohesive and Frictional Soil

1.1 Problem Statement
A common problem encountered in engineering soil mechanics is the stability of soil slopes in
frictional materials. In this example, three slope conditions are analyzed. First, a slope in sand
with zero cohesion is modeled with an initial slope steeper than the angle of repose of the sand.
This slope, of course, should collapse; the progression of this collapse is calculated as it develops.
Second, a small cohesion is added to the material and the slope is re-examined to determine whether
it is stable. Third, the water level in the slope is raised, and the effect on stability is examined.
In this example, the soil is homogeneous, and the stability and factor of safety of the slope can
readily be determined using an analytic or graphical technique.* However, the power of the FLAC
code lies in its ability to model more complex slope geometries in which, for example, several
layers of soil with differing material properties and/or constitutive behaviors may exist. This type
of problem can be examined with no greater effort than the homogeneous case by simply assigning
different material properties and/or models to different zones.
This example also demonstrates two approaches to analyze the effect of a phreatic surface in
the slope. In one approach, an effective-stress analysis is performed after adding a pore-pressure
distribution directly to the zones in the model. In the other approach, a groundwater flow calculation
is performed first to establish the phreatic surface; then the effective-stress analysis is performed.
1.2 Modeling Procedure
1.2.1 Initial Model State
The following sequence of operations is used to initialize the slope and create a restart file from
which the boundary conditions and/or material properties may be varied for the three cases of this
example. Please refer to SLOPE.DAT in Section 1.3 for a complete listing of the commands for
this example.
An initial grid of 20 10 square zones is set up. Note that this initial grid is assigned, by default, the
dimensions of 20 units in the x-direction by 10 units in the y-direction unless otherwise redefined
by the user. The command to do this is:
grid 20 10
* See Section 10 for a comparison of FLAC results to a limit equilibrium analysis.
FLAC Version 5.0
1-2
A Mohr-Coulomb constitutive model is assigned to all zones (assumed because no range is given)
with the following properties:
density
shear modulus
bulk modulus
friction angle
cohesion
tensile strength
1500 kg/m3
0.3 108 Pa
108 Pa
20
1010 Pa
1010 Pa
Note that a high cohesion and tensile strength are assigned to prevent slope failure during the
initialization of gravitational stresses in the model (see below).
The basic grid is deformed into the shape of a slope and the soil base beneath it. This is done using
the GENERATE command. Two different quadrilateral regions are created with the following two
GENERATE commands:
gen 0,0 0,3 20,3 20,0
gen same 9,10 20,10 same
j = 1,4
i = 6,21
j = 4,11
The first GENERATE command defines the base of the slope, and the second GENERATE command
creates the slope. Note that the zones are aligned with the angle of the slope so that the zones
along the slope face are all quadrilateral-shaped. This is recommended because all zones are then
composed of two overlaid sets of triangular elements. These zones are well-suited for plasticity
analysis (see Section 1.3.3.2 in Theory and Background). It is also possible to create a slope
using the GENERATE line command. However, with this command, single triangular zones will be
created along the slope face; these zones are not as accurate for plasticity analysis.
The area directly to the left of the slope face is excavated by declaring the appropriate zones as
null. This is done by creating a region (i.e., the grid is divided into two regions separated by a
boundary) that is defined by marking selected gridpoints as boundaries between regions. The
following commands mark the boundary of the excavated region and then null the zones within that
region:
mark i = 1,6 j = 4
mark i = 6
j = 4,11
model null region 1,10
The marked boundaries can be verified by issuing the PRINT mark command. The MODEL null
command will delete zones in the region containing zone (1,10). Figure 1.1 shows the resulting
FLAC grid.
FLAC Version 5.0
1-3
JOB TITLE : SLOPE UNDER GRAVITATIONAL LOAD
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
24-Aug-04 16:47
step
715
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Grid plot
0
5E 0
0.400
0.000
-0.400
Itasca Consulting Group, Inc.

Minneapolis, Minnesota USA
0.200
Figure 1.1
0.600
1.000
(*10^1)
1.400
1.800
Grid plot of initial slope
The acceleration of gravity is set to 9.81 m/sec2 (positive means acting downward). The boundary
conditions consist of roller boundaries on the left and right sides of the model as well as a fixed base.
An initial elastic state in which gravitational stresses are equilibrated is desired. This is achieved
with the SOLVE command, using default limits. Equilibrium is obtained when the out-of-balance
force ratio limit of 103 is reached. To examine the progression of the solution, the y-displacement
history is requested at a gridpoint at the slope crest. This is done using the HISTORY command.
When the SOLVE command has reached its limit, the history may be plotted to verify that the mesh
is, indeed, at an equilibrium state (i.e., PLOT history 1). A numerical and graphical representation
of the elastic stresses can be obtained by issuing the following commands:
print sxx syy
plot sxx yellow
plot syy yellow
bound
bound
Initially, very large values for the cohesion and tensile strength are assigned to the slope. In justification, let us re-examine the way in which an explicit model works. An initial grid is created first
and, in this case, gravity applied to the gridpoints and gravitational stresses allowed to equilibrate.
For most problems, it is desirable that this process occur as rapidly as possible. This can be done
by requiring the material to behave elastically during the equilibration process.
Once stresses have equilibrated, actual material properties are assigned, excavation is made, loads
are applied, etc., and the simulation process continued. In the case illustrated here, a plastic
constitutive model is assigned initially, with high cohesion and tensile strength, forcing the material
to behave elastically. Then the cohesion and tensile strength are reset to the desired values. This
FLAC Version 5.0
1-4
eliminates the necessity of redefining the constitutive model and properties after equilibrium, thus
saving a few steps of input. The same effect could be obtained by using an elastic model initially
for the equilibration process, followed by a change to the Mohr-Coulomb model, followed by a
definition of the material friction, cohesion and tensile strength. Either methodology for initializing
a gravity-loaded grid is acceptable and is selected based on the preference of the user.
A restart file is created to save the elastic equilibrium state. This is done to save time in case future
runs will be made in which material parameters or constitutive models are varied. Performing
these studies requires only that the elastic state be restored, therefore eliminating the necessity to
recompute the equilibrium state.
The command
save sl1.sav
will create a restart file on the default drive called SL1.SAV. FLAC could be halted at this point
and the program run with the saved state restored at a later time, simply by typing
restore sl1.sav
1.2.2 Slope Collapse: Dry Conditions

For the next stage of the simulation, the material properties are set to the actual soil values and the
calculation continued while examining the possible failure process. During this process, plots of the
progressive displacement of the slope are made. To avoid any confusion in analyzing the data, only
the change in displacement is monitored not the cumulative displacements from the beginning
of the simulation. The calculation procedure in FLAC does not involve displacements, but keeps
the cumulative total for each gridpoint as a convenience to the user. Therefore, the displacements
may be initialized at any point in the run without affecting the results. This is done by using the
command
ini
xdis=0
ydis=0
From this point on, plots or printouts will show only the change in displacement from the previous
state.
Next, the material properties of the zones are reset by using the PROPERTY command. The cohesion
is set to zero for all zones that are currently composed of soil.* Finally, the calculation mode is
set to large-strain to provide a more accurate geometrical representation of the slope failure as it
progresses. Because slope collapse will occur due to the angle of repose of the soil being smaller
than the slope angle, the SOLVE command is not used (because equilibrium will not be reached).
The STEP command is used to step the simulation a small number of calculational steps at a time,
stopping to print and plot intermediate stages. Here, the power of the explicit method is evident in
* Note that the tensile strength will also be set to zero in FLAC, because the tensile strength is
calculated from cohesion/tan (friction).
FLAC Version 5.0
1-5
its ability to follow highly nonlinear problems, which may never converge to an equilibrium state,
through progressive failure.
The model is now stepped in intervals to evaluate the progressive collapse of the slope. The collapse
is revealed when printing and plotting results after each step. The following commands are used:
print xv yv xd yd state
plot xv z disp bou
Figures 1.2 and 1.3 show the state of the slope at 915 steps and 1715 steps. These figures illustrate
the progressive collapse and, in particular, indicate the location of the failure (slip) surface. The
slope is collapsing in an attempt to reach its angle of repose. At some point, the displacements of the
gridpoints become unrealistic because of extreme distortion of the grid. FLAC automatically checks
for excessive grid deformation and will stop the calculation process if the condition is detected,
displaying an error message.
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
24-Aug-04 16:47
step
915
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Boundary plot
0
5E 0
Displacement vectors
max vector = 2.674E-02
0.400
5E -2
X-velocity contours
Contour interval= 2.50E-05
Minimum: -2.25E-04
Maximum: 0.00E+00
(zero contour omitted)
0.000
-0.400

0.200
Figure 1.2
0.600
1.000
(*10^1)
1.400
1.800
Plot of displacement vectors and x-velocity contours at step 915
FLAC Version 5.0
1-6
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
24-Aug-04 16:47
step
1715
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Boundary plot
0
5E 0
0.400
1E 0
X-velocity contours
Minimum: -1.00E-03
Maximum: 0.00E+00
0.000
-0.400

0.200
Figure 1.3
0.600
1.000
(*10^1)
1.400
1.800
Plot of displacement vectors and x-velocity contours at step 1715
An interesting comparison can be obtained by running another simulation in which a small cohesion
and zero tensile strength are assigned to the soil. Because the initial elastic equilibrium state has
been saved, the problem can be restored from this state:
rest sl1.sav
Following the procedure used earlier, enter the following commands:

ini xdis=0 ydis=0
prop coh=1e4 tens=0.0
set large
solve
Under these conditions, the results will show that the slope is stable. (Note the small magnitude of
the calculated displacements.)
FLAC Version 5.0
1-7
1.2.3 Effective Stress Analysis with WATER table

Next, we wish to assess the stability of the slope with a water table present. Continuing with the
model at the present state, the water level is raised in the slope to a height of 9 m on the right side
of the model and 5 m on the left side (i.e., 2 m above the base of the excavation). It is a simple
matter to use the WATER table command to specify the phreatic surface. Note that the correct
wet and dry densities must also be supplied. The FISH function wet den in SLOPE.DAT at
BRANCH:WATER TABLE prescribes the wet density to zones below the water table, assuming
a soil porosity of 0.3. The APPLY press command must also be used to apply the weight of the
water in the excavation to the surface of the excavation. Figure 1.4 illustrates the location of the
water table, the applied forces representing the weight of the water in the excavation, and the wet
and dry densities in the zones.
The calculation is continued with the STEP command. Now, after an additional 6000 steps, the
slope is observed to fail. The pore pressure distribution and velocity vectors indicating the slope
failure in the model are plotted in Figure 1.5.
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
24-Aug-04 16:47
step
1473
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Density
1.45E+03
1.50E+03
1.55E+03
1.60E+03
1.65E+03
1.70E+03
1.75E+03
1.80E+03
0.400
Contour interval= 5.00E+01

Grid plot
0
0.000
5E 0
Water Table
Net Applied Forces
max vector =
2.000E+04
-0.400
0 Consulting Group,
5E 4Inc.
Itasca
0.200
Figure 1.4
0.600
1.000
(*10^1)
1.400
1.800
Location of water table, applied forces along slope and wet and
dry densities
FLAC Version 5.0
1-8
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
24-Aug-04 16:47
step
7473
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
Boundary plot
5E 0
Velocity vectors
0
0.800
E
F
5E -5
0.400
Pore pressure contours

B: 1.000E+04
I: 8.000E+04
H
I
0.000
-0.400

0.200
Figure 1.5
0.600
1.000
(*10^1)
1.400
1.800
Pore pressure distribution and velocity vectors in failed slope (using WATER table)
1.2.4 Effective Stress Analysis with CONFIG gw

Alternatively, the groundwater flow option in FLAC can be used to find the phreatic surface and
establish the pore pressure distribution before the mechanical response is investigated. The model
is run in groundwater flow mode by using the CONFIG gw command. This command must be given
before the GRID command, so we create a new model state in SLOPE.DAT, at BRANCH:SLOPE
WITH GW in Section 1.3 for this case. We turn off the mechanical calculation (SET mech off) in
order to establish the initial pore pressure distribution. We apply pore pressure boundary conditions
to raise the water level to 5 m at the left boundary, and 9 m at the right. The slope is initially dry (INI
sat 0). We also set the bulk modulus of the water to a low value (1.0 104 ) because our objective
is to reach the steady-flow state as quickly as possible. The groundwater time scale is wrong in this
case, but we are not interested in the transient time response. The steady-flow state is determined by
using the SOLVE ratio command. When the groundwater flow ratio falls below the set value of 0.01,
steady-state flow is achieved. This can be checked by using the FISH function QRATIO.FIS to
assess the flow state. (See Section 3 in the FISH volume for a description of this function.) The
steady-flow state is indicated by the plot of flow vectors and phreatic-surface contour in Figure 1.6.
(The FISH function PS.FIS is used to locate the phreatic surface see Section 3 in the FISH
volume.) The bumpy phreatic-surface line is due to the coarse discretization.
FLAC Version 5.0
1-9
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
24-Aug-04 16:47
step
2054
Flow Time
5.9240E+07
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Boundary plot
0
Flow vectors
max vector =
0
5E 0
5.672E-07
0.400
1E -6
phreatic surface
Minimum: 0.00E+00
Maximum: 5.00E-01
0.000
-0.400

0.200
Figure 1.6
0.600
1.000
(*10^1)
1.400
1.800
Steady-state flow through slope
Mechanical equilibrium is then established including the pore pressure. This is accomplished with
the following commands:
set flow off mech on
apply press 2e4 var 0
water bulk = 0.0
-2e4
from 1,4
to 6,6
These commands turn off the flow calculation, turn on the mechanical calculation, apply the weight
of the water to the excavation surface, and set the bulk modulus of the water to zero. The last
command prevents pore pressures from generating as a result of mechanical deformation. This is
done so that the results can be compared to the previous case using the water table. The SOLVE
command is then used to find the equilibrium state.
Finally, the cohesion is reduced to 1.0 104 , the tensile strength to zero, and the calculation is
continued with the STEP command. The slope fails, as shown in Figure 1.7. The result is the same
as that which occurs using WATER table.
The approach using the groundwater flow mode can take longer to reach a solution because of the
extra calculation needed to establish the pore pressure distribution. However, this method is helpful
when the pore pressure distribution or phreatic surface location is unknown. Also, this approach
avoids the necessity of assigning wet density values to zones beneath the phreatic surface, which
can become difficult for complex geometries.
FLAC Version 5.0
1 - 10
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
24-Aug-04 16:47
step
3865
Flow Time
5.9240E+07
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
B
Boundary plot
5E 0
Velocity vectors
0.800
0.400
5E -4

B: 1.000E+04
I: 8.000E+04
I
0.000
-0.400

0.200
Figure 1.7
FLAC Version 5.0
0.600
1.000
(*10^1)
1.400
1.800
Pore pressure distribution and velocity vectors in failed slope (using CONFIG gw)
1 - 11
1.3 Data File SLOPE.DAT

;Project Record Tree export
;*** BRANCH: SLOPE WITHOUT GW ****
new
;... STATE: SL1 ....
config ats
grid 20,10
;Mohr-Coulomb model
m m
; soil properties --- note large cohesion to force initial elastic
; behavior for determining initial stress state. This will prevent
; slope failure when initializing the gravity stresses
prop s=.3e8 b=1e8 d=1500 fri=20 coh=1e10 ten=1e10
; warp grid to form a slope :
gen 0,0 0,3 20,3 20,0 j 1,4
gen same 9,10 20,10 same i 6 21 j 4 11
mark i=1,6 j=4
mark i=6 j=4,11
; displacement boundary conditions
fix x i=1
fix x i=21
fix x y j=1
; apply gravity
set grav=9.81
; displacement history of slope
his ydis i=10 j=10
; solve for initial gravity stresses
solve
; save initial state
save sl1.sav
;*** BRANCH: DRY ****
;... STATE: SL2 ....
; reset displacement components to zero
ini xdis=0 ydis=0
; set cohesion to 0
prop coh=0
; use large strain logic
set large
step 200
save sl2.sav
FLAC Version 5.0
1 - 12
;... STATE: SL3 ....

step 800
save sl3.sav
;*** BRANCH: WATER TABLE ****
restore sl1.sav
;... STATE: SL4 ....
ini xdis=0 ydis=0
prop coh=1e4 tens 0.0
set large
solve
;
ini xdis=0.0 ydis=0.0
; install phreatic surface in slope
water table 1 den 1000
table 1 (0,5) (6.11,5) (20,9)
def wet den
loop i (1,izones)
loop j (1,jzones)
if model(i,j)>1 then
xa=(x(i,j)+x(i+1,j)+x(i+1,j+1)+x(i,j+1))
xc=0.25*xa
ya=(y(i,j)+y(i+1,j)+y(i+1,j+1)+y(i,j+1))
yc=0.25*ya
if yc<table(1,xc) then
density(i,j) = 1800
end if
end if
end loop
end loop
end
wet den
apply press 2e4 var 0 -2e4 from 1,4 to 6,6
save sl4.sav
;... STATE: SL5 ....
step 6000
sclin 1 19 0 19 10
save sl5.sav
;*** BRANCH: SLOPE WITH GW ****
new
;... STATE: SLGW1 ....
FLAC Version 5.0
1 - 13
config gw ats ex 1
grid 20,10
;Mohr-Coulomb model
m m
; soil properties --- note large cohesion to force initial elastic
; behavior for determining initial stress state. This will prevent
; slope failure when initializing the gravity stresses
prop s=.3e8 b=1e8 d=1500 fri=20 coh=1e10 ten=1e10
; warp grid to form a slope :
gen 0,0 0,3 20,3 20,0 j 1,4
gen same 9,10 20,10 same i 6 21 j 4 11
mark i=1,6 j=4
mark i=6 j=4,11
prop perm 1e-10 por .3
water den 1000 bulk 1e4
; displacement boundary conditions
fix x i=1
fix x i=21
fix x y j=1
; pore pressure boundary conditions
apply pp 9e4 var 0 -9e4 i 21 j 1 10
apply pp 5e4 var 0 -3e4 i 1 j 1 4
ini
pp 2e4 var 0 -2e4 mark i 1 6 j 4 6
fix pp mark
; apply gravity
set grav=9.81
call qratio.fis
hist gwtime
hist qratio
hist inflow
hist outflow
set mech off
solve
save slgw1.sav
set flow off mech on
app press 2e4 var 0 -2e4 from 1 4 to 6 6
water bulk 0.0
; displacement history of slope
hist reset
his ydis i=10 j=10
solve
save slgw3.sav
FLAC Version 5.0
1 - 14

ini xdis 0.0 ydis 0.0
prop coh 1e4 ten 0.0
set large
step 1000
sclin 1 19 0 19 10
save slgw4.sav
;*** plot commands ****
;plot name: grid
plot hold grid
;plot name: Displacement vectors
plot hold bound displacement xvel zero
;plot name: Water Table
plot hold density fill inv grid water apply lmagenta
;plot name: Pore pressure distribution
plot hold bound velocity pp
;plot name: Steady-state flow
plot hold bound flow saturation alias phreatic surface min 0.0 max 0.5 &
int 0.5 lmagenta
FLAC Version 5.0
Axisymmetric Modeling of Post-Pillar Mining
2-1
2 Axisymmetric Modeling of Post-Pillar Mining

Post pillars are created when mechanized cut-and-fill stopes are mined on all sides, creating a long
thin pillar that is confined by backfill. This example is an examination of the potential instability
of a post pillar.
The axisymmetric geometry in FLAC is used to approximate the post-pillar mining state. The
axisymmetry provides an analysis of pillar deformation closely related to the three-dimensional
condition.*
The FLAC model is shown in Figure 2.1. In this figure, the first cut of the cut-and-fill stope has
been removed. The post-pillar is modeled as a cylinder with a radius of 10 m. The width of the
stope is 20 m, and the height of the first cut is 10 m.
Figure 2.2 illustrates the history of horizontal displacement at the midpoint of the wall of the pillar
following the first cut. This figure indicates that the wall is stable.
Each successive cut of the cut-and-fill stope is excavated while filling the previous cut with sandfill.
The sandfill was modeled using the Mohr-Coulomb model and the double-yield model in two
different runs beginning from the first excavation stage. The shear strength properties are the same
for both models (friction angle = 35 ; cohesion = 0). For the double-yield model, the cap pressure
varies from 0.01 MPa at zero plastic volumetric strain, to 50 MPa at a plastic volumetric strain of
0.2.
2.3 Results
Figures 2.3 through 2.7 show the stress trajectories for successive mining stages. As seen, the
sandfill does not develop appreciable load, and the pillar becomes de-stressed in the horizontal
direction. These plots are from the run with the Mohr-Coulomb model; the results are similar for
the run with the double-yield model.
The difference between the two runs can be appreciated by comparing the vertical heave in the
fill. For the Mohr-Coulomb model, a maximum displacement of 0.19 m is calculated after the fifth
cut-and-fill (Figure 2.8). For the double-yield model, the sandfill heaves approximately 0.16 m at
the final stage (Figure 2.9). This difference is due to the ability of the fill to compact using the
double-yield model.
* FLAC 3D is recommended for a more rigorous three-dimensional analysis.
FLAC Version 5.0
2-2
If the model is stepped further beyond the state where the equilibrium ratio is below 0.5 105
after the final cut, the floor continues to heave a small amount. In fact, several hundred additional
steps are required to reach an equilibrium state. This is because very low forces are able to cause
displacement of the weak sandfill.
The plots of the plasticity state in Figures 2.10 through 2.14 show that the pillar yields adjacent to
the cut-and-fill stope but eventually stabilizes due to the surrounding fill. Figure 2.15 presents a
plot of horizontal displacement histories at different locations along the pillar, also indicating that
the pillar is stable. For these plots, the Mohr-Coulomb model was used for the sandfill.
JOB TITLE : MODEL OF POST-PILLAR MINING
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:49
step
2111
-3.501E+01 <x< 1.250E+02
-2.002E+01 <y< 1.400E+02
0.900
Grid plot
0
2E 1
0.700
0.500
0.300
0.100
-0.100

-0.200
Figure 2.1
FLAC Version 5.0
0.000
0.200
0.400
(*10^2)
0.600
0.800
1.000
1.200
FLAC axisymmetry model showing excavation of first cut
2-3
FLAC (Version 5.00)

(10
-02
LEGEND
6.500
25-Aug-04 11:49
step
2111
6.000
HISTORY PLOT
Y-axis :
X displacement( 5, 8)
X-axis :
Number of steps
5.500
5.000
4.500
4.000
3.500
3.000
12
16
20
(10
02

Figure 2.2
History of horizontal displacement at midpoint of pillar wall following excavation of first cut
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:49
step
2111
-3.501E+01 <x< 1.250E+02
-2.002E+01 <y< 1.400E+02
0.900
Boundary plot
0
2E 1
0.700
Principal stresses
Max. Value = -2.300E+05
Min. Value = -8.468E+07
0
0.500
5E 8
0.300
0.100
-0.100

-0.200
Figure 2.3
0.000
0.200
0.400
(*10^2)
0.600
0.800
1.000
1.200
Principal stresses after first cut
FLAC Version 5.0
2-4
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:49
step
3658
-3.501E+01 <x< 1.250E+02
-2.002E+01 <y< 1.400E+02
0.900
Boundary plot
0
2E 1
0.700
Principal stresses
Max. Value = 1.428E+01
Min. Value = -8.681E+07
0
0.500
5E 8
0.300
0.100
-0.100

-0.200
Figure 2.4
0.000
0.200
0.400
(*10^2)
0.600
0.800
1.000
1.200
Principal stresses after second cut and fill of first cut (M-C sandfill)
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:49
step
5293
-3.502E+01 <x< 1.250E+02
-2.003E+01 <y< 1.400E+02
0.900
Boundary plot
0
2E 1
0.700
Principal stresses
Min. Value = -8.899E+07
0
0.500
5E 8
0.300
0.100
-0.100

-0.200
Figure 2.5
FLAC Version 5.0
0.000
0.200
0.400
(*10^2)
0.600
0.800
1.000
1.200
Principal stresses after third cut and fill of second cut (M-C sandfill)
2-5
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:49
step
7070
-3.502E+01 <x< 1.250E+02
-2.003E+01 <y< 1.400E+02
0.900
Boundary plot
0
2E 1
0.700
Principal stresses
Min. Value = -9.035E+07
0
0.500
5E 8
0.300
0.100
-0.100

-0.200
Figure 2.6
0.000
0.200
0.400
(*10^2)
0.600
0.800
1.000
1.200
Principal stresses after fourth cut and fill of third cut (M-C sandfill)
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:49
step
9472
-3.502E+01 <x< 1.250E+02
-2.003E+01 <y< 1.400E+02
0.900
Boundary plot
0
2E 1
0.700
Principal stresses
Min. Value = -9.041E+07
0
0.500
5E 8
0.300
0.100
-0.100

-0.200
Figure 2.7
0.000
0.200
0.400
(*10^2)
0.600
0.800
1.000
1.200
Principal stresses after fifth cut and fill of fourth cut (M-C sandfill)
FLAC Version 5.0
2-6
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:49
step
9472
-3.502E+01 <x< 1.250E+02
-2.003E+01 <y< 1.400E+02
0.900
Boundary plot
0
2E 1
0.700
scaled to max = 5.000E-01
0
0.500
1E 0
0.300
0.100
-0.100

-0.200
Figure 2.8
0.000
0.200
0.400
(*10^2)
0.600
0.800
1.000
1.200
Displacement of fill after fifth cut, using the Mohr-Coulomb

model for the sandfill
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:50
step
8819
-3.502E+01 <x< 1.250E+02
-2.003E+01 <y< 1.400E+02
0.900
Boundary plot
0
2E 1
0.700
0
0.500
1E 0
0.300
0.100
-0.100

-0.200
Figure 2.9
FLAC Version 5.0
0.000
0.200
0.400
(*10^2)
0.600
0.800
1.000
1.200
Displacement of fill after fifth cut, using the double-yield model

for the sandfill
2-7
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:49
step
2111
-3.501E+01 <x< 1.250E+02
-2.002E+01 <y< 1.400E+02
0.900
Boundary plot
0
2E 1
0.700
Plasticity Indicator
* at yield in shear or vol.
X elastic, at yield in past
0.500
0.300
0.100
-0.100

-0.200
0.000
0.200
0.400
(*10^2)
0.600
0.800
1.000
1.200
Figure 2.10 Plasticity state after first cut
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:49
step
3658
-3.501E+01 <x< 1.250E+02
-2.002E+01 <y< 1.400E+02
0.900
Boundary plot
0
2E 1
0.700
o at yield in tension
0.500
0.300
0.100
-0.100

-0.200
0.000
0.200
0.400
(*10^2)
0.600
0.800
1.000
1.200
Figure 2.11 Plasticity state after second cut (M-C sandfill)
FLAC Version 5.0
2-8
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:49
step
5293
-3.502E+01 <x< 1.250E+02
-2.003E+01 <y< 1.400E+02
0.900
Boundary plot
0
2E 1
0.700
0.500
0.300
0.100
-0.100

-0.200
0.000
0.200
0.400
(*10^2)
0.600
0.800
1.000
1.200
Figure 2.12 Plasticity state after third cut (M-C sandfill)
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:49
step
7070
-3.502E+01 <x< 1.250E+02
-2.003E+01 <y< 1.400E+02
0.900
Boundary plot
0
2E 1
0.700
0.500
0.300
0.100
-0.100

-0.200
0.000
0.200
0.400
(*10^2)
0.600
0.800
Figure 2.13 Plasticity state after fourth cut (M-C sandfill)
FLAC Version 5.0
1.000
1.200
2-9
(*10^2)
FLAC (Version 5.00)
1.300
LEGEND
1.100
25-Aug-04 11:49
step
9472
-3.502E+01 <x< 1.250E+02
-2.003E+01 <y< 1.400E+02
0.900
Boundary plot
0
2E 1
0.700
0.500
0.300
0.100
-0.100

-0.200
0.000
0.200
0.400
(*10^2)
0.600
0.800
1.000
1.200
Figure 2.14 Plasticity state after fifth cut (M-C sandfill)
FLAC (Version 5.00)

(10
-02
LEGEND
25-Aug-04 11:49
step
9472
8.000
7.000
HISTORY PLOT
Y-axis :
6.000
5.000
4.000
X-axis :
Number of steps
3.000
2.000
1.000
0.000
9
(10
03

Figure 2.15 History of horizontal displacements at different locations along

pillar (M-C sandfill)
FLAC Version 5.0
2 - 10
2.4 Data File POSTP. DAT

;... STATE: INI ....
;**********************
;Mining Example Problem
;**********************
;
;Post pillar mining simulation using axisymmetric model in FLAC
;
;specify axisymmetric configuration
config axi
grid 18,24
mod mohr
prop s=14e9 b=19e9 de=2500 fri=30 coh=3e6 ten=0
gen 0,0 0,120 30,120 30,0 i=1,13
gen 30,0 30,120 90,120 90,0 i=13,19
;
;apply boundary pressures
apply p=30e6 from 1,25 to 19,25
;
;note out-of-plane stress component
ini sxx=-45e6 syy=-30e6 szz=-45e6
;
;histories
his unbal
his sxx i=1 j=8
his sxx i=1 j=10
his sxx i=1 j=12
his sxx i=1 j=14
his syy i=1 j=8
his syy i=1 j=10
his syy i=1 j=12
his syy i=1 j=14
his szz i=1 j=8
his szz i=1 j=10
his szz i=1 j=12
his szz i=1 j=14
his xdis i=5 j=8
his xdis i=5 j=10
his xdis i=5 j=12
his xdis i=5 j=14
his nste=10
FLAC Version 5.0
2 - 11
set sratio 5e-5

;
;solve to ensure at equilibrium
solve
save ini.sav
;... STATE: PP1 ....
his nste=50
set large
;
;excavate cut-and-fill stope lift 1
mod null i=5,12 j=7,8
solve
save pp1.sav
;*** BRANCH: M-C SANDFILL ****
;... STATE: PPMC2 ....
;
;
;backfill previous cut using Mohr-Coulomb model
mod mohr i=5,12 j=7,8
prop s=.3e8 b=1e8 d=1600 fric=35 coh=0 ten=0 i=5,12 j=7,8
solve
save ppmc2.sav
mod null i=5,12 j=11,12
mod mohr i=5,12 j=9,10
solve
save ppmc3.sav
;
mod null i=5,12 j=13,14
;
mod mohr i=5,12 j=11,12
solve
FLAC Version 5.0
2 - 12
save ppmc4.sav
;
mod null i=5,12 j=15,16
;
mod mohr i=5,12 j=13,14
solve
save ppmc5.sav
;*** BRANCH: D-Y SANDFILL ****
restore pp1.sav
;... STATE: PPDY2 ....
;
;
;backfill previous cut using DY model
mod dy i=5,12 j=7,8
prop s=.3e8 b=1e8 d=1600 fric=35 coh=0 ten=0 cptable=1 i=5,12 j=7,8
; relation between cap pressure and volumetric strain
table 1 0 1e4 .02 .2e6 .04 .8e6 .06 1.3e6 .08 3.5e6 .1 5.5e6
table 1 .12 8.5e6 .14 11.5e6 .16 19e6 .18 34e6 .2 50e6
;history of horizontal stress in fill
hist sxx i=9 j=7
solve
save ppdy2.sav
;
mod null i=5,12 j=11,12
;
mod dy i=5,12 j=9,10
prop s=.3e8 b=1e8 d=1600 fric=35 coh=0 ten=0 cptable 1 i=5,12 j=9,10
hist sxx i=9 j=9
solve
save ppdy3.sav
FLAC Version 5.0
2 - 13
;
mod null i=5,12 j=13,14
;
mod dy i=5,12 j=11,12
hist sxx i=9 j=11
solve
save ppdy4.sav
;
mod null i=5,12 j=15,16
;
mod dy i=5,12 j=13,14
hist sxx i=9 j=13
solve
save ppdy5.sav
;plot name: Excavation
plot hold grid
;plot name: Horizontal Disp-midpoint
plot hold history 14 line
;plot name: Principal stresses
plot hold bound stress
;plot name: Displacement of fill
plot hold bound displacement max 0.5
;plot name: Plasticity state
plot hold bound plasticity
;plot name: Horizontal disp - different locations
plot hold history 14 line 15 line 16 line 17 line
FLAC Version 5.0
2 - 14
FLAC Version 5.0
Cemented Backfill Pillar Performance
3-1
3 Cemented Backfill Pillar Performance

This example studies the failure mode of cemented backfill pillars which are represented as a MohrCoulomb material. The pillar is three-dimensional in geometry, measuring 75 m in height by 27 m
in length and 15 m in width. The material properties of the fill are as follows:
density ()
bulk modulus (K)
shear modulus (G)
cohesion (c)
friction angle ()
c
tensile strength ( tan
)
2100 kg/m3
110 MPa
37 MPa
0.1 MPa
35
0.14 MPa

The analysis is made along a two-dimensional longitudinal section through the pillar center. The
problem geometry and analysis plane are illustrated in Figure 3.1. A sliding interface is used
along one of the orebody-sandfill contacts to allow downward settling of the sand during collapse.
The boundary at the fill-ore contact, which is to be excavated, is given a roller boundary, and the
right-hand boundary is fixed in the x- and y-directions; the top is free (Figure 3.1).
The modeling sequence is as follows. First, the gravity stresses are allowed to develop in the
sandfill, and forces equilibrate across the interface. This is done elastically so that the fill will not
yield. At equilibrium, displacements are reset, cohesion is set to the proper value, and vertical
retreat mining is simulated by removing the x-direction fix along the left-face boundary in small
increments that simulate the blast height (6 m). Both x-displacement and x-velocity histories are
used to evaluate whether the system is coming to equilibrium at each step. Excavation is continued
until active collapse of the pillar occurs.
FLAC Version 5.0
3-2
15m
27m
75m
unmined
vertical retreat stope
backfill
unmined
boundaries fixed
in one direction
boundary fixed
in two directions
sliding interface
Figure 3.1
Schematic illustrating true three-dimensional backfill pillar geometry and two-dimensional representation
3.3 Results
Figures 3.2 through 3.7 show displacement vectors in the model at different blast heights. The
model comes to equilibrium for the first three blast heights (6 m, 12 m and 18 m), as indicated by
Figures 3.2 through 3.5. Only a small localized displacement is shown in each of these figures.
Note that after excavation of the 12 m blast height, the zones nearest to the left-face boundary
fail in tension. These zones are deleted in order to remove elements that will produce a bad zone
geometry as the excavation continues. The model is still in equilibrium after these zones are deleted;
Figures 3.3 and 3.4 show displacements before and after tension-failed zones were deleted. For the
FLAC Version 5.0
3-3
first three stages (6 m, 12 m and 18 m blast heights), the model can be brought to equilibrium after
each excavation is made by using the SOLVE command.
At the fourth blast height (24 m), collapse of the pillar begins to occur. The failure is shown by the
large region of downward movement that is shown in Figure 3.6. The failure is even more evident
at the 30 m blast height, as shown by Figure 3.7. The deformed boundary and plasticity states at
collapse of the pillar are shown in Figure 3.8. A history of x-displacement at the left-face boundary
is shown in Figure 3.9. This gridpoint history is reset at the fourth blast height, and indicates that
the model reaches equilibrium at this stage and collapsing in the next.
JOB TITLE : CEMENTED BACKFILL PILLAR - 6 M BLAST HEIGHT
(*10^1)
FLAC (Version 5.00)

8.000
LEGEND
15-Dec-04 14:48
step
4641
-4.155E+01 <x< 5.845E+01
-1.251E+01 <y< 8.750E+01
6.000
Boundary plot
0
2E 1
4.000
0
2E -1
2.000
0.000

-3.000
Figure 3.2
-1.000
1.000
(*10^1)
3.000
5.000
Displacement vectors at 6 m blast height
FLAC Version 5.0
3-4
JOB TITLE : CEMENTED BACKFILL PILLAR - 12 M BLAST HEIGHT
(*10^1)
FLAC (Version 5.00)

8.000
LEGEND
15-Dec-04 14:48
step
6604
-4.155E+01 <x< 5.845E+01
-1.251E+01 <y< 8.750E+01
6.000
Boundary plot
0
2E 1
4.000
0
2E -1
2.000
0.000

-3.000
Figure 3.3
-1.000
1.000
(*10^1)
3.000
5.000
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

8.000
LEGEND
13-Jan-05 14:30
step
8342
-4.157E+01 <x< 5.844E+01
-1.251E+01 <y< 8.750E+01
6.000
Boundary plot
0
2E 1
4.000
0
2E -1
2.000
0.000

-3.000
Figure 3.4
FLAC Version 5.0
-1.000
1.000
(*10^1)
3.000
5.000
Displacement vectors at 12 m blast height (tension failed zones

removed)
3-5
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

8.000
LEGEND
13-Jan-05 14:30
step
8356
-4.157E+01 <x< 5.844E+01
-1.251E+01 <y< 8.750E+01
6.000
Boundary plot
0
2E 1
4.000
0
2E -6
2.000
0.000

-3.000
Figure 3.5
-1.000
1.000
(*10^1)
3.000
5.000
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

8.000
LEGEND
13-Jan-05 14:33
step 35856
-4.157E+01 <x< 5.844E+01
-1.251E+01 <y< 8.750E+01
6.000
Boundary plot
0
2E 1
4.000
max vector = 1.228E+00
0
2E 0
2.000
0.000

-3.000
Figure 3.6
-1.000
1.000
(*10^1)
3.000
5.000
FLAC Version 5.0
3-6
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

8.000
LEGEND
13-Jan-05 14:33
step 39856
-4.157E+01 <x< 5.844E+01
-1.251E+01 <y< 8.750E+01
6.000
Boundary plot
0
2E 1
4.000
0
2E 1
2.000
0.000

-3.000
Figure 3.7
-1.000
1.000
(*10^1)
3.000
5.000
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

8.000
LEGEND
13-Jan-05 14:33
step 39856
-4.157E+01 <x< 5.844E+01
-1.251E+01 <y< 8.750E+01
6.000
Boundary plot
0
2E 1
4.000
2.000
0.000

-3.000
Figure 3.8
FLAC Version 5.0
-1.000
1.000
(*10^1)
3.000
5.000
Plasticity indicators and deformed grid at 30 m blast height
3-7
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
13-Jan-05 14:33
step 39856
0.000
HISTORY PLOT
Y-axis :
7 X displacement( 1, 11)
X-axis :
Number of steps
-0.500
-1.000
-1.500
-2.000
-2.500
-3.000
10
15
20
25
30
35
(10
03

Figure 3.9
History of x-displacement at 30 m height on pillar wall
FLAC Version 5.0
3-8
3.4 Data File SAND. DAT

;... State: init.sav ....
config
;**********************
;Mining Example Problem
;**********************
;
;Stability of cemented sand pillars
;
gri 17,25
;
;Mohr-Coulomb model, 30:1 sand:cement ratio, no tension cutoff
;
m mo
;sand
prop s=37e6 bu=110e6 de=2100 coh=1e10 fri=35 tens=1.43e10
;rock
prop s=2.292e10 b=3.056e10 d=2700 fric 35 coh 1e7 i 17 j 1 25
;
;pillar 75m high by 15m wide
gen 0,0 0,75 17,75 17,0
;
;create interface
mod null i=16
ini x add -1 i=17
;
;declare interface
int 1 as from 16,26 to 16,1 bs from 17,26 to 17,1
int 1 coh=0 fric=0 kn=1e9 ks=1e9 tbond=0
set grav=9.8
;
;boundary conditions
fix y j=1
fix x i=1
fix x i=18
his unbal
his syy i=8 j=1
;
;step to initial gravity equilibrium
solve
save init.sav
;... State: step1.sav ....
FLAC Version 5.0
3-9
;
;large strain mode on
set large
;
;turn down cohesion and tension
prop coh=0.1e6 tens=0.14e6
;turn up friction for interface
int 1 fric=35
;fix y at right face boundary
fix y i=18
;
;reset displacements
ini xdis=0 ydis=0
his xdis i=1 j=3
his xdis i=1 j=5
his xdis i=1 j=7
his xdis i=1 j=9
his xdis i=1 j=11
his xdis i=1 j=13
his xdis i=1 j=15
his xdis i=1 j=17
his xdis i=1 j=19
;
set sratio 1e-4
;
;simulate vertical retreat mining in adjacent stope by
;freeing xfix on node points
; 6 m blast height
free x i=1 j=4,5
solve
save step1.sav
;
; 12 m blast height
ini xdis=0 ydis=0
free x i=1 j=6,7
solve
save step2.sav
;... State: step2del.sav ....
;
;delete zones that failed in tension
model null i 1 j 4 6
solve
save step2del.sav
FLAC Version 5.0
3 - 10

;
; 18 m blast height
ini xdis=0 ydis=0
free x i=1 j=8,9
model null i 1 j 7
solve
save step3.sav
;
; 24 m blast height
;
these final steps will require a greater number of timesteps
;
to determine potential instability and deformation pattern
ini xdis=0 ydis=0
free x i=1 j=10,11
model null i 1 j 8
solve
save step4.sav
;
; 30 m blast height
ini xdis=0 ydis=0
free x i=1 j=12,13
step 4000
save step5.sav
;plot name: Displacement Vector
plot hold bound displacement
;plot name: Plasticity Indicator
;plot name: History of x-displacement
FLAC Version 5.0
Post-Peak Pillar Behavior and the Effects of Backfill Confinement
5-1
5 Post-Peak Pillar Behavior and the Effects of Backfill Confinement

A room-and-pillar mine is presently experiencing a slow but progressive collapse of pillars. An
attempt to answer the following questions is made. Should backfill be poured around the pillars,
which have not yet failed, in an attempt to increase their peak strength and/or residual strength?
How is the stress-strain curve for rib pillars affected by backfilling? And how important is a gap
between the backfill and the roof?
The orebody seam is approximately 8 m high and dips at 20 . Underground observation shows
that hangingwall and footwall contacts are coated with talcy minerals and exhibit substantial recent
movement.
Three cases are studied with FLAC for this example: Case 1, no backfill; Case 2, tight backfill;
and Case 3, backfill with a 10 cm gap. The FLAC grid for all three cases is shown in Figure 5.1.
Interfaces are placed between the pillar and the hangingwall, and the pillar and the footwall, to
simulate the sliding contacts. Backfill regions are created around the pillar; excavation and fill
operations are simulated via the MODEL null command.
JOB TITLE : POST-PEAK BEHAVIOR AND THE EFFECTS OF BACKFILL CONFINEMENT
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
12-Mar-04 14:37
step
1685
-1.895E+01 <x< 4.895E+01
-1.941E+01 <y< 4.849E+01
3.000
Grid plot
0
2.000
2E 1
1.000
0.000
-1.000

-1.000
0.000
1.000
2.000
3.000
4.000
(*10^1)
Figure 5.1
FLAC model of a pillar in a dipping seam
FLAC Version 5.0
5-2
The properties used in this problem are shown in Table 5.1. The strain-softening model in FLAC is
used to produce a peak/residual strength behavior for the pillar.
Backfill in confined compression exhibits a pore-collapse behavior, as shown inFigure 5.2. The
double-yield model in FLAC is used to model the backfill. The relation between cap pressure (cp )
and plastic volumetric strain (epv ) is found by fitting a simple model to a uniaxial strain test on
backfill (Clark 1991). (See Figure 5.2.) The relation found to fit the test is

cp = 1.15 10
epv
0.28 epv
1.5
(5.1)
+ 104
A table of discrete values is calculated from this relation and used in the problem. Slipping interfaces
(with = 20 and c = 100 MPa) between the pillar and hangingwall/footwall simulate a weak contact.
FISH functions are used to monitor the pillar behavior. The strain is obtained by averaging the
displacements across the ends of the pillar and dividing by the original length. Averaging the
stresses is done along the interface between pillar and footwall.
The three cases are run sequentially from the data file PILL.DAT (see Section 5.5).
Table 5.1
Data used for FLAC simulations
Model
Cohesion (MPa)
Initial
Residual
Friction Angle ( )
Initial
Residual
Bulk
(Pa)
Shear
(Pa)
2.3 109
2.3 109
600 106
Elastic (hw, fw)
Strain-Softening (pillar)
20
40
30
Double-Yield (backfill)
40
40
3 109
3 109
450 106
12
16
20
24
0.01
0.80
3.50
8.50
19.0
50.0
170.0
Cap-Pressure Table
Strain, epv (%)
Pressure, cp (MPa)
FLAC Version 5.0
5-3
50
- Applied Stress (MPa)
FLAC model
40
Laboratory test
30
20
10
0
0%
5%
10%
15%
20%
25%
- Vertical Strain (%)
Figure 5.2
Comparison of experimental and FLAC-modeled uniaxial strain

test (Clark 1991)
5.3 Results
Figures 5.3 through 5.5 show the plastic shear strain in the pillar for the three cases studied once a
vertical strain of approximately 11% in the pillar is reached. In Case 1 (no backfill), a well-formed
shear band crosses the pillar; this band is inhibited when the backfill is placed (compare Figures 5.4
and 5.5 to Figure 5.3).
The vertical-stress versus the vertical-strain (both measured in the pillar) results are plotted for the
three cases in Figure 5.6.* The backfill has little effect on the peak strength but it has a significant
effect on the post-peak behavior due to an increase in confining stress. When backfill is present,
the average pillar stress does not drop below 55 MPa, as opposed to a residual average stress of
approximately 35 MPa with no backfill present. An expected delay in stress build-up in the pillar
occurs for the case that includes a gap between roof and backfill.
* The pillar stress/strain response for each case is obtained in PILL.DAT by writing the history results to tables that are then copied in TABLE command format to three separate files (P1 RES.DAT,
P2 RES.DAT and P3 RES.DAT) using FISH I/O routines (see LOG IT.FIS in Section 5.6),
and retrieved at the completion of the three cases.
FLAC Version 5.0
5-4
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
12-Mar-04 14:50
step 11685
-1.895E+01 <x< 4.895E+01
-1.941E+01 <y< 4.849E+01
3.000
Boundary plot
0
2.000
2E 1
e_plastic
0.00E+00
5.00E-02
1.00E-01
1.50E-01
2.00E-01
2.50E-01
3.00E-01
3.50E-01
1.000
0.000

-1.000

-1.000
0.000
1.000
2.000
3.000
4.000
(*10^1)
Figure 5.3
Plastic shear strain in pillar Case 1
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
12-Mar-04 14:53
step 17362
-1.895E+01 <x< 4.895E+01
-1.941E+01 <y< 4.849E+01
3.000
Boundary plot
0
2.000
2E 1
e_plastic
0.00E+00
2.50E-02
5.00E-02
7.50E-02
1.00E-01
1.25E-01
1.50E-01
1.75E-01
1.000
0.000

-1.000

-1.000
0.000
1.000
2.000
3.000
4.000
(*10^1)
Figure 5.4
FLAC Version 5.0
Plastic shear strain in pillar and backfill Case 2
5-5
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
12-Mar-04 14:57
step 17449
-1.895E+01 <x< 4.895E+01
-1.941E+01 <y< 4.849E+01
3.000
Boundary plot
0
2.000
2E 1
e_plastic
0.00E+00
2.50E-02
5.00E-02
7.50E-02
1.00E-01
1.25E-01
1.50E-01
1.75E-01
1.000
0.000

-1.000

-1.000
0.000
1.000
2.000
3.000
4.000
(*10^1)
Figure 5.5
Plastic shear strain in pillar and backfill Case 3
JOB TITLE : VERTICAL STRESS VERSUS VERTICAL STRAIN IN PILLAR
FLAC (Version 5.00)

(10
08
LEGEND
12-Mar-04 14:58
step
0
1.400
1.200
Table Plot
backfill with 10 cm gap
tight backfill
1.000
no backfill
0.800
0.600
0.400
0.200
10
(10
-02

Figure 5.6
Vertical stress versus vertical strain in pillar
FLAC Version 5.0
5-6
5.4 Reference
Clark, I. H. The Cap Model for Stress Path Analysis of Mine Backfill Compaction Processes, in
Computer Methods and Advances in Geomechanics, pp. 1293-1298. Rotterdam: A. A. Balkema,
1991.
FLAC Version 5.0
5-7
5.5 Data File PILL. DAT

;*** Branch: no backfill ****
new
;... State: pill1eq.sav ....
config
; ***********************
; Mining Example Problem
; ***********************
;Data file illustrating pillar compression tests
gr 20,22
m e
;base case with no backfill
prop s=2.292e9 b=3.056e9 d=2700
mod null j=6
mod null j=17
gen 0,0 0,16 9.5,12.54 9.5,-3.46 rat 0.6667 0.6667 i=1,6 j=1,6
gen 9.5,-3.46 9.5,12.54 20.5,8.54 20.5,-7.46 rat 1.0 0.6667 i=6,16 j=1,6
gen 20.5,-7.46 20.5,8.54 30,5.08 30,-10.92 rat 1.5 0.6667 i=16,21 j=1,6
gen 0,16 0,24 9.5,20.54 9.5,12.54 rat 0.6667 1.0 i=1,6 j=7,17
gen 9.5,12.54 9.5,20.54 20.5,16.54 20.5,8.54 i=6,16 j=7,17
gen 20.5,8.54 20.5,16.54 30,13.08 30,5.08 rat 1.5 1.0 i=16,21 j=7,17
gen 0,24 0,40 9.5,36.54 9.5,20.54 rat 0.6667 1.5 i=1,6 j=18,23
gen 9.5,20.54 9.5,36.54 20.5,32.54 20.5,16.54 rat 1.0 1.5 i=6,16 j=18,23
gen 20.5,16.54 20.5,32.54 30,29.08 30,13.08 rat 1.5 1.5 i=16,21 j=18,23
mod null i=16,20 j=7,16
;fric of hw, fw=20 intact coh=20, , res. coh=5e6, no backfill
;strain softening model
mod ss i=6,15 j=7,16
prop s=2.292e9 b=3.056e9 d=2700 ftab=1 ctab=2 i=6,15 j=7,16
prop fric=40 coh=20e6 ten=23.8e6 i=6,15 j=7,16
table 1 0,40 .01,30 1 30
table 2 0,20e6 .01,5e6 1 5e6
inter 1 as from 1,6 to 21,6 bs from 16,7 to 6,7
inter 1 ks=1e9 kn=1e9 fric=20 coh=1e5
fix x y j=1
fix x i=1
fix x i=21
set large
;
FLAC Version 5.0
5-8
;;Define FISH function to monitor pillar performance

; vertical strain
def deforpil
altura=.5*(y(6,17)-y(6,7)+y(16,17)-y(16,7))
deforpil=(8.0-altura)/8.0
end
;
; averaging of pillar stress
def sigma
s=0
loop i (6,15)
s=s-syy(i,7)
end loop
sigma=.1*s
n num = n num + 1
end
;
;averaging of reaction stress at lower boundary
def avers
force=0.0
loop i (1,igp)
force=force-yforce(i,1)
end loop
avers=force/30.0
end
;
;averaging major and minor principal stress in pillar
def pillar1
sum1=0.0
sum3=0.0
loop i (6,15)
loop j (7,16)
temp1=-0.5*(sxx(i,j)+syy(i,j))
temp2=sqrt(sxy(i,j) 2+0.25*(sxx(i,j)-syy(i,j)) 2)
s1=max(temp1+temp2,-szz(i,j))
s3=min(temp1-temp2,-szz(i,j))
sum1=sum1+s1
sum3=sum3+s3
end loop
end loop
pillar1=sum1/100.0
pillar3=sum3/100.0
end
;
;horizontal strain at 3 locations across pillar
def hst1
FLAC Version 5.0
5-9
hst1=((xdisp(16,12)-xdisp(6,12))*c20-(ydisp(16,12)+ydisp(6,12))*c70)/l0
end
;
def const
c20=cos(.35)
c70=cos(1.22)
l0=11.7
n num = 0
end
const
;
;let interfaces settle down under gravity before applying loads
set grav 10
his unbal
solve
save pill1eq.sav
;... State: pill1.sav ....
;
ini xvel=0 yvel=0
;apply y-oriented velocities to compress pillar
ini yvel=-1e-4 j=23
ini yvel=1e-4 j=1
fix x y j=1
fix x y j=23
fix x i=1
fix x i=21
;
;servo to control y-velocity
def servo
while stepping
if unbal > 1e6 then
loop i (1,21)
yvel(i,23)=yvel(i,23)*.975
yvel(i,1)=yvel(i,1)*.975
end loop
end if
if unbal < 1e5 then
loop i (1,21)
yvel(i,23)=yvel(i,23)*1.025
yvel(i,1)=yvel(i,1)*1.025
end loop
end if
end
FLAC Version 5.0
5 - 10
;
history sigma
history deforpil
history pillar1
history pillar3
history hst1
history hst2
history hst3
history avers
history yvel i=1 j=23
;
step 10000
hist write 2 vs 3 table 1
call log it.fis
set filename = p1 res.dat
set tabin 1
log it
save pill1.sav
;*** Branch: tight backfill ****
new
config
; ***********************
; ***********************
;Data file illustrating backfilled pillar compression tests
gr 20,22
m e
;backfilled. No gap.
prop s=2.292e9 b=3.056e9 d=2700
mod null j=6
mod null j=17
mod null i=5 j=7,16
mod null i=16 j=7,16
gen 0,0 0,16 9.5,12.54 9.5,-3.46 rat 0.6667 0.6667 i=1,6 j=1,6
gen 9.5,-3.46 9.5,12.54 20.5,8.54 20.5,-7.46 rat 1.0 0.6667 i=6,16 j=1,6
gen 20.5,-7.46 20.5,8.54 30,5.08 30,-10.92 rat 1.5 0.6667 i=16,21 j=1,6
gen 0,16 0,24 9.5,20.54 9.5,12.54 rat 0.6667 1.0 i=1,5 j=7,17
gen 9.5,12.54 9.5,20.54 20.5,16.54 20.5,8.54 i=6,16 j=7,17
gen 20.5,8.54 20.5,16.54 30,13.08 30,5.08 rat 1.5 1.0 i=17,21 j=7,17
gen 0,24 0,40 9.5,36.54 9.5,20.54 rat 0.6667 1.5 i=1,6 j=18,23
gen 9.5,20.54 9.5,36.54 20.5,32.54 20.5,16.54 rat 1.0 1.5 i=6,16 j=18,23
gen 20.5,16.54 20.5,32.54 30,29.08 30,13.08 rat 1.5 1.5 i=16,21 j=18,23
;backfill a Double-Yield material, 30:1 sand/cement
FLAC Version 5.0
mod dy i=1,4 j=7,16

mod dy i=17,20 j=7,16
pro den 1000 she 6e8 bul 4.5e8 fric 40 dil 5 cptable
pro den 1000 she 6e8 bul 4.5e8 fric 40 dil 5 cptable
tab 3 0 1e4 .02 .2e6 .04 .8e6 .06 1.3e6 .08 3.5e6 .1
tab 3 .12 8.5e6 .14 11.5e6 .16 19e6 .18 34e6 .2 50e6
tab 3 .24 17e7 .26 54e7
mod ss i=6,15 j=7,16
prop s=2.292e9 b=3.056e9 d=2700 ftab=1 ctab=2 i=6,15
table 1 0,40 .01,30
table 2 0,20e6 .01,5e6
fix x y j=1
fix x i=1
fix x i=21
set large
;
; vertical strain
def deforpil
altura=.5*(y(6,17)-y(6,7)+y(16,17)-y(16,7))
end
;
def sigma
s=0
loop i (6,15)
s=s-syy(i,7)
end loop
5 - 11
3 i=1,4 j=7,16
3 i=17,20 j=7,16
5.5e6
.22 80e6
j=7,16
FLAC Version 5.0
5 - 12
sigma=.1*s
n num = n num + 1
end
;
def avers
force=0.0
loop i (1,igp)
end loop
avers=force/30.0
end
;
def pillar1
sum1=0.0
sum3=0.0
loop i (6,15)
loop j (7,16)
sum1=sum1+s1
sum3=sum3+s3
end loop
end loop
pillar1=sum1/100.0
pillar3=sum3/100.0
end
;
def hst1
end
;
def const
c20=cos(.35)
c70=cos(1.22)
l0=11.7
n num = 0
end
const
;
FLAC Version 5.0
5 - 13

set grav 10
his unbal
solve
save pill2eq.sav
;
ini xvel=0 yvel=0
ini yvel=-1e-4 j=23
ini yvel=1e-4 j=1
fix x y j=1
fix x y j=23
fix x i=1
fix x i=21
;
def servo
while stepping
if unbal > 1e6 then
loop i (1,21)
yvel(i,23)=yvel(i,23)*.975
end loop
end if
if unbal < 1e5 then
loop i (1,21)
yvel(i,23)=yvel(i,23)*1.025
end loop
end if
end
;
history sigma
history deforpil
history pillar1
history pillar3
history hst1
history hst2
history hst3
history avers
;
step 15000
FLAC Version 5.0
5 - 14
call log it.fis

set tabin 2
log it
save pill2.sav
;*** Branch: backfill with 10 cm gap ****
new
config
; ***********************
; ***********************
;Data file illustrating backfilled pillar compression tests
gr 20,22
m e
;backfilled. with 0.1m gap.
prop s=2.292e9 b=3.056e9 d=2700
mod null j=6
mod null j=17
mod null i=5 j=7,16
mod null i=16 j=7,16
gen 0,0 0,16 9.5,12.54 9.5,-3.46 rat 0.6667 0.6667 i=1,6 j=1,6
gen 9.5,-3.46 9.5,12.54 20.5,8.54 20.5,-7.46 rat 1.0 0.6667 i=6,16 j=1,6
gen 20.5,-7.46 20.5,8.54 30,5.08 30,-10.92 rat 1.5 0.6667 i=16,21 j=1,6
gen 0,16 0,23.9 9.5,20.44 9.5,12.54 rat 0.6667 1.0 i=1,5 j=7,17
gen 9.5,12.54 9.5,20.54 20.5,16.54 20.5,8.54 i=6,16 j=7,17
gen 20.5,8.54 20.5,16.44 30,12.98 30,5.08 rat 1.5 1.0 i=17,21 j=7,17
gen 0,24 0,40 9.5,36.54 9.5,20.54 rat 0.6667 1.5 i=1,6 j=18,23
gen 9.5,20.54 9.5,36.54 20.5,32.54 20.5,16.54 rat 1.0 1.5 i=6,16 j=18,23
gen 20.5,16.54 20.5,32.54 30,29.08 30,13.08 rat 1.5 1.5 i=16,21 j=18,23
;backfill a Double-Yield material, 30:1 sand/cement
mod dy i=1,4 j=7,16
mod dy i=17,20 j=7,16
pro den 1000 she 6e8 bul 4.5e8 fric 40 dil 5 cptable 3 i=1,4 j=7,16
pro den 1000 she 6e8 bul 4.5e8 fric 40 dil 5 cptable 3 i=17,20 j=7,16
tab 3 0 1e4 .02 .2e6 .04 .8e6 .06 1.3e6 .08 3.5e6 .1 5.5e6
tab 3 .12 8.5e6 .14 11.5e6 .16 19e6 .18 34e6 .2 50e6 .22 80e6
tab 3 .24 17e7 .26 54e7
mod ss i=6,15 j=7,16
prop s=2.292e9 b=3.056e9 d=2700 ftab=1 ctab=2 i=6,15 j=7,16
table 1 0,40 .01,30
table 2 0,20e6 .01,5e6
FLAC Version 5.0
5 - 15

fix x y j=1
fix x i=1
fix x i=21
set large
;
; vertical strain
def deforpil
altura=.5*(y(6,17)-y(6,7)+y(16,17)-y(16,7))
end
;
def sigma
s=0
loop i (6,15)
s=s-syy(i,7)
end loop
sigma=.1*s
n num = n num + 1
end
;
def avers
force=0.0
loop i (1,igp)
end loop
avers=force/30.0
end
;
FLAC Version 5.0
5 - 16

def pillar1
sum1=0.0
sum3=0.0
loop i (6,15)
loop j (7,16)
sum1=sum1+s1
sum3=sum3+s3
end loop
end loop
pillar1=sum1/100.0
pillar3=sum3/100.0
end
;
def hst1
end
;
def const
c20=cos(.35)
c70=cos(1.22)
l0=11.7
n num = 0
end
const
;
set grav 10
his unbal
solve
save pill3eq.sav
;
ini xvel=0 yvel=0
ini yvel=-1e-4 j=23
ini yvel=1e-4 j=1
fix x y j=1
FLAC Version 5.0
5 - 17
fix x y j=23
fix x i=1
fix x i=21
;
def servo
while stepping
if unbal > 1e6 then
loop i (1,21)
yvel(i,23)=yvel(i,23)*.975
end loop
end if
if unbal < 1e5 then
loop i (1,21)
yvel(i,23)=yvel(i,23)*1.025
end loop
end if
end
;
history sigma
history deforpil
history pillar1
history pillar3
history hst1
history hst2
history hst3
history avers
;
step 15000
call log it.fis
set tabin 3
log it
save pill3.sav
;*** Branch: compare vertical stress ****
new
;... State: compare.sav ....
config
call p1 res.dat
call p2 res.dat
FLAC Version 5.0
5 - 18
call p3 res.dat
save compare.sav
;plot name: Plastic shear strain in pillar
plot hold bound e plastic fill
;plot name: FLAC model of a pillar in a dipping seam
plot hold grid
;plot name: Vertical stress versus vertical strain in pillar
label table 1
no backfill
label table 2
tight backfill
label table 3
backfill with 10 cm gap
plot hold table 3 line 2 line 1 line
FLAC Version 5.0
5 - 19
5.6 Data File LOG IT. FIS

def log it
array p val(1500)
narr = 0
loop ii (1,n num)
narr = narr + 1
tabi = tabin
xval = xtable(tabi,ii)
yval = ytable(tabi,ii)
p val(narr) = table +string(tabi)+ +string(xval)+ +string(yval)
end loop
stat = open(filename,1,1)
stat = write(p val,narr)
stat = close
end
FLAC Version 5.0
5 - 20
FLAC Version 5.0
Full-Scale Test Wall in Sand
6-1
6 Full-Scale Test Wall in Sand

A test wall is constructed in well-graded sand and supported by shotcrete, tiebacks, and soil nails
(ASCE 1988). The nails consist of No. 10 grade-60 steel bars grouted into the soil mass. The
tieback is anchored over a 24 ft length by pressure grouting. A cross section through the test wall
is shown in Figure 6.1.
Key features of this problem include:
(1) eight excavation stages;
(2) emplaced support components at each excavation stage; and
(3) development of forces in support components as a result of soil deformation.
Figure 6.1
Cross section through test wall
The prototype wall consists of nails and tiebacks at regularly spaced vertical and horizontal intervals.
Reducing three-dimensional problems with regularly spaced reinforcement to two-dimensional
problems involves averaging the reinforcement effect in three dimensions over the distance between
the reinforcement. Donovan et al. (1984) suggest that linear scaling of material properties is a simple
and convenient way of distributing the discrete effect of reinforcement over the distance between
reinforcement in a regularly spaced pattern.
This approach is used in this analysis, based upon a one-foot problem width and the spacing for
reinforcement shown in Table 6.1. See Section 1.9.4 in Structural Elements for a description of
the properties that are scaled using this approach.
FLAC Version 5.0
6-2
Table 6.1
Reinforcement spacing
Row
1
2
3
4
5
6
7
Spacing (feet)
4.5
3.5
9.0
4.5
4.5
4.5
4.5
Note that the spacing is the horizontal reinforcement spacing (in feet) for each of the seven rows
for which the support components are installed.
The nails and tiebacks are assumed to be homogeneous, isotropic, linearly elastic materials represented as cables with the properties shown in Table 6.2.
Table 6.2
Properties for soil nails and tiebacks
Property
Row 1 Nails
Row 2 Nails
Row 3 Nails
Grouted Tieback
Ungrouted Portion
Row 4-7 Nails
Youngs Modulus (psf)
4.2 10
4.2 10
4.2 10
4.2 10
4.2 10
4.2 10
.0103
.0103
Area (ft )
Bond Stiffness (lbf /ft/ft)
Bond Strength (lbf /ft)
Yield Strength (lbf )
8.5 10
-3
6.3 10
8.5 10
-3
6.3 10
8.5 10
-3
8.5 10
-3
6.3 10
6.3 10
6.3 10
990
5000
6000
9000
6000
73620
73620
73620
222750
222750
73620
The pressure dependency of the grout bond strength is neglected for this analysis. See Section 9
for an example that includes the pressure dependency of the bond strength.
The shotcrete is assumed to be a homogeneous, isotropic, linearly elastic material represented as a
liner with the following properties:
Youngs modulus
4.80 108 psf
Poissons ratio
0.2
moment of inertia
3.00 103 ft4
area
0.333 ft2
Note that the Youngs modulus is adjusted in the FLAC analysis to account for plane-strain conditions
by dividing by (1 - 2 ). A value for E of 5.00 108 psf is used in the analysis.
FLAC Version 5.0
6-3
The soil is assumed to be homogeneous and to behave as a Mohr-Coulomb material with the
following properties:
density
3.63 slugs/ft3
friction
36
cohesion
dilation
7.5
bulk modulus
1.33 106 psf
shear modulus
0.8 106 psf

The numerical analysis is performed by first compacting the soil mass under gravity to establish
equilibrium in-situ conditions, and then sequentially excavating to various levels and introducing
support elements. The complete input data file for the analysis is contained in WALL. DAT. See
Section 6.5.
Note that the spacing property is used with the STRUCT prop command to assign the spacing in
accordance with Table 6.1 for the soil nails and tiebacks. By assigning spacing, the appropriate
input properties in Table 6.2 will be automatically scaled, and the actual forces in the cable elements
will be automatically calculated for output of results.
The various stages in the modeling sequence are discussed below.
Initial Equilibrium Stage
The FLAC grid for the problem region, shown in Figure 6.1, is created using the Build / Block tool.
The model dimensions and zoning selected for the grid are shown in the Block dialog, in Figure 6.2.
The Utility / Table tool and the Alter / Shape tool (with the Table mode) are used to define the irregular
ground surface. The final grid is shown in Figure 6.3. The grid contains 1225 zones; elements
representing soil behind the wall are one foot square, and elements farther from the wall are slightly
larger.
FLAC Version 5.0
6-4
Figure 6.2
Build
/ Block dialog
JOB TITLE : Test Wall
(*10^2)
1.150
FLAC (Version 5.00)

LEGEND
1.050
27-Aug-04 14:55
step
0
-5.333E+01 <x< 1.333E+01
5.067E+01 <y< 1.173E+02
0.950
Grid plot
0
2E 1
0.850
0.750
0.650
0.550

-4.500
-3.500
-2.500
-1.500
-0.500
0.500
(*10^1)
Figure 6.3
Plot of grid used in test wall analysis
Roller-boundary conditions are applied to the bottom and side boundaries. The ground surface is
free (i.e., no applied stress). An initial stress state given by a coefficient of earth pressure at rest of
0.45 is assumed. The problem is then timestepped to equilibrium.
Excavation Stage I
The first increment of excavation is modeled by deleting elements to a depth of 5 feet. At the same
time, liner elements representing the shotcrete face support and cable elements representing the
first row of soil nails are introduced. The model is stepped to equilibrium for this stage.
FLAC Version 5.0
6-5
Excavation Stage II
The next increment is modeled by deleting elements equivalent to 2 feet of excavation. Again,
shotcrete and nail support are installed via liner and cable elements, and the problem is stepped to
equilibrium.
Excavation Stage III
The next increment is modeled by deleting zones equivalent to 5 feet of excavation. Shotcrete and
nail support are installed, as before. The tieback at this level is modeled using a six-segment cable
with a grouted unstressed length of 24 feet connected to a one-segment cable with a pre-tensioned
ungrouted length of 21 feet. A pre-tension force of 186.2 kips is applied to the ungrouted cable to
simulate the effect of the pressure grouting. (Note that the pre-tensioning parameter is also scaled
automatically by the tieback spacing of 9 feet.) The end of the tieback at the excavation face is
connected directly to a gridpoint common to the shotcrete and soil mass, in order to simulate the
effect of the reinforced concrete pads.
In order to create the tieback, the ungrouted cable is installed first, and then the grouted cable is
installed. In this way, the properties of the grouted portion of the cable will be assigned to the node
connecting the grouted and ungrouted cables.
Excavation Stages IV VII
The next four stages consist of deleting elements to simulate soil excavation, simultaneously installing support elements, and stepping to equilibrium to allow passive forces to develop in the
reinforcement.
Excavation Stage VIII
The final stage consists of deleting elements equivalent to one foot of excavation and simultaneously
installing a liner element equivalent to one foot of shotcrete. The problem is then stepped to
equilibrium.
The model grid at the completion of all excavation stages is shown in Figure 6.4.
FLAC Version 5.0
6-6
JOB TITLE : .
(*10^2)
1.150
FLAC (Version 5.00)

LEGEND
1.050
13-Jan-05 15:24
step 106218
-5.333E+01 <x< 1.333E+01
5.069E+01 <y< 1.174E+02
0.950
Grid plot
0
2E 1
0.850
0.750
0.650
0.550

-4.500
-3.500
-2.500
-1.500
-0.500
0.500
(*10^1)
Figure 6.4
Plot of grid following removal of elements representing soil in

front of the wall
6.3 Results
The results presented in Figures 6.5 through 6.8 represent the state of the analysis following all
excavation and support installation (i.e., the end of the analysis).
Figure 6.5 plots the actual axial forces in the soil nails and tiebacks at the last stage. Note that
when spacing is specified with the STRUCT prop command, the actual forces (and moments) are
displayed in printed and plotted output for spaced structural elements.
Figures 6.6 and 6.7 plot the actual forces along the tieback and soil nails, respectively, at the
last stage. These plots are created using the FISH function tabforce listed in WALL.FIS
in Section 6.6. Scaled values for axial force are accessed directly in FLAC by the FISH offset
$kelfax. These forces must be multiplied by the spacing to obtain the actual values, which are
then stored in tables. The table numbers for the soil nail forces correspond to the excavation stages
in which the nails were installed. Note that the peak value in the tieback force distribution plot in
Figure 6.6 corresponds to the axial force in cable ID 4 (161.7 kips), shown in Figure 6.5.
Figure 6.8 shows the displacement vectors at the last stage and indicates the effect of the pretensioning in the tieback on reducing the movement of the wall.
Intermediate results at the end of each stage, as well as displacement and loads in the shotcrete, are
also available but are not presented here. Further discussion on this application is given by Lorig
(1991).
FLAC Version 5.0
6-7
JOB TITLE : .
(*10^2)
1.150
FLAC (Version 5.00)

LEGEND
1.050
13-Jan-05 15:24
step 106218
-5.333E+01 <x< 1.333E+01
5.069E+01 <y< 1.174E+02
0.950
Boundary plot
0
2E 1
Cable Plot
0.850
Axial Force on
Structure
Max. Value
# 2 (Cable) -3.728E+03
# 3 (Cable) -6.511E+03
# 4 (Cable) -1.690E+03
# 5 (Cable) -1.618E+05
# 6 (Cable) -7.486E+04
# 7 (Cable) -1.098E+04
# 8 (Cable) -1.423E+04
# 9 (Cable) -1.450E+04
#10 (Cable) -1.061E+04
Cable Plot
Shear Spring Bond Yields
0.750
0.650
0.550

-4.500
-3.500
-2.500
-1.500
-0.500
0.500
(*10^1)
Figure 6.5
Actual axial tensile forces in the soil nails and tieback at the end
of the analysis
JOB TITLE : .
FLAC (Version 5.00)

(10
02
LEGEND
1.600
13-Jan-05 15:24
step 106218
Table Plot
Axial Tensile Force (kips)
1.400
1.200
1.000
0.800
0.600
0.400
0.200
-40
-35
-30
-25
-20
-15

Figure 6.6
Distribution of axial tensile forces in the tieback distance from

the face (ft) vs actual axial force (kips)
FLAC Version 5.0
6-8
JOB TITLE : .
FLAC (Version 5.00)

(10
01
LEGEND
13-Jan-05 15:24
step 106218
1.400
1.200
Table Plot
Stage VII nail
Stage VI nail
1.000
Stage V nail
Stage IV nail
0.800
Stage III nail

Stage II nail
0.600
Stage I nail
0.400
0.200
-10
-9
-8
-7
-6
-5
-4
-3
-2

Figure 6.7
Distribution of axial tensile forces in the soil nails distance

from the face (ft) vs actual axial force (kips)
JOB TITLE : .
(*10^2)
1.150
FLAC (Version 5.00)

LEGEND
1.050
13-Jan-05 15:24
step 106218
-5.333E+01 <x< 1.333E+01
5.069E+01 <y< 1.174E+02
0.950
0
0.850
1E -1
Boundary plot
0
2E 1
0.750
0.650
0.550

-4.500
-3.500
-2.500
-1.500
-0.500
(*10^1)
Figure 6.8
FLAC Version 5.0
Displacement vectors for test wall analysis
0.500
6-9
6.4 References
ASCE (American Society of Civil Engineers). Full Scale Wall Tests Soil Nailing in Sand, Civil
Engineering, 58(5) (1988).
Donovan, K., W. G. Pariseau and M. Cepak. Finite Element Approach to Cable Bolting in Steeply
Dipping VCR Stopes, in Geomechanics Applications in Underground Hardrock Mining, pp. 6590. New York: AIME, 1984.
Lorig, L. J. Analysis of Novel Retaining Structures Using Explicit Finite Difference Codes, in
Computer Methods and Advances in Geomechanics, pp. 157-164. Rotterdam: A. A. Balkema,
1991.
FLAC Version 5.0
6 - 10
6.5 Data File WALL. DAT

;Title:Test Wall in Sand
;... State: tw0.sav ....
config
grid 35,35
gen (-50.0,65.0) (-50.0,71.0) (-20.0,71.0) (-20.0,65.0) ratio 0.9,1.0 &
i 1 11 j 1 4
gen (-50.0,71.0) (-50.0,103.0) (-20.0,103.0) (-20.0,71.0) ratio 0.9,1.0 &
i 1 11 j 4 36
gen (-20.0,65.0) (-20.0,71.0) (0.0,71.0) (0.0,65.0) i 11 31 j 1 4
gen (-20.0,71.0) (-20.0,103.0) (0.0,103.0) (0.0,71.0) i 11 31 j 4 36
gen (0.0,65.0) (0.0,71.0) (10.0,71.0) (10.0,65.0) ratio 1.2,1.0 &
i 31 36 j 1 4
gen (0.0,71.0) (0.0,103.0) (10.0,103.0) (10.0,71.0) ratio 1.2,1.0 &
i 31 36 j 4 36
model elastic i=1,35 j=1,35
table 101 delete
table 101 -60.00 100.0 -44.79 100.0 -40.29 99.00 -36.59 98.50 &
-33.59 99.00 -30.60 100.1 -27.29 102.0 -25.20 102.8 -24.00 103.0
gen table 101
ini x -25.0 y 102.8 i 8 j 36
model null region 4 34
group null region 4 34
group delete null
save tw0.sav
group sand notnull
model mohr notnull group sand
prop density=3.63 bulk=1330000.0 shear=800000.0 cohesion=0.0 &
friction=36.0 dilation=7.5 tension=0.0 notnull group sand
fix x y j 1
fix x i 36
fix x i 1 j 1 33
initial syy -4446.0 var 0.0,4446.0
initial sxx -2000.0 var 0.0,2000.0
initial szz -2000.0 var 0.0,2000.0
set gravity=32.18504
set =large
history 1 xdisp i=31, j=36
FLAC Version 5.0
6 - 11

history 7 ydisp i=31, j=36
history 999 unbalanced
solve
save tw1.sav
model null i 31 35 j 31 35
group null i 31 35 j 31 35
group delete null
initial xdisp 0 ydisp 0
initial xvel 0 yvel 0
struct node 1 grid 31,31
struct liner begin node 1 end node 2
struct prop 5001
struct prop 5001 spacing 1.0 e 5.0E8
thickness 0.333
struct node 7 -11.9,98.65
struct cable begin node 7 end node 8
struct prop 2001 spacing 4.5 e 4.2e9
yield 73620.0
solve
save tw2.sav
prop
prop
prop
prop
prop
5001
5001
5001
5001
5001
area 0.333 I 0.0030 pratio 0.2 &
seg 6 prop 2001

area 0.0085 kbond 6.3e7 sbond 990.0 &

group delete null
FLAC Version 5.0
6 - 12
struct liner begin node 14 end

struct node 16 -11.9,94.15
struct cable begin node 16 end
struct prop 2002
struct prop 2002 spacing 3.5 e
yield 73620.0
solve
save tw3.sav
node 15 prop 5001

node 1 prop 5001
node 17 seg 6 prop 2002

4.2E9 area 0.0085 kbond 6.3E7 sbond 5000.0 &

group delete null
struct liner begin node 23 end node 24 prop 5001
struct node 28 -11.9,89.65
struct cable begin node 28 end node 29 seg 6 prop 2003
struct prop 2003
struct prop 2003 spacing 9.0 e 4.2E9 area 0.0085 kbond 6.3E7 sbond 6000.0 &
yield 73620.0
struct node 36 4.3501117E-4,92.00395 slave x y 24
struct cable begin node 35 end node 36 prop 2004 tension 186200.0
struct prop 2004
struct prop 2004 spacing 9.0 e 4.2E9 area 0.0103 yield 222750.0
struct cable begin node 37 end node 35 seg 6 prop 2005
struct prop 2005
struct prop 2005 spacing 9.0 e 4.2E9 area 0.0103 kbond 6.3E7 sbond 9000.0 &
yield 222750.0
solve
save tw4.sav
FLAC Version 5.0
group delete null
struct node 47 -11.9,85.17
struct prop 2006
struct prop 2006 spacing 4.5 e
yield 73620.0
solve
save tw5.sav
group delete null
struct node 59 -11.9,80.67
solve
save tw6.sav
6 - 13
node
node
node
node
44
45
46
23
prop
prop
prop
prop
5001
5001
5001
5001

4.2E9 area 0.0085 kbond 6.3E7 sbond 6000.0 &
node
node
node
node
node
55
56
57
58
43
prop
prop
prop
prop
prop
5001
5001
5001
5001
5001

group delete null
FLAC Version 5.0
6 - 14

struct liner begin node 66
struct node 70 -11.9,76.17
struct cable begin node 70
solve
save tw7.sav
end
end
end
end
node
node
node
node
67
68
69
54
prop
prop
prop
prop
5001
5001
5001
5001
end node 71 seg 6 prop 2006

group delete null
struct node 82 -11.9,71.67
solve
save tw8.sav
node
node
node
node
node
78
79
80
81
66
prop
prop
prop
prop
prop
5001
5001
5001
5001
5001

model null i 31 35 j 5
group null i 31 35 j 5
group delete null
solve
save tw9.sav
set echo off
call Wall.fis
tabforce
save tw10.sav
FLAC Version 5.0
6 - 15

;plot name: Tieback/nail forces
plot hold bound struct cable axial fill white struct cable bond
plot hold displacement max 0.05 bound
;plot name: Grid
plot hold grid
;plot name: Axial forces in nails
label table 1
Stage I nail
label table 2
Stage II nail
label table 3
Stage III nail
label table 4
Stage IV nail
label table 5
Stage V nail
label table 6
Stage VI nail
label table 7
Stage VII nail
plot hold table 1 both 2 both 3 both 4 both 5 both 6 both 7 both
;plot name: tieback axial forces
label table 9
Axial Tensile Force (kips)
plot hold table 9 both
FLAC Version 5.0
6 - 16
6.6 Data File WALL. FIS

;Name:
;Diagram:
; --- Save cable axial forces in table --ca str.fin
def tabforce
i1 = 0
; Table 1: nail at Stage I
i2 = 0
; Table 2: nail at Stage II
i3 = 0
; Table 3: nail at Stage III
i4 = 0
; Table 4: nail at Stage IV
i5 = 0
; Table 5: nail at Stage V
i6 = 0
; Table 6: nail at Stage VI
i7 = 0
; Table 7: nail at Stage VII
i8 = 0
; Table 8: tieback
nin = imem(str pnt + $ksels)
loop while nin # 0
if imem(nin+$kelcod) = 2 then
if imem(nin+$keltyp) # 2004 then
if imem(nin+$keltyp) # 2005 then
p1 = imem(nin + $keln1)
xe = (fmem(p2+$kndx)+fmem(p1+$kndx))*0.5
ye = (fmem(p2+$kndy)+fmem(p1+$kndy))*0.5
if ye < 74.0 then
i7 = i7 + 1
xtable(7,i7) = xe
ytable(7,i7) = -fmem(nin+$kelfax)*4.5/1000.0
else
if ye < 79.0 then
i6 = i6 + 1
xtable(6,i6) = xe
else
if ye < 83.0 then
i5 = i5 + 1
xtable(5,i5) = xe
else
if ye < 88.0 then
i4 = i4 + 1
xtable(4,i4) = xe
else
if ye < 92.0 then
i3 = i3 + 1
FLAC Version 5.0
6 - 17
xtable(3,i3) = xe
else
if ye < 97.0 then
i2 = i2 + 1
xtable(2,i2) = xe
else
if ye < 102.0 then
i1 = i1 + 1
xtable(1,i1) = xe
end if
end if
end if
end if
end if
end if
end if
end if
end if
end if
; property numbers 2004 and 2005 correspond to the tieback
if imem(nin+$keltyp) = 2005 then
i8 = i8 + 1
xtable(8,i8) = xe
end if
if imem(nin+$keltyp) = 2004 then
i8 = i8 + 1
xtable(8,i8) = xe
end if
nin = imem(nin)
end loop
loop m (1,i8)
x value = xtable(8,m)
y value = ytable(8,m)
FLAC Version 5.0
6 - 18
table(9,x value) = y value

endloop
end
tabforce
FLAC Version 5.0
Stresses around a Pressurized Concrete Tunnel
7-1
7 Stresses around a Pressurized Concrete Tunnel

This example demonstrates the stress analysis of a water tunnel that is excavated in rock, subsequently lined, and then pressurized. The analysis presented here considers the lining as a monolithic
structure, and includes gravitational loading of the surrounding rock mass.
The objectives of this analysis are to simulate stress relaxation in the rock after excavation and
prior to liner installation, and to examine rock stresses after the lining is installed and the tunnel
pressurized.
The section to be analyzed is illustrated in Figure 7.1. The circular water tunnel has an excavated
radius of 5 ft 10 in, and the monolithic liner is 10 in thick. The centerline location has been
positioned, for convenience, at coordinate (0,0) and is located 30 ft below the ground surface. The
water table is 5 ft below ground surface. The overburden is composed completely of a mica-schist
rock with negligible amounts of decomposed rock and clay sediments over the top.
C
L
elev 0'
y = 30
5'
mica-schist
rock
5' 10"
5'
elev -30
y=0
0"
elev -90
y = -60
x=0
Figure 7.1
Geometry for water tunnel
It is assumed that the rock is at equilibrium under gravity loading prior to making the excavation.
Further, it is assumed that there is sufficient time delay between excavation and liner installation to
permit the perturbed stress field to come to equilibrium.
FLAC Version 5.0
7-2
The following properties describe the rock behavior:

unit weight
170 pcf
elastic modulus (E) 2 106 psi

Poissons ratio ()
0.32
friction angle ()
45
cohesion (c)
30 psi
tensile strength (t ) 30 psi
For this study, the rock is assumed to be a homogeneous, isotropic material with no anisotropy
(petrographic, bedding or jointing). The vertical in-situ stress is assumed to increase hydrostatically
with depth as the sum of the water weight and the buoyant rock weight. The ratio of horizontal
to vertical effective stress is 0.3. Depth stress gradients are assumed linear, with zero stress at the
ground surface. Gravity is specified in the analysis.
Groundwater flow is neglected for this analysis. The unlined tunnel boundary is assumed to be
impermeable, and change in pore pressure around the tunnel is assumed to only occur as a result of
mechanical deformations. In order to evaluate the effect of the tunnel excavation on pore pressure
change, two different effective-stress analyses are performed: one neglecting pore pressure change;
and the other including mechanical generation of pore pressure.
The concrete liner is a continuous monolithic structure and is assumed to behave as a homogeneous,
isotropic, linearly elastic material with the following elastic properties:
elastic modulus (E)
Poissons ratio
3.6 106 psi

0.2
The analysis focuses on the installation of a 1 ft section of the liner. In order to account for the
plane-strain conditions assumed for the rock mass, the input value for E of the liner is divided by
(1 - 2 ). The moment of inertia, I , for the lining is determined from I = t3 /12, where t is the liner
thickness.
After the liner is installed, the tunnel is pressurized to 200 psi. For this analysis, the pressurization
is performed at the same stage during which the liner is installed.
FLAC Version 5.0
7-3

The basic assumptions that are applied for the FLAC analysis include the following:
1. The geometry of the tunnel is the same along the tunnel length, permitting the
three-dimensional problem to be modeled in two dimensions as a plane-strain
analysis.
2. The layout and geometry of the tunnel are symmetric about the center of the
tunnel, permitting only half of the problem to be simulated along the line of
symmetry.
3. The excavation face is advanced instantaneously and left unlined for sufficient
time to permit the stress field to come to equilibrium. However, the time is not
sufficient for flow of groundwater into the tunnel.
4. The liner is installed and pressurized instantaneously, permitting the stress field
to respond to radial strain of the liner. The liner is also assumed to be rigidly
connected to the rock, and acts as an impermeable barrier to groundwater.*
The Build / Library tool is used to create the grid for this problem. We select the Single tunnel - refined
region item from the library list, and then alter the grid to fit the problem dimensions. The FLAC
mesh is shown in Figure 7.2. The model extends 60 ft below the centerline of the tunnel.
* For other scenarios for modeling forces on submerged lined tunnels, see Section 1.9.5 in FluidMechanical Interaction.
FLAC Version 5.0
7-4
JOB TITLE : .
(*10^1)
4.000
FLAC (Version 5.00)

LEGEND
30-Aug-04 9:45
step
0
-1.000E+01 <x< 1.100E+02
-7.500E+01 <y< 4.500E+01
2.000
Grid plot
0
0.000
2E 1
-2.000
-4.000
-6.000

0.000
0.200
0.400
0.600
0.800
1.000
(*10^2)
Figure 7.2
FLAC grid for water tunnel analysis
A roller boundary is used to model zero x-displacement along the line of symmetry at x = 0 and
the right boundary at x = 100. The bottom of the mesh is pinned in both x- and y-directions.
The modeling sequence consists of the following stages:
Stage I
Stage II
Stage III
establish equilibrium conditions to initialize stresses

excavate water tunnel and allow stresses to come to equilibrium
install tunnel lining, add pressure, and cycle to equilibrium
The lining is represented by structural liner elements connected from node to node around the
excavation. Thus, the liner segments are effectively bonded to the rock.
Each structural element has nodal endpoints that coincide with gridpoints describing the excavation
periphery. The internal tunnel pressure (200 psi) at the tunnel boundary is reduced to account for
the thickness of the liner.
The effect of groundwater pressure on the liner is obtained automatically, because the total stress
exerted by each rock zone contains a pore pressure component. As the liner deforms under load,
it will affect the rock stresses and local pore pressures. The influence on pore pressures can be
evaluated by the following approach.
Two different effective-stress analyses are performed to evaluate the influence of pore pressure. In
the first analysis, the effect of pore-pressure change on the response of the system is neglected.
FLAC Version 5.0
7-5
The WATER table command sets the pore pressure distribution, and this does not change during the
modeling stages.
In the second analysis, mechanical generation of pore pressures is included. The CONFIG gw
command and a groundwater bulk modulus of 4 107 psf are used in this case. SET ow off
is specified in order to exclude groundwater flow. The INITIAL pp command sets the initial pore
pressure distribution for this case.
The initial total stress distribution is also specified in both cases, using INITIAL sxx syy szz commands.
As a result, both analyses are at a stress equilibrium state with the same total and effective stress
distributions, before the tunnel is excavated. The initial states can be compared by viewing states
W2.SAV and GW W2.SAV. The data file (PRESSTUNNEL.DAT) including both analyses
is listed in Section 7.4.
FLAC Version 5.0
7-6
7.3 Results and Discussion

The results for the first case (no change in pore pressure) are plotted in a sequence of eight figures.
Figures 7.3 and 7.4 show effective and total principal stress plots (in psf) after stress equilibrium
for the far field and the near field. It is seen that the major principal stress is vertical. As expected,
these stresses increase with depth. The location of the water table is also shown in Figure 7.3.
The next plots, Figures 7.5 and 7.6, are of the effective and total principal stresses after the excavation
is made and stresses have come to equilibrium, but prior to liner installation. It is clear that the
principal stress tensors rotate about the circumference of the excavation for a distance of about 1
tunnel diameter. Comparing the unexcavated to excavated stresses, it can be seen that the maximum
principal stress has increased around the tunnel. However, the radial stresses are relaxed in the
vicinity of the tunnel, and there is a tensile-stress region immediately surrounding the tunnel, as
shown in Figure 7.5. Change in principal stress tensor direction between the unexcavated and
excavated states represents shear stresses due to excavation.
Figures 7.7 and 7.8 indicate the principal stress distribution after the lining is installed and pressurized, and stresses have come to equilibrium. The tensile-stress region now extends farther away
from the tunnel, all the way to the ground surface. The radial stresses surrounding the tunnel
increase as a result of the pressurization, but the tangential stresses decrease and become tensile.
This can be seen by comparing Figure 7.8 to Figure 7.6.
The plasticity plot in Figure 7.9 shows that many zones around the tunnel either are at yield, or were
at yield in the past. However, none of the zones has failed yet in tension. The currently yielded
zones are in the higher shear stress regions.
Figure 7.10 plots the pore pressure distribution at the final state. Note that the pore pressures do
not change for this analysis.
For the second analysis, pore pressures can generate as a result of mechanical volume change
during the tunnel construction stages. Pore pressure change is governed by the water bulk modulus.
Figure 7.11 shows the pore pressure distribution after the tunnel excavation, and Figure 7.12 shows
the distribution after the liner is installed and pressurized.
Figures 7.13 and 7.14 show the principal stress distribution for the analysis, including mechanical
generation of pore pressure. The effect of mechanical deformation on pore pressure is indicated in
these figures by the change in the pore-pressure contours around the tunnel (as shown in Figures 7.11
and 7.12). However, as these figures also indicate, the influence of change in pore pressure on the
effective stress distribution, and the extent of the tensile stress region, is insignificant. Compare
Figures 7.13 and 7.14 to Figures 7.5 and 7.7.
FLAC Version 5.0
7-7
JOB TITLE : .
(*10^1)
4.000
FLAC (Version 5.00)

LEGEND
30-Aug-04 9:45
step
21
-1.000E+01 <x< 1.100E+02
-7.500E+01 <y< 4.500E+01
2.000
Effective Principal Stress

Max. Value = -2.269E+01
Min. Value = -9.798E+03
0
0.000
5E 4
Water Table
-2.000
Boundary plot
0
2E 1
Tension Region Contour
-4.000
-6.000

0.000
0.200
0.400
0.600
0.800
1.000
(*10^2)
Figure 7.3
Effective principal stress distribution under gravity loading
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

2.000
LEGEND
30-Aug-04 9:45
step
21
-2.504E+01 <x< 2.904E+01
-2.704E+01 <y< 2.704E+01
1.000
Principal stresses
Max. Value = -1.589E+02
Min. Value = -9.597E+03
0
5E 4
0.000
Water Table
Boundary plot
0
1E 1
-1.000
-2.000

-2.000
Figure 7.4
-1.000
0.000
(*10^1)
1.000
2.000
Close-up view of initial total stresses before tunnel excavation
FLAC Version 5.0
7-8
JOB TITLE : .
(*10^1)
4.000
FLAC (Version 5.00)

LEGEND
30-Aug-04 9:46
step
2649
-1.000E+01 <x< 1.100E+02
-7.500E+01 <y< 4.500E+01
2.000

Min. Value = -1.009E+04
0
0.000
5E 4
Water Table
-2.000
Boundary plot
0
2E 1
-4.000
-6.000

0.000
0.200
0.400
0.600
0.800
1.000
(*10^2)
Figure 7.5
Effective principal stress distribution and tensile stress region

after tunnel excavation
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

2.000
LEGEND
30-Aug-04 9:46
step
2649
-2.504E+01 <x< 2.904E+01
-2.704E+01 <y< 2.704E+01
1.000
Principal stresses
Max. Value = -6.432E+01
Min. Value = -1.180E+04
0
5E 4
0.000
Water Table
Boundary plot
0
1E 1
-1.000
-2.000

-2.000
Figure 7.6
FLAC Version 5.0
-1.000
0.000
(*10^1)
1.000
2.000
Close-up view of total principal stress distribution after tunnel

excavation
7-9
JOB TITLE : .
(*10^1)
4.000
FLAC (Version 5.00)

LEGEND
30-Aug-04 9:47
step
7856
-1.000E+01 <x< 1.100E+02
-7.500E+01 <y< 4.500E+01
2.000

Min. Value = -1.818E+04
0
0.000
1E 5
Water Table
-2.000
Boundary plot
0
2E 1
-4.000
-6.000

0.000
0.200
0.400
0.600
0.800
1.000
(*10^2)
Figure 7.7
Effective principal stress distribution and tensile stress region

after liner installed and pressurized
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

2.000
LEGEND
30-Aug-04 9:47
step
7856
-2.504E+01 <x< 2.904E+01
-2.704E+01 <y< 2.704E+01
1.000
Principal stresses
Min. Value = -1.947E+04
0
1E 5
0.000
Water Table
Boundary plot
0
1E 1
-1.000
Liner Plot
Axial Force on
Structure
Max. Value
# 1 (Liner) -5.605E+04
Net Applied Forces
max vector =
2.320E+04
-2.000
5E 4

-2.000
Figure 7.8
-1.000
0.000
(*10^1)
1.000
2.000
Close-up view of total principal stress distribution, axial forces in

liner and applied forces at tunnel boundary after liner installed
and pressurized
FLAC Version 5.0
7 - 10
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

2.000
LEGEND
30-Aug-04 9:47
step
7856
-2.504E+01 <x< 2.904E+01
-2.704E+01 <y< 2.704E+01
1.000
Boundary plot
0
1E 1
0.000
-1.000
-2.000

-2.000
Figure 7.9
-1.000
0.000
(*10^1)
1.000
2.000
Plasticity indicators after liner installed and pressurized
JOB TITLE : .
(*10^1)
4.000
FLAC (Version 5.00)

LEGEND
30-Aug-04 9:47
step
7856
-1.000E+01 <x< 1.100E+02
-7.500E+01 <y< 4.500E+01
2.000

0.00E+00
1.00E+03
2.00E+03
3.00E+03
4.00E+03
5.00E+03
0.000
-2.000

Boundary plot
0
-4.000
2E 1
-6.000

0.000
0.200
0.400
0.600
0.800
1.000
(*10^2)
Figure 7.10 Pore pressure distribution after liner installed and pressurized
(no mechanical generation of pore pressure)
FLAC Version 5.0
7 - 11
JOB TITLE : .
(*10^1)
4.000
FLAC (Version 5.00)

LEGEND
30-Aug-04 9:49
step
4345
-1.000E+01 <x< 1.100E+02
-7.500E+01 <y< 4.500E+01
2.000

0.00E+00
1.00E+03
2.00E+03
3.00E+03
4.00E+03
5.00E+03
0.000
-2.000

Boundary plot
0
-4.000
2E 1
-6.000

0.000
0.200
0.400
0.600
0.800
1.000
(*10^2)
Figure 7.11 Pore pressure distribution after tunnel excavation (mechanical

generation of pore pressure)
JOB TITLE : .
(*10^1)
4.000
FLAC (Version 5.00)

LEGEND
30-Aug-04 9:52
step
9713
-1.000E+01 <x< 1.100E+02
-7.500E+01 <y< 4.500E+01
2.000

0.00E+00
1.00E+03
2.00E+03
3.00E+03
4.00E+03
5.00E+03
0.000
-2.000

Boundary plot
0
-4.000
2E 1
-6.000

0.000
0.200
0.400
0.600
0.800
1.000
(*10^2)
Figure 7.12 Pore pressure distribution after liner installed and pressurized
(mechanical generation of pore pressure)
FLAC Version 5.0
7 - 12
JOB TITLE : .
(*10^1)
4.000
FLAC (Version 5.00)

LEGEND
30-Aug-04 9:49
step
4345
-1.000E+01 <x< 1.100E+02
-7.500E+01 <y< 4.500E+01
2.000

Min. Value = -9.774E+03
0
0.000
5E 4
Boundary plot
0
-2.000
2E 1

-4.000
-6.000

0.000
0.200
0.400
0.600
0.800
1.000
(*10^2)
Figure 7.13 Effective principal stress distribution, pore pressure contours and
tensile stress region after tunnel excavation (mechanical generation of pore pressure included)
JOB TITLE : .
(*10^1)
4.000
FLAC (Version 5.00)

LEGEND
30-Aug-04 9:52
step
9713
-1.000E+01 <x< 1.100E+02
-7.500E+01 <y< 4.500E+01
2.000

Min. Value = -1.529E+04
0
0.000
5E 4
Boundary plot
0
-2.000
2E 1

-4.000
-6.000

0.000
0.200
0.400
0.600
0.800
1.000
(*10^2)
Figure 7.14 Effective principal stress distribution, pore pressure contours and
tensile stress region after liner installed and pressurized (mechanical generation of pore pressure included)
FLAC Version 5.0
7 - 13
7.4 Data File PRESSTUNNEL.DAT

;Title:Pressurized Concrete Tunnel
;*** Branch: effective stress ****
new
;... State: w1.sav ....
config
grid 58,51
gen 0.0,-60.0 0.0,-20.0 20.0,-20.0 20.0,-60.0 i 2 14 j 1 21
gen 20.0,-60.0 20.0,-20.0 100.0,-20.0 100.0,-60.0 ratio 1.1,1.0 i 14 34 &
j 1 21
gen 0.0,-20.0 0.0,20.0 20.0,20.0 20.0,-20.0 i 35 59 j 1 43
gen 20.0,-20.0 20.0,20.0 100.0,20.0 100.0,-20.0 ratio 1.1,1.0 i 14 34 &
j 21 42
gen 0.0,20.0 0.0,30.0 20.0,30.0 20.0,20.0 i 2 14 j 42 52
gen 20.0,20.0 20.0,30.0 100.0,29.999992 100.0,20.0 ratio 1.1,1.0 i 14 34 &
j 42 52
attach aside from 2 21 to 14 21 bside from 35 1 to 59 1
gen arc 0.0,0.0 0.0,-5.833 180.0
group rock mass - wet notnull j 1 46
model mohr notnull group rock mass - wet
prop density=5.28 bulk=2.67E8 shear=1.09E8 cohesion=4320.0 &
friction=45.0 dilation=0.0 tension=4320.0 notnull group rock mass - wet
group rock mass - dry notnull j 47 51
model mohr notnull group rock mass - dry
friction=45.0 dilation=0.0 tension=4320.0 notnull group rock mass - dry
fix x y i 2 34 j 1
fix x i 34
fix x i 2 j 42 52
fix x i 35 j 1 42
fix x i 2 j 1 21
save w1.sav
;... State: w2.sav ....
FLAC Version 5.0
7 - 14
table 1 delete
table 1 0 25.00 100.0 25.00
water density=1.938
water table=1
initial syy -15207.83 var 0.0,14451.36
initial syy -756.47
var 0.0,756.47
initial syy -8407.19 var 0.0,6800.64
initial sxx -8275.37 var 0.0,8048.43 i
initial sxx -226.94 var 0.0,226.94 i
initial sxx -4487.87 var 0.0,3787.49 i
initial szz -8275.37 var 0.0,8048.43 i
initial szz -226.94 var 0.0,226.94 i
initial szz -4487.87 var 0.0,3787.49 i
solve elastic
save w2.sav
;... State: w3.sav ....
group delete null
solve
save w3.sav
;... State: w4.sav ....
FLAC Version 5.0
i 2 33
i 2 33
i 35 58
2 33 j
2 33 j
35 58 j
2 33 j
2 33 j
35 58 j
j 1 46
j 47 51
j 1 42
1 46
47 51
1 42
1 46
47 51
1 42
7 - 15

struct prop 5001
struct node 1 fix r
struct node 27 fix r
struct prop 5001 e 5.4E8 area 0.83 I 0.048225 thickness 0.83 pratio 0.2
apply pressure 24685.7 from 35,16 to 35,28
solve
save w4.sav
;*** Branch: pore press generation ****
new
FLAC Version 5.0
7 - 16
;... State: gw w1.sav ....

config gw
grid 58,51
gen 0.0,-60.0 0.0,-20.0 20.0,-20.0 20.0,-60.0 i 2 14 j 1 21
gen 20.0,-60.0 20.0,-20.0 100.0,-20.0 100.0,-60.0 ratio 1.1,1.0 i 14 34 &
j 1 21
gen 0.0,-20.0 0.0,20.0 20.0,20.0 20.0,-20.0 i 35 59 j 1 43
gen 20.0,-20.0 20.0,20.0 100.0,20.0 100.0,-20.0 ratio 1.1,1.0 i 14 34 &
j 21 42
gen 0.0,20.0 0.0,30.0 20.0,30.0 20.0,20.0 i 2 14 j 42 52
gen 20.0,20.0 20.0,30.0 100.0,29.999992 100.0,20.0 ratio 1.1,1.0 i 14 34 &
j 42 52
gen arc 0.0,0.0 0.0,-5.833 180.0
group rock mass - dry notnull j 1 51
model mohr notnull group rock mass - dry
friction=45.0 dilation=0.0 tension=4320.0 notnull group rock mass - dry
fix x y i 2 34 j 1
fix x i 34
fix x i 2 j 42 52
fix x i 35 j 1 42
fix x i 2 j 1 21
save gw w1.sav
prop por=0.3 notnull
water density=1.938
set flow=off
initial pp 5304.31 var 0.0,-5304.31 i 2 34 j 1 47
initial pp 2808.16 var 0.0,-2496.14 i 35 59 j 1 43
initial saturation 0.0 i 2 34 j 47 52
initial syy -15207.83 var 0.0,14451.36 i 2 33 j 1 46
initial syy -756.47
var 0.0,756.47
i 2 33 j 47 51
initial syy -8407.19 var 0.0,6800.64 i 35 58 j 1 42
initial sxx -8275.37 var 0.0,8048.43 i 2 33 j 1 46
initial sxx -226.94
var 0.0,226.94
i 2 33 j 47 51
FLAC Version 5.0
initial
initial
initial
initial
history
solve
save gw
sxx
szz
szz
szz
999
-4487.87 var
-8275.37 var
-226.94
var
-4487.87 var
unbalanced
0.0,3787.49
0.0,8048.43
0.0,226.94
0.0,3787.49
7 - 17
i
i
i
i
35 58
2 33
2 33
35 58
j
j
j
j
1 42
1 46
47 51
1 42
w2.sav

group delete null
water bulk=4.0E7
solve
save gw w3.sav
FLAC Version 5.0
7 - 18

struct prop 5001
struct node 1 fix r
struct prop 5001 e 5.4E8 area 0.83 I 0.048225 thickness 0.83 pratio 0.2
solve
save gw w4.sav
;plot name: Unbalanced force
plot hold history 999
;plot name: eff. stresses
plot hold estress water white bound tcont lmagenta
;plot name: total stresses closeup
plot hold stress water lmagenta bound struct liner axial fill &
max 200000.0 white apply lmagenta
;plot name: plasticity
;plot name: tot. stresses
plot hold stress bound
FLAC Version 5.0
7 - 19
;plot name: pp
plot hold pp fill bound
;plot name: fail Mohr
set pltc 4320.0 pltf 45.0 pltt 4320.0
plot hold fail normal
;plot name: grid
plot hold grid
FLAC Version 5.0
7 - 20
FLAC Version 5.0
Displacements near the Face of an Advancing Shaft
8-1
8 Displacements near the Face of an Advancing Shaft

A circular shaft is excavated in chalk and lined with monolithic precast concrete segments. The aim
of this exercise is to determine the displacements that take place before the lining is installed. This
type of information can then be used to enable a two-dimensional plane-strain analysis to include
the effect of shaft or tunnel advancement on relaxation of shaft or tunnel loads (see Section 12).
For this example, the shafts diameter is 5.44 m, and the thickness of the lining is 22 cm. The
excavation increments are 2 m long and, after each round, a 2 m precast concrete segment is placed
right up to the face. Figure 8.1 shows a cross-section view of the process.
Symmetry
Line
r=2.5 m
0.22 m
2m
2m
STEP 1
Figure 8.1
STEP 2
STEP 3
Problem geometry and excavation steps
FLAC Version 5.0
8-2
In reality, the chalk exhibits time-dependent behavior associated with creep. The analysis presented
here is limited to the elastic response which accounts for the closure near the face. The following
properties are used:
Chalk
density
bulk modulus
shear modulus
2350 kg/m3
600 MPa
360 MPa
Concrete Liner
density
bulk modulus
shear modulus
2500 kg/m3
14 GPa
8.4 GPa
The in-situ stress state is considered to be hydrostatic, with xx = yy = zz = 1.5 MPa.

Panet (1979) published an expression relating shaft closure of an unsupported shaft to the distance
to the shaft face. The expression provides a starting point for this example. The solution calculates
radial displacements, ur , at the shaft circumference:
|y|
ur = c0 + c1 1 exp
0.7R
|y|
ur = c0 c1 1 exp
0.7R

y0
(8.1)

y0
with:
where:
R = unlined shaft radius;

ho
= 1/3;
o = in-situ pressure;
G = shear modulus; and
FLAC Version 5.0
c0 =
o R
ho
2G
c1 =
o R
(1 ho )
2G
(8.2)
8-3
|y|= distance to the face (positive in the direction of the advancing face).
Two analyses are performed for the advancing shaft:
(1) the unlined shaft is modeled and the results compared to Panets (1979) solution; and
(2) the effect of the lining is studied.

Because the axis of the shaft is an axis of radial symmetry, an axisymmetric elastic model is used
in this example. The FLAC model has dimensions 30 m by 60 m and contains 1300 zones. In order
to limit the effect of the boundary condition, 30 meters of shaft have been modeled (6 diameters).
For the unlined case, the excavation of the 30 meters has been modeled in only one step (see
BRANCH: UNLINED in SHAX.DAT in Section 8.5). For the lined case, the excavation
sequence is followed. The FISH function excav in BRANCH: LINED in SHAX.DAT) simulates
the excavation sequence. For each round, rows jj1 to jj2 are excavated, and the concrete liner is
placed in rows jj3 to jj4. Figures 8.2 and 8.3 show the zone geometry at the end of the run.
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)
3.500
LEGEND
2.500
30-Aug-04 11:37
step
1037
-2.500E+01 <x< 5.500E+01
-4.000E+01 <y< 4.000E+01
1.500
Grid plot
0
2E 1
0.500
-0.500
-1.500
-2.500
-3.500

-2.000
-1.000
0.000
1.000
2.000
3.000
4.000
5.000
(*10^1)
Figure 8.2
Zone geometry
FLAC Version 5.0
8-4
JOB TITLE : .
FLAC (Version 5.00)

5.000
LEGEND
30-Aug-04 11:37
step
1037
-6.338E-01 <x< 1.175E+01
-6.100E+00 <y< 6.284E+00
3.000
Grid plot
0
1.000
2E 0
-1.000
-3.000
-5.000

0.100
Figure 8.3
0.300
0.500
(*10^1)
0.700
0.900
1.100
Zone geometry (detail)
8.3 Results
For the analysis of the unlined shaft, the radial displacements, ur , calculated at the shaft circumference closely match Panets solution (Eq. (8.1)). Figure 8.4 compares the radial displacements as a
function of the distance to the face obtained from FLAC with Panets solution.
For the lined shaft analysis, the radial displacements that take place at the shaft interface with the
precast liner are plotted in Figure 8.5. The displacements are approximately 60% of those with no
liner. The jagged shape is due to the uneven radial displacement of the unsupported span, which is
partially confined at both ends (last ring of lining and shafts face) and relatively free to move in
the middle.
Figure 8.6 shows a magnified grid plot for the unlined shaft, while Figure 8.7 shows the same plot
for the lined shaft. The jagged shape of the deformed liner is also evident in the second plot.
Note that the use of the axisymmetry option prevents FLAC from modeling the bending resistance
accurately. However, the compression resistance is represented adequately.
FLAC Version 5.0
8-5
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
30-Aug-04 11:37
step
1037
-1.000
Table Plot
FLAC
-2.000
Panet Solution
-3.000
-4.000
-5.000
-20
-15
-10
-5
10
15
20

Figure 8.4
Comparison of radial displacements (mm) as a function of the

distance to the face (m) for the unlined shaft
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
16-Dec-04 12:26
step 27102
-1.000
Table Plot
Panet Solution
-2.000
FLAC
-3.000
-4.000
-5.000
-20
-15
-10
-5
10
15
20

Figure 8.5
Comparison of radial displacements (mm) as a function of the

distance to the face (m) for the lined shaft versus the unlined
shaft
FLAC Version 5.0
8-6
JOB TITLE : .
FLAC (Version 5.00)

5.000
LEGEND
30-Aug-04 11:37
step
1037
-6.338E-01 <x< 1.175E+01
-6.100E+00 <y< 6.284E+00
3.000
Grid plot
0
1.000
2E 0
Exaggerated Grid Distortion

Magnification = 5.000E+01
Max Disp = 5.438E-03
-1.000
-3.000
-5.000

0.100
Figure 8.6
0.300
0.500
(*10^1)
0.700
0.900
1.100
Deformed grid for the unlined shaft (magnification factor = 50)
JOB TITLE : .
FLAC (Version 5.00)

5.000
LEGEND
16-Dec-04 12:26
step 27102
-6.338E-01 <x< 1.175E+01
-6.100E+00 <y< 6.284E+00
3.000
Grid plot
0
1.000
2E 0

-1.000
-3.000
-5.000

0.100
Figure 8.7
FLAC Version 5.0
0.300
0.500
(*10^1)
0.700
0.900
1.100
Deformed grid for the lined shaft (magnification factor = 50)
8-7
8.4 Reference
Panet, M. Time-Dependent Deformations in Underground Works, in Proceedings of the 4th
ISRM Congress (Montreux), Vol. 3, pp. 279-289. Rotterdam: A. A. Balkema and the Swiss
Society for Soil and Rock Mechanics, 1979.
FLAC Version 5.0
8-8
8.5 Data File SHAX.DAT

;*** BRANCH: UNLINED ****
new
;... State: shax1.sav ....
config axi
g 20 65
mo el
gen 0 -30 0 0 2.5 0 2.5 -30 i 1 5 j 1 21 rat 1 0.9
gen s 0 30 2.5 30 s i 1 5 j 21 66
gen 2.5 0 2.5 30 30 30 30 0 i 5 21 j 21 66 rat 1.2 1
gen 2.5 -30 2.5 0 30 0 30 -30 i 5 21 j 1 21 rat 1.2 .9
ini x 2.72 i 6
pro bulk 6e8 she 3.6e8 den 2350
ini sxx=-1.5e6 syy=-1.5e6 szz=-1.5e6
app sxx=-1.5e6 i 21
fix y j 1
fix y j 66
hist unbal
hist xd i 6 j 21
hist xd i 6 j 59
mo null i 1 5 j 21 65
solve
save shax1.sav
;... State: ExApp 08a.sav ....
def cons
h0 = 1. / 3.
a1 = 1.5e-2 * 2.72 / (2. * 3.6)
a2 = -1. / (0.7 * 2.72)
c0 = -a1 * h0
c1 = -(1.0 - h0) * a1
end
cons
def panet
i = 6
k = 0
loop j (1,66)
yf = y(i,j)
if abs(yf) < 21. then
k = k + 1
xtable(10,k) = yf
yv = abs(yf)
FLAC Version 5.0
8-9
yval
= 1.-exp(yv * a2)
if yf < 0.0 then
yval = (c0 - c0 * yval)*1e3
else
yval = (c0 + c1 * yval)*1e3
end if
ytable(10,k) = yval
xtable(11,k) = yf
ytable(11,k) = xdisp(i,j)*1e3
end if
end loop
end
panet
save ExApp 08a.sav
;*** BRANCH: LINED ****
new
;... State: shax2.sav ....
config axi
g 20 65
mo el
gen 0 -30 0 0 2.5 0 2.5 -30 i 1 5 j 1 21 rat 1 0.9
gen s 0 30 2.5 30 s i 1 5 j 21 66
gen 2.5 0 2.5 30 30 30 30 0 i 5 21 j 21 66 rat 1.2 1
gen 2.5 -30 2.5 0 30 0 30 -30 i 5 21 j 1 21 rat 1.2 .9
ini x 2.72 i 6
pro bulk 6e8 she 3.6e8 den 2350
ini sxx=-1.5e6 syy=-1.5e6 szz=-1.5e6
app sxx=-1.5e6 i 21
fix y j 1
fix y j 66
hist unbal
hist xd i 5 j 21
hist xd i 5 j 61
mo null i 1 5 j 63 65
solve
def excav
loop k (1,14)
jj1=63-3*k
jj2=jj1+2
jj3=jj2+1
jj4=jj3+2
command
mo null i 1 5 j jj1 jj2
mo el i 5 j jj3 jj4
FLAC Version 5.0
8 - 10
pro bulk 14e9 shear 8.4e9 den 2500 i 5 j jj3 jj4

solve
end command
end loop
end
excav
save shax2.sav
;... State: ExApp 08b.sav ....
def cons
h0 = 1. / 3.
a1 = 1.5e-2 * 2.72 / (2. * 3.6)
a2 = -1. / (0.7 * 2.72)
c0 = -a1 * h0
c1 = -(1.0 - h0) * a1
end
cons
def panet
i = 6
k = 0
loop j (1,66)
yf = y(i,j)
if abs(yf) < 21. then
k = k + 1
xtable(10,k) = yf
yv = abs(yf)
yval
= 1.-exp(yv * a2)
if yf < 0.0 then
yval = (c0 - c0 * yval)*1e3
else
yval = (c0 + c1 * yval)*1e3
end if
ytable(10,k) = yval
xtable(11,k) = yf
ytable(11,k) = xdisp(i,j)*1e3
end if
end loop
end
panet
save ExApp 08b.sav
;plot name: grid
plot hold grid
;plot name: zone geometry(detail)
plot hold grid
FLAC Version 5.0
8 - 11
;plot name: Deformed grid

plot hold grid red grid magnify 50.0 white
;plot name: Radial displacement
label table 10
Panet Solution
label table 11
FLAC
plot hold table 11 line 10 line
FLAC Version 5.0
8 - 12
FLAC Version 5.0
Simulation of Pull-Tests for Grouted Cable Anchors
9-1
9 Simulation of Pull-Tests for Grouted Cable Anchors

The most common method for determination of cable bolt properties is to perform pull-out tests on
small segments of grouted cables in the field. Typically, segments from 10 to 50 cm in length are
grouted into boreholes. The ends of these segments are pulled with a jack mounted to the surface
of the tunnel and connected to cable via a barrel-and-wedge type anchor. The force applied to the
cable, and the deformation of the cable are plotted to produce an axial force-deflection curve. From
this curve, the peak shear strength of the grout bond is determined and converted to a strength in
tons/m cable length.
The results for pull-tests on one-half meter segments of several types of cables are illustrated in
Figure 9.1. These plots are expressed in terms of tons/m versus deformation in mm. For all cables,
a water/cement ratio of one-third was used.
30
15.2 mm
2 x 15.2 mm
25
24 mm
Force (tons/m)
20
26 mm
15
10
0
0
Figure 9.1
10
15
20
Displacement (mm)
25
30
Field results for pull-tests on various types of cables for a bond

length of 0.5 m and a water/cement ratio of 1/3
FLAC Version 5.0
9-2
The properties of typical concrete-reinforcing tendons are given in Table 9.1.
Table 9.1
Typical stranded cable properties
Nominal
diameter (mm)
Nominal
mass (kg/m)
Nominal
area (mm2 )
Grade
Force (kN) at the following

% of ultimate tensile strength
55% 65% 75% 100%
15
1.54
177
Super
105
125
145
190
19
2.42
283
Super
170
200
230
305
23
3.49
415
Reg
Super
235
250
280
295
325
340
430
450
26
4.43
530
Reg
Super
300
315
355
370
410
430
545
570
29
5.48
660
Reg
Super
375
390
440
460
510
535
680
710
32
6.64
804
Reg
Super
455
480
540
565
625
655
830
870
35
7.91
962
Reg
Super
545
570
645
675
745
780
990
1040
38
9.29
1140
Reg
Super
645
675
765
800
880
925
1175
1230
In this example problem, FLAC is used to model the previous tests and reproduce the field data. It is
noted that the current cable bolt model in FLAC describes the response of the cable-rock interaction
as cohesive and frictional in nature. The model assumes the grout behaves as an elastic, perfectly
plastic material with confining stress dependence but no loss of strength after failure. Recent field
and laboratory work (e.g., Kaiser et al. 1992) has shown that dilation introduced by the spiral cable
strands can be an important factor in cable response. This behavior is not addressed here, but could
be accounted for through the use of a FISH routine.
FLAC Version 5.0
9-3

First, we consider the case where the confining stress dependence on the cable shear bond strength
may be neglected. The cable properties required by FLAC s cable-bolt model must be extracted
from the field pull-test curve. This is easily done when the field test data are presented in terms of
force/unit length versus deformation, as shown in Figure 9.1. Assuming no yielding in the cable,
the value of the grout shear stiffness, kbond, is simply the slope of the curve, with the ultimate bond
strength, sbond, being the peak-pull strength value per unit length.
For example, all of the pull-test results shown here have roughly the same loading slope; so, an
average value of kbond is chosen for all:
kbond
28 tons/meter
= 1.12 107 N/m/m
3
25 10 meter
This value of kbond is very low, indicating a rather poor grouting job for the cable. Typical kbond
values would be approximately one order of magnitude or more higher than this.
The value of sbond for the single 15.2 mm wire is simply the peak shear resistance in tons/m. In this
case, sbond 17.5 tons/m, or 17.5 104 N/m. To check this value of sbond for reasonableness,
it can be converted to grout shear strength by dividing by the approximate surface area of the
wire (assuming the bond fails at the grout/cable interface). We find that the peak shear strength
is 3.66 MPa. This value should equal roughly half the uniaxial compressive strength of the grout,
indicating either a very poor grout or that the cable was allowed to rotate during the pull-test,
yielding artificially low grout shear-strength values.*
At Branch: Single 15.2 mm Cable in the data file PULL.DAT (see Section 9.4) is the command
listing for a simple example of a pull-test using FLAC. The cable end-node is pulled at a small,
constant y-oriented velocity (Figure 9.2). A FISH function, ff, is used to sum the reaction forces
and monitor nodal displacement generated by the pull-tests for comparison to field test results.
* This effect is explored in some detail in Hyett et al., 1992.
FLAC Version 5.0
9-4
End of cable pulled at velocity, V
This surface fixed in y-direction
Grouted cable length 0.5 m
Figure 9.2
Schematic of geometry of FLAC model for a pull test
A plot of history 1 versus history 2 (pull force versus cable displacement) for the case of a single
15.2 mm cable is shown in Figure 9.3. This figure illustrates the general force-displacement behavior
given in Figure 9.1. The peak force is reached at a displacement of approximately 17 mm. After
this point, the cable is simply pulled out of the borehole in much the same fashion as a block sliding
on a plane. Figures 9.4(a-c) show the axial force distribution on the cable for displacements of 10
mm, 17 mm and 17.5 mm, respectively. Superimposed on the axial forces are locations at which
the grout bond is yielding. At 10 mm [Figure 9.4(a)], the grout bond has not failed. At 17 mm
[Figure 9.4(b)], bond failure is initiated and rapidly propagates [Figure 9.4(c)] down the entire cable
length. At that stage, the force on the cable end is simply the sum of sbond li (where li is the length
of cable segments) for all n-slipping segments. If the embedded length were long enough, the cable
axial force would eventually reach the yield force limit of the cable itself. The cable should then
break when the extension strain equals the ultimate breaking strain of the cable (generally, around
3%). The cable model does not have an extension strain limit; the rockbolt model (see Section 1.6
in Structural Elements) should be used to simulate this condition.
FLAC Version 5.0
9-5
JOB TITLE : PULL OUT TEST FOR SINGLE 15.2 MM CABLE
FLAC (Version 5.00)

(10
05
LEGEND
1.600
30-Aug-04 15:26
step 20000
1.400
HISTORY PLOT
Y-axis :
ff
(FISH)
X-axis :
dd
(FISH)
1.200
1.000
0.800
0.600
0.400
0.200
10
12
14
16
18
20
(10
-03

Figure 9.3
Cable pull force in N/m versus cable displacement in meters for

the case of a single 15.2 mm grouted cable
FLAC Version 5.0
9-6
FLAC (Version 5.00)

0.750
LEGEND
0.650
30-Aug-04 15:26
step 10000
-2.733E-01 <x< 6.733E-01
-1.183E-01 <y< 8.283E-01
0.550
Boundary plot
0.450
2E -1
Cable Plot
Axial Force on
Structure
Max. Value
# 1 (Cable) -4.938E+04
Cable Plot
0.350
0.250
0.150
0.050
-0.050

-0.200
-0.100
0.000
0.100
0.200
0.300
0.400
0.500
0.600
(a) at 10 mm deformation
FLAC (Version 5.00)

0.750
LEGEND
0.650
30-Aug-04 15:26
step 17000
-2.780E-01 <x< 6.780E-01
-1.195E-01 <y< 8.365E-01
0.550
Boundary plot
0.450
2E -1
Cable Plot
Axial Force on
Structure
Max. Value
# 1 (Cable) -8.258E+04
Cable Plot
0.350
0.250
0.150
0.050
-0.050

-0.200
-0.100
0.000
0.100
0.200
0.300
(b) at 17 mm deformation
FLAC Version 5.0
0.400
0.500
0.600
9-7
FLAC (Version 5.00)

0.750
LEGEND
0.650
30-Aug-04 15:26
step 17500
-2.783E-01 <x< 6.783E-01
-1.196E-01 <y< 8.371E-01
0.550
Boundary plot
0.450
2E -1
Cable Plot
Axial Force on
Structure
Max. Value
# 1 (Cable) -8.402E+04
Cable Plot
0.350
0.250
0.150
0.050
-0.050

-0.200
Figure 9.4
-0.100
0.000
0.100
0.200
0.300
0.400
0.500
0.600
(c) at 17.5 mm deformation

Plot of axial force and cable bond yield points for pull-test simulation on a 15.2 mm cable bolt (Note that cable-bond-slip progresses
rapidly after peak strength is reached at the first cable element.)
The evolution of the force profile along the cable is illustrated for the case of the 26 mm cables (Figure 9.5). See Branch:26 mm Cable in PULL.DAT in Section 9.4. Here, the force-displacement
profiles for various snapshots in displacement are compared. Table 10 is labeled 5 mm displacement, and the remaining tables are labeled at 5 mm increments up to 30 mm. (Note that the last
three values in each table are for structural nodes outside the grid and should be neglected.) Up
to the point of approximately 20 mm, bond slippage does not occur, and the increase in axial load
is essentially elastic; however, bond slippage occurs rapidly between 20 and 25 mm (Tables 40
and 50), with a constant force distribution thereafter. The slope of the final curve is approximately
sbond/L, where L is the initial grouted length of the cable.
FLAC Version 5.0
9-8
JOB TITLE : PULL OUT TEST FOR SINGLE 26 MM CABLE
FLAC (Version 5.00)

(10
01
LEGEND
30-Aug-04 15:26
step 30000
2.400
axial force along cable

30 mm displ
2.000
25 mm displ
20 mm displ
1.600
15 mm displ
10 mm displ
1.200
5 mm displ
0.800
0.400
10
20
30
40
50
60
(10
-02

Figure 9.5
Evolution of the final axial force distribution (in tons/m) for a

pull-test simulation on 26 mm cables in a single borehole (Note
that the last three points in each table are outside the grid and
should be neglected.)
The cable shear bond strength will, in general, increase with increasing effective pressure p acting
on the cable. A linear law is implemented in FLAC, whereby the cable shear bond strength is defined
as a constant (sbond) plus the effective pressure on the cable multiplied by the cable perimeter
(perimeter) times a friction angle (sfriction). This pressure dependence is activated automatically in
FLAC by issuing the cable properties perimeter and sfriction. Note that, in this case, the input data
for sbond must correspond to the shear bond strength in a cable pull-out test carried out without
confining pressure. Numerical results of pull-out tests on the 15.2 mm cable are presented for a
friction angle of 20 and three levels of initial confining pressure, namely p = 105 , 106 and 107
N/m2 , in Figures 9.6 to 9.8. Those figures indicate an increasing failure level with increasing initial
confining pressure illustrating the frictional character of the cable-rock interface. Results for the
pull-out test with confined pressure on the 15.2 mm cable were obtained in PULL.DAT at the
branch named Branch: 15.2 mm cab. with confining pressure in Section 9.4.
FLAC Version 5.0
9-9
FLAC (Version 5.00)

(10
05
LEGEND
31-Aug-04 8:44
step 40000
3.500
3.000
HISTORY PLOT
Y-axis :
ff
(FISH)
X-axis :
dd
(FISH)
2.500
2.000
1.500
1.000
0.500
10
15
20
25
30
35
40
(10
-03

Figure 9.6
Pull-out test on 15.2 mm cable p = 105 N/m2
FLAC (Version 5.00)

(10
05
LEGEND
31-Aug-04 8:44
step 40000
3.500
3.000
HISTORY PLOT
Y-axis :
ff
(FISH)
X-axis :
dd
(FISH)
2.500
2.000
1.500
1.000
0.500
10
15
20
25
30
35
40
(10
-03

Figure 9.7
FLAC Version 5.0
9 - 10
FLAC (Version 5.00)

(10
05
LEGEND
31-Aug-04 8:44
step 40000
3.500
3.000
HISTORY PLOT
Y-axis :
ff
(FISH)
X-axis :
dd
(FISH)
2.500
2.000
1.500
1.000
0.500
10
15
20
25
30
35
40
(10
-03

Figure 9.8
9.3 References
Hyett, A. J., W. F. Bawden and R. D. Reichert. The Effect of Rock Mass Confinement on the
Bond Strength of Fully Grouted Cable Bolts, Int. J. Rock Mech. Min. Sci. & Geomech. Abstr.,
29(5), pp. 503-524, 1992.
Kaiser, P. K., M. Diederichs and S. Yazici. Cable Bolt Performance During Mining Induced Stress
Change Three Case Examples, in Rock Support in Mining and Underground Construction
(Proceedings of the International Symposium on Rock Support, Sudbury, Ontario, Canada, June,
1992), pp. 377-384. P. K. Kaiser and D. R. McCreath, Eds. Rotterdam: A. A. Balkema, 1992.
FLAC Version 5.0
9 - 11
9.4 Data File PULL.DAT

;*** Branch: Single 15.2 mm cable ****
new
;... State: pull a.sav ....
config
;pull.dat
g 4 6
mo el
ini x mul .1
ini y mul .1
pro bulk 5e9 she 3e9 den 2000
fix y j 7
set large
set st damp struc combined
; --- Cable installation --stru cable beg .2 0.1 end .2 .7 seg 12
stru pro 1 e 98.6e9 yield 232.13e3 a 181e-6 sbond 1.75e5 kbond 1.12e7
; --- Fish functions --;
ff : Pull force per unit cable length
;
dd : Displacement of cable end
def ff
sum = 0.0
loop i (1,igp)
sum = sum+yforce(i,7)
end loop
ff=2.0*sum
; multiply by 2.0 because length is 0.5 m
dd=step*1e-6
end
; --- Histories --his nstep 100
hist ff
hist dd
hist unbal
; --- Pull out test --stru node 13 fix y ini yvel 1e-6
step 10000
; 10 mm deformation
;plot hold bou struc ax struc bo
save pull a.sav
;... State: pull b.sav ....
step 7000
FLAC Version 5.0
9 - 12
; 17 mm deformation
save pull b.sav
;... State: pull c.sav ....
step 500
; 17.5 mm deformation
save pull c.sav
;... State: pull.sav ....
step 2500
; 20 mm deformation
;plot hold his 1 vs 2
save pull.sav
;*** Branch: 26 mm cable ****
new
;... State: pull1.sav ....
config
;pull1.dat
g 4 6
mo el
ini x mul .1
ini y mul .1
fix y j 7
set large
set st damp struct combined
stru pro 1 e 98.6e9 yield 545.e3 a 530.e-6 sbond 2.63e5 kbond 1.12e7
; --- Fish functions --;
;
def ff
sum = 0.0
loop i (1,igp)
end loop
ff=2.0*sum
dd=step*1e-6
end
ca str.fin
def fill tab
FLAC Version 5.0
;
;
9 - 13
xval: distance from initial cable deep end [m]

yval: axial force by unit cable length
[tons/m]
k = 0
eind = imem(str pnt+$ksels)
loop while eind # 0
pt1 = imem(eind+$keln1)
pt2 = imem(eind+$keln2)
k
= k + 1
xval = (fmem(pt1+$kndy)+fmem(pt2+$kndy))*0.5 - 0.1
yval = -fmem(eind+$kelfax)*2.0e-4
xtable(itab,k) = xval
ytable(itab,k) = yval
eind = imem(eind)
end loop
end
def test
command
step 5000
end command
itab = itab + 10
fill tab
end
hist ff
hist dd
hist unbal
test
save pull1.sav
test
save pull2.sav
test
save pull3.sav
test
save pull4.sav
test
FLAC Version 5.0
9 - 14
save pull5.sav
test
save pull6.sav
;*** Branch: 15.2 mm cab. with confining pressure ****
new
;... State: pull2p.sav ....
config
;pull2.dat
g 4 6
mo el
ini x mul .1
ini y mul .1
fix y j 7
; --- Settings --set large
set st damp struc combined
stru pro 1 e 98.6e9 yield 232.13e3 a 181e-6 sbond 1.75e5 kbond 1.12e7
stru pro 1 sfriction 20.0 perimeter 47.75e-3
; --- Fish functions --def con p
cpm = - cp
command
ini sxx cpm szz cpm
app pr cp i 1
app pr cp i 5
end command
end
;
;
def ff
sum = 0.0
loop i (1,igp)
end loop
ff=2.0*sum
dd=step*1e-6
end
FLAC Version 5.0
9 - 15
hist ff
hist dd
hist unbal
save pull2p.sav
;*** Branch: c.p. =1e5 ****
;... State: pullp1.sav ....
set cp = 1e5
con p
step 40000
save pullp1.sav
;*** Branch: c.p.=1e6 ****
restore pull2p.sav
set cp = 1e6
con p
step 40000
save pullp2.sav
;*** Branch: c.p.=1e7 ****
restore pull2p.sav
set cp = 1e7
con p
step 40000
save pullp3.sav
;plot name: Pull force vs displ
plot hold history 1 line vs 2
;plot name: Axial force and cable bond yield points
plot hold bound struct cable axial max 90000.0 struct cable bond
;plot name: Axial force along cable
label table 10
5 mm displ
label table 20
10 mm displ
label table 30
15 mm displ
label table 40
FLAC Version 5.0
9 - 16
20 mm displ
label table 50
25 mm displ
label table 60
30 mm displ
plot hold table 60 both 50 both 40 both 30 both 20 both 10 both alias &
axial force along cable
FLAC Version 5.0
Slope Stability Analysis of a Rock Slope
10 - 1
10 Slope Stability Analysis of a Rock Slope

Limit equilibrium methods are commonly used to evaluate the stability of slopes in rock or soil.
For materials in which well-defined structural patterns (e.g., features such as bedding planes or
joints) are not present, rupture surfaces develop naturally along planes or curved surfaces within
the slope. In soils, the rupture surface is commonly observed to be circular: this is the basis for
many limit-equilibrium stability theories.
One common method is the Bishop simplified method of slices (Bishop 1955). This method is
one of several used by Hoek and Bray (1981) to produce series of slope stability charts for circular
failure. A limiting equilibrium condition for the development of a tensile crack can be included
with the circular failure, as well as the influence of water pressure due to the presence of a phreatic
surface in the slope. The charts developed by Hoek and Bray can be used to determine the location
of the critical failure circle and tension crack, and find the associated factor of safety for stability
of the slope.
An example application of Bishops method is given by Hoek and Bray for the case of a slope
excavated in highly weathered granitic rock. The slope contains three 15 m high benches with two
8 m wide beams. The bench faces are inclined at 75 to the horizontal, and the top of the slope is
cut at 45 from the top of the third bench to the ground surface. Figure 10.1 illustrates the geometry
of the slope.
critical center for = 45
tension
crack
a
R
5m
10
20
2
Scale m
3
4
a/R = 0.401
5
8
Figure 10.1 Failure surface solution from Bishops method for a rock slope
[Hoek and Bray (1981)]
FLAC Version 5.0
10 - 2
The rock mass is classified as a Hoek-Brown material with strength parameters of:
m
s
c
cm
=
=
=
=
0.13
0.00001
150
MPa
s c = 0.47 MPa
The tensile strength is estimated to be 0.012 MPa. For the Bishops method, a tangent to the curved
Hoek-Brown failure envelope is drawn at a normal stress level estimated from the slope geometry.
Mohr-Coulomb properties for friction angle and cohesive strength are then estimated to be (see
HOEK.FIS in Section 3 in the FISH volume):
= 45
c = 0.14 MPa
The mass density of the dry rock mass is 2500 kg/m3 , and the mass density of the saturated rock
mass is 2800 kg/m3 . The phreatic surface is located as shown in Figure 10.1, and the mass density
of water is 1000 kg/m3 .
Based upon these parameters, Hoek and Bray report that the Bishop method produces a location
for the circular failure surface and tension crack, as shown in Figure 10.1, and a factor of safety of
1.423.
In FLAC, the failure surface can evolve during the calculation in a way that is representative of the
natural evolution of the physical failure plane in the slope. It is not necessary to make an estimate
for the location of the circular failure line when beginning an analysis, as it is with limit-equilibrium
methods. FLAC will find the failure plane and the failure mechanism by simulating the material
behavior directly. Run the data file ROCKSL.DAT in Section 10.5 to perform this analysis.
A reasonably fine grid should be selected to ensure that the failure plane will be well-defined as
it develops. It is best to use the finest grid possible when studying problems involving localized
failure (see Section 3.2.4 in the Users Guide). For the bench-cut slope, a model grid is created as
shown in Figure 10.2.
FLAC Version 5.0
10 - 3
JOB TITLE : SLOPE STABILITY ANALYSIS OF A ROCK SLOPE
(*10^2)
FLAC (Version 5.00)

1.000
LEGEND
31-Aug-04 12:07
step
0
-5.000E+01 <x< 1.500E+02
-8.000E+01 <y< 1.200E+02
0.600
Grid plot
0
5E 1
Water Table
0.200
-0.200
-0.600

-0.200
0.200
0.600
1.000
1.400
(*10^2)
Figure 10.2 FLAC model grid with water table

FLAC can perform a factor-of-safety calculation by using the strength reduction method as described
in note 12 in Section 3.8 in the Users Guide. The method is implemented by invoking the SOLVE
fos command. In this approach, the strength of the material is reduced until a failure of the slope
occurs. Both friction angle and cohesion are reduced simultaneously by a constant factor, and
FLAC runs are automatically made with each new pair of strength parameters using a bracketing
approach until a safety factor is found.
We begin this analysis at the strength parameters selected by Hoek and Bray to characterize the
slope ( = 45 , c = 0.14 MPa and tf = 0.012 MPa). The model is first brought to an equilibrium
state assuming an unsaturated condition. We use the SOLVE elastic command to minimize plastic
yield during the initial development of gravitational stresses. We then introduce the water table and
adjust the mass density below the phreatic surface to its saturated value (using the FISH function
wet den). Note that the total stresses are adjusted automatically (by specifying CONFIG ats at
the start) to correspond to the increase in pore pressure resulting from the WATER table command.
The model is stepped to equilibrium again, and the resulting stress state is almost entirely within
the elastic range for the imposed conditions, as can be seen from the failure envelope plot shown
in Figure 10.3.
FLAC Version 5.0
10 - 4
FLAC (Version 5.00)

(10
06
LEGEND
31-Aug-04 12:43
step
4909
Failure Surface Plot
Shear Stress vs
Normal Stress
Zone Stress States
Mohr-Coulomb Fail. Surf.
Friction = 4.5000E+01
Cohesion = 1.4000E+05
Tension = 1.2000E+04
1.400
1.200
1.000
0.800
0.600
0.400
0.200
10
12
14
(10
05

Figure 10.3 Failure surface and initial zone stresses

The factor-of-safety calculation is then begun by specifying SOLVE fos. With the strength reduction
approach, the factor of safety, fs , adjusts the friction and cohesion as follows.
The reduced friction angle, r , is
r = arctan(tan /fs )
and the reduced cohesion, cr , is
cr = c/fs
Note also that we simulate a loss in tensile strength when the tensile strength limit is reached. We
assume that the tensile strength drops to zero instantaneously; this is prescribed by default with the
Mohr-Coulomb model. (Tensile softening as a function of plastic tensile strain can be prescribed
with the strain-softening model.)
FLAC Version 5.0
10 - 5
10.3 Results
While FLAC is executing the SOLVE fos command, the bracketing values for fs are printed continuously to the screen. When completed, the final value for fs is displayed. In this case, the calculated
fs is 1.41.
The failure surface is identified by the plots in Figures 10.4 and 10.5. These plots are generated
after restoring the file ROCKSL FOS.SAV,
In Figure 10.4, the contours of shear strain rate indicate the plane of shear failure, while the plot of
zero tensile strength within zones denotes the region in which tensile failure occurs (and the tensile
strength drops to zero). In Figure 10.5, the plot of velocity vectors also indicates the pattern of
motion at the initiation of failure.
Compare Figures 10.4 and 10.5 to Figure 10.1. The failure surface in FLAC closely resembles that
produced from the Bishop solution. However, the tensile failure extends farther up the slope in
the FLAC solution. It is important to recognize that the limit equilibrium solution only identifies
the onset of failure, whereas the FLAC solution includes the effect of stress redistribution and
progressive failure after failure has been initiated. In this problem, tensile failure continues up the
slope as a result of the tensile softening. The resulting factor of safety allows for this weakening
effect.
This rock slope example is also run using FLAC/Slope, see Section 1.4.5 in the FLAC/Slope Users
Guide. The factor of safety calculated in the FLAC/Slope simulation is 1.38, and the failure pattern
is slightly different. For example, compare Figure 10.5 to Figure 1.96 in the FLAC/Slope Users
Guide. This difference is related to the effect of the loading path on the development of tensile
failure in the model. In FLAC/Slope, the SOLVE fos solution is invoked from an initial stress state
of zero in the model. In this example, the model is brought to an equilibrium stress state, with only
a minor amount of tensile failure, before the SOLVE fos solution is performed. The approach in
FLAC/Slope produces slightly more tensile failure in the model. Note that if the tensile strength
is set to a high value (e.g., 0.14 MPa), then no tensile failure occurs, and the calculated factor of
safety is the same for both simulations: 1.41.
More detailed comparisons of FLAC to limit equilibrium solutions have been made by others (e.g.,
Thompson 1993 and Dawson et al., 1999).
FLAC Version 5.0
10 - 6
(*10^2)
FLAC (Version 5.00)

1.000
LEGEND
31-Aug-04 13:18
step 25225
-5.000E+01 <x< 1.500E+02
-8.000E+01 <y< 1.200E+02
0.600
Factor of Safety 1.41

tension
0.000E+00
1.200E+04
Max. shear strain increment
0.200

Minimum: 0.00E+00
Maximum: 4.00E+00
Boundary plot
0
5E 1
-0.200
-0.600

-0.200
0.200
0.600
1.000
1.400
(*10^2)
Figure 10.4 Failure surface in slope indicated by shear strain contours and
zero tension zones
(*10^2)
FLAC (Version 5.00)

1.000
LEGEND
31-Aug-04 13:18
step 25225
-5.000E+01 <x< 1.500E+02
-8.000E+01 <y< 1.200E+02
0.600

User-defined Groups
rock mass
Water Table
Velocity vectors
0
0.200
1E -2
-0.200
-0.600

-0.200
0.200
0.600
1.000
1.400
(*10^2)
Figure 10.5 Failure surface in slope indicated by velocity vectors and plasticity
indicators
FLAC Version 5.0
10 - 7
10.4 References
Bishop, A. W. The Use of the Slip Circle in the Stability Analysis of Earth Slopes, Gotechnique,
5, 7-17 (1955).
Dawson, E. M., W. H. Roth and A. Drescher. Slope Stability Analysis by Strength Reduction,
Gotechnique, 49(6), 835-840 (1999).
Hoek, E., and J. Bray. Rock Slope Engineering. London: IMM, 1981.
Thompson, R. J. The Location of Critical Slip Surfaces in Slope-Stability Problems, J. S. Afr.
Inst. Min. Metall., 93(4), 85-95 (1993).
FLAC Version 5.0
10 - 8
10.5 Data File ROCKSL.DAT

;... State: rocsl 1.sav ....
;---------------------------------------------------------------------; rocksl.dat
rock slope stability analysis
;
comparison to solution in Hoek & Bray, 1981
;---------------------------------------------------------------------config ats
grid 60,40
gen -40.0,-40.0 -40.0,80.0 140.0,80.0 140.0,-40.0 i=1,61 j=1,41
model elastic
; Add top surface
table 1 (-130.0 -40.00) (-130.0 0) (-40.00 0) (0 0) (3.880 15.00) (11.88 &
15.00) (15.76 30.00) (23.76 30.00) (27.63 45.00) (43.15 60.50) (80.00 &
76.00) (140.0 80.00) (230.0 80.00) (230.0 -40.00)
gen table 1
;
fix x y j=1
fix x i=1
fix x i=61
;
group rock mass reg tab 1
model mohr group rock mass notnull
prop density=2500.0 bulk=1E8 shear=3E7 cohesion=140000.0 friction=45.0 &
dilation=0.0 tension=12000.0 group rock mass notnull
;
set gravity=9.81
solve elastic
save rocsl 1.sav
;... State: rocsl 2.sav ....
; Add water table
table 2 (-40.00 0) (0 0) (3.880 4.500) (11.88 12.00) (15.76 16.00) &
(23.76 24.00) (27.64 27.00) (43.15 41.00) (52.00 43.00) (80.00 45.00) &
(140.0 46.00)
def wet den
loop i (1,izones)
loop j (1,jzones)
if model(i,j)>1 then
xa=(x(i,j)+x(i+1,j)+x(i+1,j+1)+x(i,j+1))
xc=0.25*xa
ya=(y(i,j)+y(i+1,j)+y(i+1,j+1)+y(i,j+1))
FLAC Version 5.0
10 - 9
yc=0.25*ya
if yc < table(2,xc) then
density(i,j) = 2800
endif
endif
endloop
endloop
end
wet den
water density=1000.0
water table=2
solve
save rocsl 2.sav
;... State: rocksl fos.fsv ....
solve fos no restore file rocksl fos.fsv
;plot name: grid
plot hold grid water iwhite
;plot name: fail Mohr
set pltc 140000.0 pltf 45.0 pltt 12000.0
plot hold fail normal
;plot name: fos - ssi
plot hold fos tension block inv ssi iwhite bound
;plot name: fos - vel. vectors
plot hold fos group water velocity plasticity no past
FLAC Version 5.0
10 - 10
FLAC Version 5.0
Undrained Cylindrical Cavity Expansion in a Cam-Clay Medium
4-1
4 Undrained Cylindrical Cavity Expansion in a Cam-Clay Medium

The stress and pore pressure changes due to the expansion of a pressuremeter in a saturated clay
mass are analyzed using the model of a cylindrical cavity in an infinite Cam-clay medium. The
effect of the finite length of the measuring device is not considered.
In the experiment, the radius a of the cavity is expanded to up to twice its original size, a0 . The
properties of the Cam-clay material, which correspond to a Boston Blue Clay, are as follows [Carter
et al., 1979]:
undrained cohesion (Cu )
shear modulus (G)
soil constant (M)
slope of normal consolidation line ()
slope of elastic swelling line ()
reference pressure (p1 )
specific volume at reference pressure (v )
density ()
1 MPa
74 Cu
1.2
0.15
0.03
Cu
2.3
103 kg/m3
The clay is normally consolidated with in-situ stresses, r = = 1.65Cu , z = 3Cu , and
initial excess pore pressure, ue = 0. The shear modulus of the material is assumed to remain
constant during the simulation. The pressuremeter membrane is considered impermeable, and the
fluid bulk modulus is much larger than that of the soil, so that the numerical simulation can be
carried out under undrained conditions.
The problem is modeled using an axisymmetric configuration and plane-strain boundary conditions,
as represented in Figure 4.1. The FLAC model is of finite extent, but the length, L, is chosen as
very large compared to a0 .
FLAC Version 5.0
4-2
a0
Figure 4.1
Model geometry
The dimensions of the FLAC grid correspond to dimensionless values L/a0 = 100 and h/a0 = 1,
as indicated in Figure 4.2, where the FLAC system of reference axes is also represented. The grid
is composed of a single layer of 31 zones of constant height and variable zone width, graded by a
factor of 1.1.
100
Figure 4.2
Grid geometry
Initially, the cavity boundary is fixed, in-situ stresses are installed, and a pressure boundary condition
of magnitude 1.65Cu is applied at the far x-boundary. The groundwater configuration (CONFIG gw)
is selected, and the no flow (SET ow off) and large-strain (SET large) options are specified.
The preconsolidation pressure must be supplied to the numerical model. Since the soil is normally
consolidated, this value is calculated from the given initial state. The corresponding values of mean
pressure and deviator stress are p0 = 2.1Cu and q0 = 1.35Cu , and the preconsolidation pressure,
evaluated from the Cam-clay yield function (see Section 2.4.7 in Theory and Background),

pc0
= p0 [1 + (q0 /(Mp0 ))2 ]
FLAC Version 5.0
(4.1)
4-3
/p , is approxis, hence, 2.70Cu . The value of the over-consolidation ratio R, defined as R = pc0
0
imately 1.29 for this problem.
As an illustration, initial values for the specific volume, v0 , and tangent bulk modulus, K0 , are
specified. They correspond to the default values that would have been assigned by the code at the
first step command:

v0 = v ln(pc0
/p1 ) + ln(pc0
/p0 )
K0 =
v0 p0
(4.2)
(4.3)
Here, information derived from v0 and K0 is used to specify the initial porosity, water and material
bulk modulus.
The initial porosity is calculated from
n0 = 1 1/v0
(4.4)
The bulk modulus of the water is set to 100 times the initial value of the product K n. This is
sufficient to represent the water as incompressible relative to the clay. The maximum bulk modulus
of the clay is set to 10 times the initial value.
A compressive velocity of magnitude 105 a0 is applied at the cavity boundary for a total of 100,000
steps to allow doubling of the cavity radius at the end of the pressure test. Stresses and pore pressure
are monitored during the calculation.
The data file for this problem, CAV.DAT, is listed at the end of this section (Section 4.5). The Camclay parameters are calculated in the FISH functions set prop and c var. The FISH function
b table creates tables to plot results for comparison to the solution by Carter et al. (1979).
The evolution of the deviator stress, q/Cu , at the cavity wall during the expansion is plotted in
Figure 4.3. The numerical results indicate a failure level at q/Cu = 1.778. This value can
be compared to the Cam-clay analytical prediction, as follows. Under undrained conditions, the
yield path, followed by a normally consolidated stress point, has the form (see Section 15 in the
Verifications volume):
p0
=
p
M 2 + 2
M 2 + 02

(4.5)
FLAC Version 5.0
4-4
where = q/p and = ( )/. The initial value 0 = q0 /p0 can be derived from Eq. (4.1).
Using the definition of R, we obtain
02 = RM 2 M 2
(4.6)
and the stress path becomes

p0
=
p
M 2 + 2
RM 2

(4.7)
Intersection of this stress path with the critical state line q = Mp or = M gives
qcr =
Mp0

2
R
(4.8)
The prediction of qcr /Cu derived from this formula is 1.771, a value in close agreement with that
obtained numerically.
JOB TITLE : UNDRAINED CYLINDRICAL CAVITY EXPANSION IN CAM-CLAY MEDIUM
FLAC (Version 5.00)

LEGEND
1.750
12-Mar-04 12:02
step 100000
1.700
HISTORY PLOT
Y-axis :
c_q
(FISH)
X-axis :
c_a
(FISH)
1.650
1.600
1.550
1.500
1.450
1.400
11
12
13
14
15
16
17
18
19
20
(10
-01

Figure 4.3
FLAC Version 5.0
Deviator stress q/Cu at the cavity wall versus a/a0
4-5
The variation of excess pore pressure and total radial stress at the cavity wall as the cavity expands is
illustrated in Figure 4.4. These curves show a sharp rise followed by a gentle slope as pore pressure
and radial stress approach a limit value.
FLAC (Version 5.00)

LEGEND
12-Mar-04 12:02
step 100000
6.000
HISTORY PLOT
Y-axis :
Rev_Ave. SXX
( 1, 1)
5.000
Pore pressure ( 1, 1)
4.000
X-axis :
c_a
(FISH)
3.000
2.000
1.000
11
12
13
14
15
16
17
18
19
20
(10
-01

Figure 4.4
Total radial stress r /Cu and excess pore pressure ue /Cu at the
cavity wall versus a/a0
The radial distribution of effective stresses and pore pressure is plotted in Figure 4.5 when a = 2a0 .
It may be seen that the stresses remain constant in an annulus around the cavity where the soil is
at the critical state. There, the distribution of stresses has been greatly affected by the process of
cavity expansion with radial and tangential stresses now in the role of minor and major principal
stresses. The excess pore pressure develops mainly in this region. Further out, the stresses and pore
pressure are shown to evolve towards their in-situ values. These results compare well with those
presented by Carter et al. (1979).
FLAC Version 5.0
4-6
FLAC (Version 5.00)

LEGEND
3.500
12-Mar-04 12:02
step 100000
Table Plot
effective tangential stress
3.000
2.500
effective axial stress

effective radial stress
2.000
Pore Pressure
1.500
1.000
0.500
0.000
10
15
20
25
30
35
40
45
50
(10
-01

Figure 4.5
Radial distribution of effective stresses and pore pressure when

a = 2a0 plotted versus ln(r/a0 )
4.4 Reference
Carter, J. P., M. F. Randolph and C. P. Wroth. Stress and Pore Pressure Changes in Clay during and
after the Expansion of a Cylindrical Cavity, International Journal for Numerical and Analytical
Methods in Geomechanics, 3, 305-322 (1979).
FLAC Version 5.0
4-7
4.5 Data File CAV.DAT

;... State: cav.sav ....
;----------------------------------------------------------; cav.dat
; undrained cylindrical cavity expansion in Cam-Clay medium
;----------------------------------------------------------config ax gw
; --- model geometry --g 31 1
gen 1 0 1 1 200 1 200 0 rat 1.1 1
; --- model properties --mo cam-clay
prop shear 74 dens 1e-3
prop mm 1.2 lambda 0.15 kappa 0.03 mp1 1.0 mv l 2.3
; --- boundary conditions --fix y
ini sxx -1.65 syy -3 szz -1.65
apply press 1.65 i 32
fix x i 1
ini xv 1.0e-5 i 1
; model settings --set flow off
set large
; --- fish functions --; ... initial specific volume, tangent bulk modulus, porosity ...
def set prop
s1 = -sxx(1,1)
s2 = -syy(1,1)
s3 = -szz(1,1)
p0 = (s1 + s2 + s3) / 3.
q0 = sqrt(((s1-s2)*(s1-s2)+(s2-s3)*(s2-s3)+(s3-s1)*(s3-s1))*0.5)
e0 = q0 / (mm(1,1) * p0)
pc = p0 * (1. + e0 * e0)
p1 = mp1(1,1)
vc = mv l(1,1) - lambda(1,1) * ln(pc/p1)
v0 = vc - kappa(1,1) * ln(p0/pc)
b0 = v0 * p0 / kappa(1,1)
n0 = (v0 - 1.) / v0
bm = 10. * b0
k0 = 100. * b0 * n0
; K w/n large compared to K
command
prop mpc pc bulk bm por n0
water bulk k0 ten 1e10
; K w = 100*b0*n0
FLAC Version 5.0
4-8
end command
end
; ... numerical variables ...
def c var
c ep = cam p(1,1)
; effective pressure
c q = cam q(1,1)
; deviatoric stress
c cq = c ep * mm(1,1)
c tp = c ep + pp(1,1)
c v = sv(1,1)
c a = 1. + xdisp(1,1)
c b = bulk current(1,1)
c pc = mpc(1,1)
end
; --- histories --hist nstep 500
hist c var
hist sxx i 1 j 1
hist pp i 1 j 1
hist c a
hist c ep
hist c cq
hist c tp
hist c v
hist c q
hist c b
hist c pc
hist unbal
; --- test --set prop
step 100000
save cav.sav
;... State: cav1.sav ....
;
table 10 : pore pressure
;
table 11 : radial-component of effective stress (xx-stress)
;
table 12 : axial-component of effective stress (yy-stress)
;
table 13 : tangential-component of effective stress (zz-stress)
;
xval
: ln(r/a0)
def b table
loop i(1,izones)
xval = (x(i+1,1) + x(i,1)) * 0.5
xval = ln(xval)
c pp = pp(i,1)
xtable(10,i) = xval
ytable(10,i) = c pp
xtable(11,i) = xval
FLAC Version 5.0
ytable(11,i)
xtable(12,i)
ytable(12,i)
xtable(13,i)
ytable(13,i)
end loop
=
=
=
=
=
4-9
-sxx(i,1) - c pp
xval
-syy(i,1) - c pp
xval
-szz(i,1) - c pp
end
b table
save cav1.sav
;plot name: Deviator stress
;plot name: Total radial stress and excess pore pressure
plot hold history -2 line 3 line vs 4
;plot name: Distribution of eff. stresses and pore pressure
label table 10
pore pressure
label table 11
effective radial stress
label table 12
effective axial stress
label table 13
effective tangential stress
plot hold table 13 line 12 line 11 line 10 line
FLAC Version 5.0
4 - 10
FLAC Version 5.0
Analysis of a Concrete Diaphragm Wall
11 - 1
11 Analysis of a Concrete Diaphragm Wall

A concrete diaphragm wall is a continuous concrete wall built into the ground from the ground
surface. The wall may consist of precast or cast-in-place concrete panels, or contiguous bored
concrete piles. The most common type of wall is a tremie-concrete diaphragm wall cast within a
slurry stabilized trench. Trenches are 0.6 to 1.0 meters wide and are excavated in 3 to 6 meter lengths.
After the individual panels are excavated, end-stops and reinforcing are installed. Concrete is placed
and the end-stops are removed. Once the concrete has set, the neighboring panel can be excavated.
After all panels have been constructed, the major excavation can begin, with internal bracing
installed as the excavation progresses. In saturated ground, dewatering must also be performed.
The various stages of construction of a diaphragm wall can be simulated with FLAC. In this
example, we model the stages following the construction of the panels. Dewatering, excavation
and installation of struts are simulated in five excavation stages.
A cross section through the diaphragm wall, given in Figure 11.1, shows the subsurface conditions,
the depth at each of the five excavation stages, and the location of the support struts. The properties
of the soil layers are summarized in Table 11.1. A porosity of 0.3 is assumed for all soil layers.
A thick aquifer of dense gravel is located below a depth of 34 meters. The pore pressure is assumed
to be a constant value of 317 kPa at the bottom of the diaphragm wall. The wall is 1 meter thick
and is assumed to behave as an elastic material with a Youngs modulus of 19.2 GPa and Poissons
ratio of 0.2. The struts are spaced at 2 meter intervals and are pre-loaded. The excavation and
dewatering depths are listed in Table 11.2. The locations and properties of the struts are summarized
in Table 11.3. The Youngs modulus for the struts is 200 GPa.
Of interest in this analysis are the distribution of shear force and bending moment in the wall, the
axial force in the struts, and the displacements of the soil behind the wall.
FLAC Version 5.0
11 - 2
2m
Strut 1H 350 x 350
3m
Soil 1
3.5 m
5m
7.8 m
CL/ML
Strut 1H 400 x 400
8.5 m
8.3 m
11.5 m
14.75 m
Strut 1H 400 x 400
10 m
Soil 2
12 m
Strut 1H 400 x 400
SM
15 m
15.25 m
17 m
Soil 3
18 m
CL/ML
20 m
17.8 m
25 m
20.5 m
Soil 4
SM
30 m
30 m
Soil 5
ML
33.5 m
34 m
Soil 6
GW
Support
Subsurface Condition
Figure 11.1 Cross section through diaphragm wall
Table 11.1 Soil properties
Soil 1
Soil 2
Soil 3
Soil 4
Soil 5
Soil 6
Dry Density
(kg/m3 )
Bulk Modulus
(MPa)
Shear Modulus
(MPa)
Friction Angle
(degrees)
1880
1980
1910
1970
1980
1700
16.67
30.30
33.33
41.67
77.78
78.43
10.17
25.60
20.25
32.50
47.25
53.20
30
32
30
34
32
38
Table 11.2 Excavation and dewatering depths

Excavation Depth (m)
4.0
8.0
12.0
15.0
18.0
FLAC Version 5.0
Dewatering Depth (m)

5.0
9.0
13.0
16.0
19.0
11 - 3
Table 11.3 Strut properties

Strut
Number
Depth
(m)
Area
(m2 )
Moment of
Inertia (m4 )
Pre-Load
(kN)
1
2
3
4
3
7
11
14
173.9 104
218.7 104
218.7 104
218.7 104
4.03 104
6.60 104
6.60 104
6.60 104
200.0
400.0
450.0
400.0

The numerical analysis for this problem provides a solution at each of the five excavation depths.
At each depth, the modeling sequence includes dewatering, excavation and strut installation steps.
A vertical line of symmetry is taken through the center of the excavation, and only one wall is
modeled. The data files for this problem are listed in Sections 11.5 through 11.8.
The groundwater flow option, adjust total stress option and structural elements are activated from
the Model Options dialog for this problem, as shown in Figure 11.2. The automatic adjustment of
total stresses is selected to facilitate the simulation of dewatering.
Figure 11.2 Settings in Model Options dialog for concrete diaphragm wall
example
FLAC Version 5.0
11 - 4
The grid for this analysis is created using the Build / Block tool. The model dimensions and zoning
selected for the grid are assigned in the Block dialog, as shown in Figure 11.3. The grid is shown
in Figure 11.4. The right boundary of the grid is the line of symmetry for the excavation. The
left boundary is located approximately 15 times the excavation width away from the excavation, in
order to minimize boundary effects.
Figure 11.3
Build
/ Block dialog
JOB TITLE : Analysis of a Concrete Diaphragm Wall
(*10^2)
FLAC (Version 5.00)

0.400
LEGEND
25-Sep-03 16:36
step
0
-1.580E+02 <x< 1.695E+01
-1.175E+02 <y< 5.745E+01
0.200
0.000
Grid plot
0
5E 1
-0.200
Beam plot
-0.400
-0.600
-0.800
-1.000

-1.400
-1.200
-1.000
-0.800
(*10^2)
-0.600
-0.400
-0.200
0.000
Figure 11.4 FLAC grid for the analysis of a concrete diaphragm wall
The diaphragm wall is also in place in the model before the initial equilibrium calculation is
performed, as shown in Figure 11.4. The wall is modeled by 28 beam elements. The upper 18
elements are each 1 meter in height and are positioned so that beam nodes coincide with the depths
of the struts. (See Figure 11.5). The structural nodes with numbers 27, 23, 19 and 16 will be
FLAC Version 5.0
11 - 5
attached to the struts during the excavation stages. The beam is connected to the soil grid via
interface elements attached on both sides of the beam elements. The bottom node (node 1) is
attached directly to a gridpoint; this provides a representation for end-bearing loading of the wall.
Note that structural element logic is a plane-stress formulation, so the value specified for the Youngs
modulus, E, is divided by (1 2 ) to correspond to the plane-strain model (see Section 1.2.2 in
Structural Elements).
Figure 11.5 Structural beam nodes in upper part of wall

The soil/wall interface properties selected for this example are for demonstration purposes: actual
values for wall friction and adhesion should be determined from physical testing or from the literature
(e.g., see Clayton et al., 1993). The interface nodes are assigned low shear strength (approximately
2/3 of the soil friction angle) to simulate a relatively smooth concrete-to-soil interface. (The
influence of the interface resistance can be investigated easily by varying the properties of the
interface nodes.) Interface stiffness values were selected to approximate the results for the case that
the wall is rigidly attached to the grid.
The analysis begins at an initial equilibrium state prior to excavation. The initial stress state is
found for the given soil conditions, assuming that the ratio of effective horizontal stress to effective
vertical stress is 0.5. A series of FISH functions contained in ININV.FIS (see the FISH Library in
Section 3 in the FISH volume) are used to establish the initial stress state and pore pressures for the
horizontally layered media, with the groundwater table located 2 meters below the ground surface.
This analysis is performed as a coupled groundwater flow and mechanical calculation (CONFIG gw)
in order for ININV.FIS to be applied. However, the groundwater flow is inhibited (SET ow off)
and the water bulk modulus is set to zero so that the specified pore pressures are maintained. The
recommended procedure to initialize a stress state in a medium with a phreatic surface is discussed
in Section 3.4.8 in the Users Guide.
FLAC Version 5.0
11 - 6
Although stresses and pore pressures are initialized in the grid, some stepping is required to bring the
model to equilibrium. This is because the additional weight of the beam elements representing the
concrete wall produces an imbalance that necessitates some stepping to equilibrate the model again.
Note that the grid is pulled down slightly as the wall settles due to gravity. This problem is run in
large-strain mode. The model at the initial equilibrium stage is saved to the file DW EQUIL.SAV.
The structural forces in the wall arising from the presence of groundwater will depend on the details
of the connection, the fluid boundary conditions, and when the support is installed. Section 1.9.5
in Fluid-Mechanical Interaction discusses the various possibilities. It is important to recognize
the conditions that the model is to represent before setting up model conditions. For this example,
we assume the following conditions apply:
1. The structure exists independently of the grid. Forces are transmitted through
the interface elements.
2. The wall provides an impermeable barrier to the groundwater, which acts
directly on the surface of the wall.
3. The wall is dewatered instantaneously; no movements or transient flow effects
are considered. (See Section 1.8.2 in Fluid-Mechanical Interaction for a
recommended procedure to determine if this assumption is appropriate.)
The pore pressure distribution in the model is changed at each stage to represent dewatering.
The distribution approximates the steady-state flow condition for that excavation stage. (It is not
necessary to perform a flow calculation for this analysis, because of the constant pore pressure
at the base of the wall imposed by the underlying aquifer.) Note that stresses are also changed
automatically by specifying CONFIG ats because total stresses change if pore pressure is changed
by some external method. (See Section 1.9.7 in Fluid-Mechanical Interaction.) The effect of
dewatering for the first excavation stage is illustrated in Figure 11.6. This plot shows the settlement
within the trench region, as well as the moments in the wall, induced by the dewatering.
The struts are modeled with beam elements. One long element segment represents each strut. The
Youngs modulus of the strut is automatically scaled by the 2 meter strut spacing when the spacing
property is given. The pre-load applied with the STRUCT node n load command is scaled manually
by dividing the actual pre-load value by the 2 meter spacing.
The struts should be installed so that moments cannot develop at their connection with the wall.
This is accomplished by slaving the strut node to the wall node using the STRUCTURE node n slave
x y m command, in which n corresponds to the strut node and m to the wall node. The strut node
is slaved in the x- and y-direction to the wall node, but is free to rotate. In this way, the wall and
strut can move without moments developing between the strut and the wall.
FLAC Version 5.0
11 - 7
JOB TITLE : .
(*10^1)
0.500
FLAC (Version 5.00)

LEGEND
14-Jan-05 12:04
step
5006
-3.543E+01 <x< 1.465E+01
-4.280E+01 <y< 7.276E+00
-0.500
Y-displacement contours
-7.50E-04
-5.00E-04
-2.50E-04
2.50E-04
5.00E-04
C
-1.500

B: 5.000E+04
I: 4.000E+05
Beam Plot
E
-2.500
Moment
on
Structure
Max. Value
# 1 (Beam )
2.364E+05
-3.500

I
-3.000
-2.000
-1.000
(*10^1)
0.000
1.000
Figure 11.6 Vertical displacement contours in grid and moment distribution

in diaphragm wall induced by first dewatering step
A pre-load is applied in the x-direction to the strut node located at the centerline of the excavation.
This node is also prevented from rotating or translating in the y-direction. Although the fixity
condition in the y-direction is not required for this problem, it may become necessary to minimize
any effect of drift for excavations requiring longer struts. The fixity condition does not affect the
solution, provided moments in the struts can be neglected. (Check moments in the struts to ensure
that this is the case.) The pre-load is applied in two steps. First, the force is applied to the strut
node with the STRUCT node n load fx fy m command, and the model is stepped to an equilibrium
state. After the equilibrium state is achieved with the pre-load, the pre-load force is removed, the
x-velocity of the node is set to zero, and the strut node is fixed so that the load in the strut can
change during subsequent excavation stages.
This problem is run for two cases: (1) pre-loading in the struts; and (2) no pre-loading. This will
allow us to assess the effect of pre-loading on the displacement of the soil and wall and on the loads
in the wall and struts.
FLAC Version 5.0
11 - 8
11.3 Results
The deflections of the diaphragm wall at each excavation stage are indicated by the plot of xdisplacement of the wall structure versus wall depth in Figure 11.7 for the pre-loaded struts, and
in Figure 11.8 for no pre-loading. These plots are table plots generated using the FISH function
WALL DISP.FIS listed in Section 11.6. The x-displacement and the y-position of each node
along the wall are stored in five tables corresponding to each excavation stage. The maximum
deformation is approximately 62 mm at 20 m depth for pre-loaded struts (shown in Figure 11.7)
and 70 mm for no pre-load (shown in Figure 11.8), at the final stage.
The actual axial loads in the struts are calculated by the FISH functions in STRUT AX LOAD.FIS
listed in Section 11.8, after the model has come to equilibrium for each excavation stage. Note
that the axial loads accessed by FISH are scaled values and must be multiplied by the spacing to
determine the actual values. The actual axial load values are then stored in tables for comparison
at the end of the calculation. The results are shown in Figure 11.9 for pre-loaded struts, and in
Figure 11.10 for no pre-load.
As shown in Figure 11.9 for the pre-load case, the axial load in strut 1 (stored in Table 1) is initially
at a pre-load value of 200 kN for the first stage, increases at the second stage and then decreases for
the later stages. The axial load in strut 2 (Table 2) increases until the third stage and then decreases.
The load in strut 3 (Table 3) increases until the fourth stage, while strut 4 (Table 4) increases through
the fifth stage.
The results are similar for the no pre-load case, as shown in Figure 11.10.
The change in axial load is related to the movement of the wall during excavation indicated
by the plot in Figure 11.11 for pre-loaded struts, and in Figure 11.12 for no pre-loading. These
figures show the evolution of the horizontal displacement at five elevations on the wall (at the top
of the wall, and at the location of each strut). The top of the wall is shown to move away from
the excavation (i.e., in the negative x-direction) after an initial inward movement, while most of
the wall moves into the excavation (i.e., in the positive x-direction). The movements increase with
depth, and coincide with the increase in axial forces for the struts.
Figures 11.13 and 11.14 plot the x-displacement contours and axial forces in the struts after the
final excavation stage, for pre-loading and no pre-loading, respectively. Again, as shown in these
figures, the movements correspond to the increase in loads in the struts. Note that actual values for
the axial forces are plotted directly for these plots.
The moment distribution and shear forces in the wall, for the analysis with pre-loaded struts, are
shown in Figures 11.15 and 11.17, and for the no pre-loading case in Figures 11.16 and 11.18. In
both cases, a large bending moment is shown to develop in the wall at the bottom of the excavation.
FLAC Version 5.0
11 - 9
JOB TITLE : .
FLAC (Version 5.00)

(10
01
LEGEND
14-Jan-05 12:19
step 41545
0.000
-0.500
Table Plot
Stage 1
Stage 2
-1.000
Stage 3
Stage 4
-1.500
Stage 5
-2.000
-2.500
-3.000
10
20
30
40
50
60
(10
-03

Figure 11.7 x-displacement of diaphragm wall at the end of each excavation

stage for pre-loaded struts
JOB TITLE : .
FLAC (Version 5.00)

(10
01
LEGEND
14-Jan-05 12:30
step 36654
Table Plot
Stage 1
Stage 2
Stage 3
0.000
-0.500
-1.000
Stage 4
Stage 5
-1.500
-2.000
-2.500
-3.000
10
20
30
40
50
60
(10
-03

Figure 11.8 x-displacement of diaphragm wall at the end of each excavation

stage for no pre-loading in struts
FLAC Version 5.0
11 - 10
JOB TITLE : .
FLAC (Version 5.00)

(10
06
LEGEND
14-Jan-05 12:19
step 41545
Table Plot
Strut 1
3.000
2.500
Strut 2
Strut 3
2.000
Strut 4
1.500
1.000
0.500
10
15
20
25
30
35
40
45
50
(10
-01

Figure 11.9 Actual axial forces in struts at the end of each excavation stage
for pre-loaded struts
JOB TITLE : .
FLAC (Version 5.00)

(10
06
LEGEND
14-Jan-05 12:30
step 36654
Table Plot
Strut 1
Strut 2
Strut 3
3.000
2.500
2.000
Strut 4
1.500
1.000
0.500
0.000
10
15
20
25
30
35
40
45
50
(10
-01

Figure 11.10 Actual axial forces in struts at the end of each excavation stage
for no pre-loading in struts
FLAC Version 5.0
11 - 11
JOB TITLE : .
FLAC (Version 5.00)

(10
-02
LEGEND
14-Jan-05 12:19
step 41545
HISTORY PLOT
Y-axis :
3.000
2.500

2.000

1.500
X-axis :
Number of steps
1.000
0.500
10
15
20
25
30
35
40
(10
03

Figure 11.11 x-displacement histories at five elevations along the wall for preloaded struts
JOB TITLE : .
FLAC (Version 5.00)

(10
-02
LEGEND
14-Jan-05 12:30
step 36654
HISTORY PLOT
Y-axis :
4.000
3.500
3.000

2.500

X-axis :
Number of steps
2.000
1.500
1.000
0.500
10
15
20
25
30
35
(10
03

Figure 11.12 x-displacement histories at five elevations along the wall for no
pre-loading in struts
FLAC Version 5.0
11 - 12
JOB TITLE : .
(*10^1)
0.750
FLAC (Version 5.00)

LEGEND
0.250
14-Jan-05 12:19
step 41545
-2.769E+01 <x< 1.787E+01
-3.677E+01 <y< 8.791E+00
-0.250
-0.750
Boundary plot
0
1E 1
-1.250
Beam Plot
Axial Force on
Structure
Max. Value
# 2 (Beam )
1.200E+05
# 3 (Beam )
9.227E+05
# 4 (Beam )
1.479E+06
# 5 (Beam )
3.344E+06
X-displacement contours
0.00E+00
2.00E-02
4.00E-02
6.00E-02
-1.750
-2.250
-2.750

Beam plot
-3.250

-2.500
-2.000
-1.500
-1.000
-0.500
(*10^1)
0.000
0.500
1.000
1.500
Figure 11.13 x-displacement contours in the grid and axial forces in the struts
at the final excavation stage for pre-loaded struts
JOB TITLE : .
(*10^1)
0.750
FLAC (Version 5.00)

LEGEND
0.250
14-Jan-05 12:30
step 36654
-2.769E+01 <x< 1.787E+01
-3.677E+01 <y< 8.791E+00
-0.250
-0.750
Boundary plot
0
1E 1
-1.250
Beam Plot
Axial Force on
Structure
Max. Value
# 2 (Beam )
8.761E+04
# 3 (Beam )
1.019E+06
# 4 (Beam )
1.598E+06
# 5 (Beam )
3.147E+06
0.00E+00
2.00E-02
4.00E-02
6.00E-02
-1.750
-2.250
-2.750

Beam plot
-3.250

-2.500
-2.000
-1.500
-1.000
-0.500
(*10^1)
0.000
0.500
1.000
1.500
Figure 11.14 x-displacement contours in the grid and axial forces in the struts
at the final excavation stage for no pre-loading in struts
FLAC Version 5.0
11 - 13
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)
1.000
LEGEND
14-Jan-05 12:19
step 41545
-3.075E+01 <x< 2.908E+01
-4.570E+01 <y< 1.413E+01
0.000
Beam plot
Boundary plot
0
-1.000
1E 1
Beam Plot
Moment
on
Structure
Max. Value
# 1 (Beam ) -2.420E+06
-2.000
-3.000
-4.000

-2.500
-1.500
-0.500
0.500
1.500
2.500
(*10^1)
Figure 11.15 Moments in the diaphragm wall at the final excavation stage (with
pre-loading)
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)
1.000
LEGEND
14-Jan-05 12:30
step 36654
-3.075E+01 <x< 2.908E+01
-4.570E+01 <y< 1.413E+01
0.000
Beam plot
Boundary plot
0
-1.000
1E 1
Beam Plot
Moment
on
Structure
Max. Value
# 1 (Beam ) -2.500E+06
-2.000
-3.000
-4.000

-2.500
-1.500
-0.500
0.500
1.500
2.500
(*10^1)
Figure 11.16 Moments in the diaphragm wall at the final excavation stage (no
pre-loading)
FLAC Version 5.0
11 - 14
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
0.000
14-Jan-05 12:19
step 41545
-2.569E+01 <x< 3.175E+01
-4.682E+01 <y< 1.062E+01
-1.000
Beam Plot
Shear Force on
Structure
Max. Value
# 1 (Beam ) -1.168E+06
Beam plot
Boundary plot
0
-2.000
1E 1
-3.000
-4.000

-2.000
-1.000
0.000
(*10^1)
1.000
2.000
3.000
Figure 11.17 Shear forces in the diaphragm wall at the final excavation stage
(with pre-loading)
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
0.000
14-Jan-05 12:30
step 36654
-2.569E+01 <x< 3.175E+01
-4.682E+01 <y< 1.062E+01
-1.000
Beam Plot
Shear Force on
Structure
Max. Value
# 1 (Beam ) -1.142E+06
Beam plot
Boundary plot
0
-2.000
1E 1
-3.000
-4.000

-2.000
-1.000
0.000
(*10^1)
1.000
2.000
3.000
Figure 11.18 Shear forces in the diaphragm wall at the final excavation stage
(no pre-loading)
FLAC Version 5.0
11 - 15
11.4 Reference
Clayton, C. R. I., J. Milititsky and R. T. Woods. Earth Pressure and Earth-Retaining Structures.
London: Blackie Academic & Professional, 1993.
FLAC Version 5.0
11 - 16
11.5 Data File DIAP.DAT

;Title:Concrete diaphragm wall
;... State: dw ini.sav ....
config gwflow ats
grid 37,35
gen (-150.0,-60.0) (-150.0,-18.0) (-4.0,-18.0) (-4.0,-60.0) ratio 0.9,0.9 &
i 1 24 j 1 18
gen (-150.0,-18.0) (-150.0,0.0) (-4.0,0.0) (-4.0,-18.0) ratio 0.9,1.0 &
i 1 24 j 18 36
gen (-4.0,-60.0) (-4.0,-18.0) (10.0,-18.0) (10.0,-60.0) ratio 1.0,0.9 &
i 24 38 j 1 18
gen (-4.0,-18.0) (-4.0,0.0) (10.0,0.0) (10.0,-18.0) i 24 38 j 18 36
group Soil 6 j 1 7
model mohr group Soil 6
prop density=1700.0 bulk=7.843E7 shear=5.32E7 cohesion=0.0 friction=38.0 &
dilation=0.0 tension=0.0 group Soil 6
group Soil 5 j 8
friction=32.0 dilation=0.0 tension=0.0 group Soil 5
group Soil 4 j 9 14
group Soil 3 j 15 18
prop density=1980.0 bulk=3.03E7 shear=2.56E7 cohesion=0.0 friction=32.0 &
dilation=0.0 tension=0.0 group Soil 2
model null i 28
group null i 28
group delete null
prop por=0.3 notnull
struct node 2 0.0,-18.0
FLAC Version 5.0
11 - 17
struct node 3 0.0,0.0

struct beam begin node 1 end node 2 seg 10 prop 1001
struct prop 1001
struct prop 2001
struct prop 3001
struct prop 4001
struct prop 5001
struct prop 6001
struct prop 1001 density 2000.0 spacing 1.0 e 2.0E10 height 1.0 width 1.0
interface 1 aside from 28,8 to 28,36 bside from node 1,4 to node 3
interface 1 unglued kn=1.0E8 ks=1.0E8 cohesion=0.0 dilation=0.0 &
friction=20.0 tbond=0.0 bslip=Off
ini x add -1.0 y add 0.0 nmregion 29 1
attach aside from 28,1 to 28,7 bside from 29,1 to 29,7
interface 3 aside from 28,7 to 28,8 bside from 29,7 to 29,8
interface 3 glued kn=1.0E8 ks=1.0E8
fix y j 1
fix x i 38
fix x i 1
set gravity=10.0
set flow=off
set echo off
call Ininv.fis
set wth=-2.0 k0x=0.5 k0z=0.5
ininv
save dw ini.sav
;... State: dw equil.sav ....
history 1 syy i=37, j=28
history 2 esyy i=37, j=28
history 3 pp i=37, j=28
set large
solve elastic
save dw equil.sav
;... State: dw w1.sav ....
initial pp 0.0 i 29 38 j 32 36
FLAC Version 5.0
11 - 18
initial pp 317000.0 var 0.0,-317000.0 i 29 38 j 8 32

scline reset
scline 1 (4.5,-60.0) (4.5,0.0)
solve
save dw w1.sav
;... State: dw e1.sav ....
history reset
group delete null
solve
save dw e1.sav
;*** Branch: Preload ****
;... State: dw s1.sav ....
struct node 31 0.0,-3.0 pin slave x y 27
struct beam begin node 30 end node 31 prop 1002
struct prop 1002
struct node 30 fix y r load -1e5,0.0 0.0
struct prop 1002 spacing 2.0 e 1.99999996E11 area 0.01739 I 4.03E-4
solve
struct node 30 fix x initial xvel=0.0 load 0.0,0.0 0.0
solve
save dw s1.sav
;... State: dw r1.sav ....
set echo off
call wall disp.fis
set nodes in wall=29 n table=1
wall disp
set echo off
call init strut table.fis
init strut table
set echo off
call strut ax load.fis
set nstage=1 nstrut=11 el num=29
strut ax load
save dw r1.sav
FLAC Version 5.0
11 - 19

initial pp 317000.0 var 0.0,-317000.0 i 29 38 j 8 28
initial pp 0.0 i 29 38 j 28 32
solve
save dw w2.sav
group delete null
solve
save dw e2.sav
struct prop 1003
struct node 32 fix y r load -200000.0,0.0 0.0
struct prop 1001 height 0.0 width 0.0
struct prop 1003 spacing 2.0 e 2e11 area 0.02187 I 6.6E-4
solve
struct node 32 fix x initial xvel 0.0 load 0.0,0.0 0.0
solve
save dw s2.sav
set echo off
call wall disp.fis
wall disp
set echo off
strut ax load
set echo off
strut ax load
save dw r2.sav
initial pp 317000.0 var 0.0,-317000.0 i 29 38 j 8 24
FLAC Version 5.0
11 - 20
initial pp 0.0 i 29 38 j 24 28
solve
save dw w3.sav
group delete null
solve
save dw e3.sav
solve
struct node 34 fix x initial xvel 0.0 load 0.0,0.0 0.0
solve
save dw s3.sav
set echo off
call wall disp.fis
wall disp
set echo off
strut ax load
set echo off
strut ax load
set echo off
strut ax load
save dw r3.sav
initial pp 317000.0 var 0.0,-317000.0 i 29 38 j 8 21
initial pp 0.0 i 29 38 j 21 24
solve
save dw w4.sav
FLAC Version 5.0
11 - 21

group delete null
solve
save dw e4.sav
solve
struct node 36 fix x initial xvelocity 0.0 load 0.0,0.0 0.0
solve
save dw s4.sav
set echo off
call wall disp.fis
wall disp
set echo off
strut ax load
set echo off
strut ax load
set echo off
strut ax load
set echo off
strut ax load
save dw r4.sav
initial pp 317000.0 var 0.0,-317000.0 i 29 38 j 8 18
initial pp 0.0 i 29 38 j 18 21
solve
FLAC Version 5.0
11 - 22
save dw w5.sav
group delete null
solve
save dw e5.sav
set echo off
call wall disp.fis
wall disp
set echo off
strut ax load
set echo off
strut ax load
set echo off
strut ax load
set echo off
strut ax load
save dw r5.sav
;*** Branch: NoPreload ****
restore dw e1.sav
;... State:
struct node
struct node
struct beam
struct prop
struct prop
struct node
solve
save noload
noload dw s1.sav ....

30 9.0,-3.0
31 0.0,-3.0 pin slave x y 27
begin node 30 end node 31 prop 1002
1002
1002 spacing 2.0 e 1.99999996E11 area 0.01739 I 4.03E-4
30 fix x y r
dw s1.sav
;... State: noload dw r1.sav ....
FLAC Version 5.0
11 - 23
set echo off

call wall disp.fis
wall disp
set echo off
call init strut table.fis
init strut table
set echo off
strut ax load
save noload dw r1.sav
;... State: noload dw w2.sav ....
initial pp 317000.0 var 0.0,-317000.0 i 29 38 j 8 28
initial pp 0.0 i 29 38 j 28 32
solve
save noload dw w2.sav
;... State: noload dw e2.sav ....
group delete null
solve
save noload dw e2.sav
;... State:
struct node
struct node
struct beam
struct prop
struct prop
struct prop
struct node
solve
save noload

32 9.0,-7.0
33 0.0,-7.0 pin slave x y 23
1003
1001 height 0.0 width 0.0
1003 spacing 2.0 e 2e11 area 0.02187 I 6.6E-4
32 fix x y r
dw s2.sav

set echo off
call wall disp.fis
wall disp
set echo off
FLAC Version 5.0
11 - 24
strut ax load
set echo off
strut ax load
initial pp 317000.0 var 0.0,-317000.0 i 29 38 j 8 24
initial pp 0.0 i 29 38 j 24 28
solve
group delete null
solve
;... State:
struct node
struct node
struct beam
struct node
solve
save noload

34 9.0,-11.0
35 0.0,-11.0 pin slave x y 19
34 fix x y r
dw s3.sav

set echo off
call wall disp.fis
wall disp
set echo off
strut ax load
set echo off
strut ax load
set echo off
strut ax load
FLAC Version 5.0
11 - 25

initial pp 317000.0 var 0.0,-317000.0 i 29 38 j 8 21
initial pp 0.0 i 29 38 j 21 24
solve
group delete null
solve
;... State:
struct node
struct node
struct beam
struct node
solve
save noload

36 9.0,-14.0
37 0.0,-14.0 pin slave x y 16
36 fix x y r
dw s4.sav

set echo off
call wall disp.fis
wall disp
set echo off
strut ax load
set echo off
strut ax load
set echo off
strut ax load
set echo off
strut ax load
FLAC Version 5.0
11 - 26

initial pp 317000.0 var 0.0,-317000.0 i 29 38 j 8 18
initial pp 0.0 i 29 38 j 18 21
solve
group delete null
solve
set echo off
call wall disp.fis
wall disp
set echo off
strut ax load
set echo off
strut ax load
set echo off
strut ax load
set echo off
strut ax load
;plot name: ydisp-pp-moments
plot hold ydisp fill int 2.5E-4 zero inv pp int 50000.0 white struct beam &
moment fill white
;plot name: xdisp histories
plot hold history 1 line 2 line 3 4 5
;plot name: wall moments
plot hold beam bound struct beam moment 1 fill max 5000000.0
FLAC Version 5.0
11 - 27
;plot name: strut loads

plot hold bound struct beam axial 2 5 fill max 8000000.0 lcyan xdisp fill &
int 0.02 beam lmagenta
;plot name: wall shear
plot hold struct beam shear 1 fill max 2000000.0 lcyan beam bound
;plot name: grid
plot hold grid beam lmagenta
;plot name: Strut load vs. stage
label table 11
Strut 1
label table 12
Strut 2
label table 13
Strut 3
label table 14
Strut 4
plot hold table 14 line 13 line 12 line 11 line
;plot name: x-disp along wall
label table 1
Stage 1
label table 2
Stage 2
label table 3
Stage 3
label table 4
Stage 4
label table 5
Stage 5
plot hold table 5 line 4 line 3 line 2 line 1 line
FLAC Version 5.0
11 - 28
11.6 Data File WALL DISP.FIS

;Name:wall disp
;Diagram:
;Input:nodes in wall/int/29/number of structural nodes along wall
;Input:n table/int/10/table number to store vertical displacement
call str.fin
def wall disp
ip = imem(str pnt+$ksnode)
mt = 0
loop while ip # 0
id node = imem(ip+$kndid)
if id node <= nodes in wall then
mt = mt + 1
xn disp = fmem(ip + $kndu1)
yn pos = fmem(ip + $kndy)
table(n table,yn pos) = xn disp
endif
ip = imem(ip)
endloop
loop m (1,mt)
x value = ytable(n table,m)
y value = xtable(n table,m)
xtable(n table,m) = x value
ytable(n table,m) = y value
endloop
end
FLAC Version 5.0
11 - 29
11.7 Data File INIT STRUT TABLE.FIS

;Name:init strut table
;Diagram:
def init strut table
nstage = 0
nstrut = 11
table(nstrut,nstage) =
nstrut = 12
nstrut = 13
nstrut = 14
;
nstage = 1
nstrut = 12
nstrut = 13
nstrut = 14
;
nstage = 2
nstrut = 13
nstrut = 14
;
nstage = 3
nstrut = 14
end
0
0
0
0
0
0
0
0
0
FLAC Version 5.0
11 - 30
11.8 Data File STRUT AX LOAD.FIS

;Name:strut ax load
;Diagram:
;Input:nstage/int/1/excavation stage number
;Input:nstrut/int/11/strut number
;Input:el num/int/29/beam element ID number
call str.fin
def get el num
ip = imem(str pnt+$ksels)
loop while ip # 0
id num = imem(ip + $kelid)
if id num = el num then
el addr = ip
exit
endif
ip = imem(ip)
endloop
end
def get ax load
ax load = 2.0 * fmem(el addr+$kelfax)
end
def store ax load
mstrut = nstrut
mstage = nstage
command
table mstrut insert mstage,ax load
endcommand
end
def strut ax load
get el num
get ax load
store ax load
end
FLAC Version 5.0
Multi-Stage Tunnel Excavation and Support
12 - 1
12 Multi-Stage Tunnel Excavation and Support

Construction of large railroad, subway and road tunnels often involves multiple stages of excavation
and support, particularly if the tunnels are located at shallow depth and/or in weak ground. A typical
example for the construction sequence of a shallow tunnel is illustrated in Figure 12.1.
In this example, the construction sequence is divided into three major excavation stages:
Stage I: side excavation;
Stage II: top-heading excavation;
Stage III: bench excavation.
Each excavation stage is accomplished in three construction steps:
Step a: initial excavation;
Step b: installation of rockbolt support;
Step c: installation of a shotcrete lining.
The three steps occur at different times during the advancement of the tunnel face. Consequently,
the loads acting on the tunnel will be changed at the time the support is installed, as a function of
the tunnel advancement.
The stress and displacement fields in the vicinity of a tunnel construction change in the direction of
the advancing tunnel face, and this is most rigorously analyzed using a three-dimensional program,
such as FLAC 3D (Itasca 2002). However, advancing tunnel problems are often analyzed in two
dimensions by neglecting displacements normal to the tunnel cross-section.
An important issue in the design of supports is the amount of change in the tunnel load that takes
place, due to the tunnel advancement, before the support is installed. If no change is assumed
to occur, the loads acting on the support will be over-predicted. If complete relaxation at the
tunnel periphery is assumed to occur, zero load will develop in the support at the installation step,
provided that the relaxation state is at equilibrium. In reality, some relaxation takes place. However,
it is difficult to quantify relaxation with a two-dimensional program, because this depends on the
distance behind the face at which the support is installed. One way to model the relaxation is to
decrease the elastic moduli of the tunnel core, equilibrate, install the support and remove the core.
This approach is typical of finite element codes. The main problem then becomes estimating how
much to reduce the moduli.
FLAC Version 5.0
12 - 2
24. 3 m
Rock Properties
Unit weight
Bulk Modulus
Shear Modulus
Cohesion
Internal
Friction Angle
Tensile
Strength
:
:
:
:
2700 kg/m
0.555 GPa
0.417 GPa
10 kPa
: 33
R=
17.0
: 1 kPa
m
6.5
: 0.5
2
: 9.81 m/sec
Shotcrete
Youngs Modulus : 5.56 GPa
10.35 m
R=
R=
=9
.63
Initial Conditions
Ko
Gravity
3. 5
m
Rock Bolt
Diameter
: 25 mm
Youngs Modulus : 205 GPa
8.12 m
CONSTRUCTION SEQUENCE
Initial
Ko State
Side Excavation (I)
Percent
Relaxation
50%
Step
Ia
0
CL
75%
CL
100%
Ib
CL
50%
Ic
CL
Bench Excavation (III)
Top Heading Excavation (II)

75%
IIa
CL
100%
IIb
CL
50%
IIc
CL
75%
IIIa
CL
100%
IIIb
CL
IIIc
CL
Construction
State
Figure 12.1 Construction conditions and sequence for a multi-stage tunnel

excavation and support
FLAC Version 5.0
12 - 3
An alternative approach to model the relaxation is based on the relation of the closure of the
unsupported tunnel to the distance to the face. Panet (1979) published such an expression. (Also
see Section 8.) Tunnel closure can also be related to traction forces acting on the tunnel periphery
via a ground reaction curve. Thus, the tunnel relaxation as a function of the distance to the face
can be specified in terms of tractions defined by a ground reaction curve and an expression relating
closure to distance to the face.
In order to simulate the relaxation, tractions are first applied to the tunnel boundary to provide
an equilibrium condition at zero relaxation; then the tractions are gradually decreased to a value
corresponding to a tunnel closure value that is related to a specified distance to the face. The
support is then installed at this relaxation state. In this example, the rockbolt support is installed at
an excavation stage corresponding to 50% relaxation of the tunnel load, and the shotcrete is installed
at a stage corresponding to 75% relaxation, as illustrated in Figure 12.1.
12.2.1 Model Setup
FLAC is well-suited to model sequential excavation and construction problems. In this example,
the three excavation stages and three construction steps within each stage are simulated as nine
sequential solutions. The data file for this analysis begins at Branch: Multi-Stage Tunnel in
MSTUNNEL.DAT, listed in Section 12.5. The data file at Branch: Ground Reaction Curve in
MSTUNNEL.DAT is included to demonstrate the process to develop a ground reaction curve for
this model. In addition, several FISH functions, listed in Section 12.6, are provided to control the
tunnel load relaxation process.
The FLAC mesh is defined with the grid distorted to align with the boundaries of the three segments
of the tunnel excavation. The Build / Radial tool is used to create a fine mesh in the vicinity of the
tunnel, and a radially graded mesh extending to the model boundaries. The model dimensions and
zoning selected for the grid are shown in the Edit Radial Grid dialog, in Figure 12.2. The Arc mode
in the Alter / Shape tool is then used to create the tunnel periphery based on the arc radii shown in
Figure 12.1, and the Alter / Mark tool is used to mark the boundaries between the three excavation
segments.
The initial grid and marked gridpoints indicating the tunnel segment boundaries are shown in
Figure 12.3. Note that because of the symmetry in the tunnel conditions, only half of the tunnel
is modeled. The symmetry condition implies that the left- and right-side drifts are excavated
simultaneously. If the effect of excavation sequence is considered an issue, then it may be necessary
to model the entire tunnel.
FLAC Version 5.0
12 - 4
Figure 12.2
Edit Radial Grid
dialog
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

0.500
LEGEND
8-Dec-03 13:16
step
0
-1.891E+01 <x< 6.891E+01
-7.391E+01 <y< 1.391E+01
-0.500
-1.500
Grid plot
0
2E 1
-2.500
Marked Gridpoints
-3.500
-4.500
-5.500
-6.500

-1.000
0.000
1.000
2.000
(*10^1)
3.000
4.000
5.000
Figure 12.3 FLAC grid for a multi-stage tunnel construction
FLAC Version 5.0
6.000
12 - 5
The rock behavior is represented by the Mohr-Coulomb model assigned the properties listed in
Figure 12.1. The rockbolts are modeled using rockbolt elements, and the shotcrete is simulated
with elastic liner elements. Note that structural element logic is a plane-stress formulation, so the
value specified for the Youngs modulus, E, is divided by (1 2 ) to correspond to the plane-strain
model (see Section 1.2.2 in Structural Elements).
The model is brought to an initial force-equilibrium state under gravitational loading, with the top
boundary of the mesh representing the ground surface. This initial stage is identified as Step 0 in
Figure 12.1.
12.2.2 Ground Reaction Curve
Before conducting the sequential excavation/support analysis, unsupported-tunnel calculations are
performed, in order to develop ground reaction curves for this model. This procedure is demonstrated for the excavation of the entire tunnel in one stage. Separate ground reaction curves can also
be developed for each tunnel segment. Branch: Ground Reaction Curve, in MSTUNNEL.DAT,
lists the data file for this analysis.
The ground reaction curve is developed by measuring the force on the tunnel boundary at zero
relaxation, and applying an incrementally decreasing amount of this force as a traction while
measuring the corresponding tunnel closure.
The FISH function apply rf, in RELAX TRACTIONS.FIS (Section 12.6.1), is used to apply
tractions along the tunnel boundary. This function first recovers the x- and y-reaction forces at
selected tunnel-boundary gridpoints, and then assigns these forces as tractions (with an opposite
sign) at the same boundary gridpoints, using the APPLY xforce and APPLY yforce commands. The
history keyword is used to reduce the tractions linearly between user-selected traction-multiplier
limits over a specified step interval, defined in relax hist.
Before applying apply rf, the tunnel-boundary gridpoints are identified by using the command
APPLY xvel 0 yvel 0 to assign x- and y-fixity conditions to the selected boundary gridpoints. The
x- and y-fixity conditions identify the selected tunnel-boundary gridpoints to receive tractions.
apply rf applies the traction forces to only those fixed gridpoints. Note that the APPLY remove
command is first used to remove previous fixity conditions before assigning the new fixity conditions.
In order to create the ground reaction curve in this example, the tractions along the entire tunnel
boundary are reduced in 20% increments from the zero relaxation state. At each increment, the
calculated vertical displacement at the tunnel crown is stored in a table versus the relaxation factor
(i.e., the ratio of the current tunnel traction to the initial tunnel traction). FISH function grc is
used to store the results (see Section 12.6.2). Figure 12.4 displays the result for load relaxation of
the entire tunnel boundary from a relaxation factor of 1.0 to 0.2. (Note that at a relaxation factor of
approximately 0.2, the tunnel collapses. Figure 12.5 illustrates the collapse.)
By also relating the tunnel closure to the distance to the tunnel face (e.g., see Figure 8.4 in Section 8),
relaxation factors can be selected to correspond to selected distances to the tunnel face.
FLAC Version 5.0
12 - 6
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
1.000
8-Dec-03 15:51
step 16213
Ground Reaction Curve
Relax Factor vs Vert. Disp.
0.900
0.800
0.700
0.600
0.500
0.400
0.300
10
20
30
40
50
60
70
80
(10
-03

Figure 12.4 Ground reaction curve: vertical displacement at tunnel crown

versus relaxation factor
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

0.000
LEGEND
8-Dec-03 15:51
step 16213
-1.359E+01 <x< 6.359E+01
-6.859E+01 <y< 8.589E+00
-1.000
-8.00E-02
-7.00E-02
-6.00E-02
-5.00E-02
-4.00E-02
-3.00E-02
-2.00E-02
-1.00E-02
0.00E+00
1.00E-02
-2.000
-3.000
-4.000

0
-5.000
2E -1
-6.000
Boundary plot
0
2E 1
-0.500
0.500
1.500
2.500
(*10^1)
3.500
4.500
5.500
Figure 12.5 Displacement field during collapse for unsupported tunnel
FLAC Version 5.0
12 - 7
For this example, we do not relate the relaxation factors to specific distances to the tunnel face.
We arbitrarily choose a relaxation factor of 0.5 (50% relaxation) to define the tunnel loading state
at which the rockbolt support is installed. The factor is then reduced to 0.25 (75% relaxation) to
develop loads in the rockbolts. The relaxation factor of 0.25 corresponds to the state at which the
shotcrete is installed, and then complete relaxation (100% relaxation) is allowed to develop loads
in the shotcrete.
12.2.3 Construction Simulation
The construction steps of the excavation/support analysis follow the same sequence for each excavation stage. First, the excavation segment is nulled, and tractions are applied and relaxed by 50%.
The tractions are reduced gradually over an interval of 1000 steps and then the model is brought
to a force-equilibrium state. At this state, indicated as Ia, IIa and IIIa in Figure 12.1, the rockbolt
elements are added, representing the rockbolt support. The tunnel tractions are then reduced to 25%
over a 1000 step interval, and the model is brought to equilibrium again. At this state, indicated as
Ib, IIb and IIIb in Figure 12.1, the liner elements are added to represent installation of the shotcrete
lining. The tunnel tractions are then reduced to zero over a 1000 step interval and the model is
brought to equilibrium. The loads that develop in the rockbolts result from tunnel-load relaxation
from 50% to zero, and the loads that develop in the shotcrete result from relaxation from 25% to
zero.
By applying the relaxation over a 1000 step interval, the effects of transient waves are minimized,
and a gradual excavation of the tunnel is simulated. This is demonstrated by Figure 12.6, which
displays radial stress histories at the crown, floor and springline of the tunnel. The histories show
gradual changes in the stresses; if the relaxation loads were applied suddenly (i.e., in one step),
sudden changes would be observed in these histories and a different final state could result. (See
Section 3.10.3 in the Users Guide for further discussion on path-dependency effects of loading.)
The FISH function apply rf is used to apply tractions along the tunnel boundary in the same
manner as discussed previously in Section 12.2.2. For the side excavation, tractions are applied
along the entire boundary of this tunnel segment. For the top-heading excavation, tractions are
applied along the crown and the floor of this tunnel segment; for the bench excavation, tractions
are applied along the floor. The model should remain in equilibrium after each tunnel segment is
nulled and the tractions are applied, before they are relaxed.
FLAC Version 5.0
12 - 8
JOB TITLE : .
FLAC (Version 5.00)

(10
06
LEGEND
14-Jan-05 13:40
step 22865
-0.100
-0.200
HISTORY PLOT
Y-axis :
2 Ave. SYY
( 1, 52)
3 Ave. SYY
( 1, 31)
4 Ave. SXX
( 17, 40)
X-axis :
Number of steps
-0.300
-0.400
-0.500
-0.600
-0.700
-0.800
-0.900
12
16
20
(10
03

Figure 12.6 Stress histories at the crown (1,52), floor (1,31) and springline
(17,40)
12.3 Results
Typical results for this analysis are shown in Figures 12.7, 12.9 and 12.10. The settlement profile
of the ground surface at the end of the analysis is shown in Figure 12.7. The profile is created with
FISH function settle: y-displacements at the gridpoints along the top of the model are stored in
table 2. See Section 12.6.3.
The axial forces in the rockbolts at the end of each excavation stage are shown in Figure 12.9, and
the axial forces in the shotcrete are shown in Figure 12.10. Note that the sense of the axial force
plot depends on the order in which the structural elements are created. The sense can be changed
by assigning a maximum value with opposite sign following the max keyword when issuing PLOT
struct axial. Figure 12.8 shows the Plot Item Switches dialog, in which the maximum value is set
to -200000 to change the sense of the liner axial force plot.
FLAC Version 5.0
12 - 9
JOB TITLE : .
FLAC (Version 5.00)

(10
-02
LEGEND
14-Jan-05 13:40
step 22865
Table Plot
Vert. Dist. vs Dist.
-0.200
-0.400
-0.600
-0.800
-1.000
-1.200
-1.400
-1.600
-1.800
-2.000
10
20
30
40
50

Figure 12.7 Final settlement profile
Figure 12.8 Liner Plot Item Switches dialog; use the

the plot sense
Maximum
switch to change
FLAC Version 5.0
12 - 10
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

-2.200
LEGEND
14-Jan-05 13:32
step
7524
-4.432E+00 <x< 1.655E+01
-4.072E+01 <y< -1.974E+01
-2.600
Rockbolt Plot
Axial Force on
Structure
Max. Value
# 1 (Rockb) -2.383E+04
# 2 (Rockb) -6.366E+04
# 3 (Rockb) -9.091E+04
# 4 (Rockb) -1.103E+05
# 5 (Rockb) -1.031E+05
# 6 (Rockb) -1.016E+05
# 7 (Rockb) -1.048E+05
# 8 (Rockb) -8.041E+04
# 9 (Rockb) -3.259E+04
Marked Gridpoints
-3.000
-3.400
-3.800

-0.200
0.200
0.600
(*10^1)
1.000
1.400
(a) side excavation

JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

-2.200
LEGEND
14-Jan-05 13:37
step 15580
-4.432E+00 <x< 1.655E+01
-4.072E+01 <y< -1.974E+01
-2.600
Rockbolt Plot
Axial Force on
Structure
Max. Value
# 1 (Rockb) -1.114E+05
# 2 (Rockb) -1.026E+05
# 3 (Rockb) -1.089E+05
# 4 (Rockb) -1.188E+05
# 5 (Rockb) -1.044E+05
# 6 (Rockb) -1.279E+05
# 7 (Rockb) -1.955E+05
# 8 (Rockb) -1.653E+05
# 9 (Rockb) -1.678E+05
#11 (Rockb) -1.630E+05
#12 (Rockb) -8.910E+04
#13 (Rockb) -8.433E+04
#14 (Rockb) -8.641E+04
Marked Gridpoints
-3.000
-3.400
-3.800

-0.200
0.200
0.600
(*10^1)
1.000
1.400
(b) top-heading excavation

JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

-2.200
LEGEND
14-Jan-05 13:40
step 22865
-4.432E+00 <x< 1.655E+01
-4.072E+01 <y< -1.974E+01
-2.600
Rockbolt Plot
Axial Force on
Structure
Max. Value
# 1 (Rockb) -1.375E+05
# 2 (Rockb) -1.464E+05
# 3 (Rockb) -1.220E+05
# 4 (Rockb) -1.850E+05
# 5 (Rockb) -1.335E+05
# 6 (Rockb) -1.626E+05
# 7 (Rockb) -1.966E+05
# 8 (Rockb) -1.665E+05
# 9 (Rockb) -1.691E+05
#11 (Rockb) -1.646E+05
#12 (Rockb) -8.967E+04
#13 (Rockb) -8.655E+04
#14 (Rockb) -8.653E+04
Marked Gridpoints
-3.000
-3.400
-3.800

-0.200
0.200
0.600
(*10^1)
1.000
1.400
(c) bench excavation

Figure 12.9 Axial forces in rockbolts at 100% relaxation for each excavation
stage
FLAC Version 5.0
12 - 11
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
-2.400
14-Jan-05 13:32
step
7524
-3.201E+00 <x< 1.198E+01
-3.705E+01 <y< -2.187E+01
-2.600
Liner Plot
-2.800
Axial Force on
Structure
Max. Value
#10 (Liner)
6.002E+05
Marked Gridpoints
-3.000
-3.200
-3.400
-3.600

-0.200
0.000
0.200
0.400
(*10^1)
0.600
0.800
1.000
(a) side excavation

JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
-2.400
14-Jan-05 13:37
step 15580
-3.201E+00 <x< 1.198E+01
-3.705E+01 <y< -2.187E+01
-2.600
Liner Plot
-2.800
Axial Force on
Structure
Max. Value
#10 (Liner)
9.710E+05
Marked Gridpoints
-3.000
-3.200
-3.400
-3.600

-0.200
0.000
0.200
0.400
(*10^1)
0.600
0.800
1.000
(b) top-heading excavation

JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
-2.400
14-Jan-05 13:40
step 22865
-3.201E+00 <x< 1.198E+01
-3.705E+01 <y< -2.187E+01
-2.600
Liner Plot
-2.800
Axial Force on
Structure
Max. Value
#10 (Liner)
9.198E+05
Marked Gridpoints
-3.000
-3.200
-3.400
-3.600

-0.200
0.000
0.200
0.400
(*10^1)
0.600
0.800
1.000
(c) bench excavation

Figure 12.10 Axial forces in shotcrete at 100% relaxation for each excavation
stage
FLAC Version 5.0
12 - 12
12.4 References
Itasca Consulting Group, Inc. FLAC 3D (Fast Lagrangian Analysis of Continua in 3 Dimensions),
Version 2.1. Minneapolis: ICG, 2002.
FLAC Version 5.0
12 - 13
12.5 Data File MSTUNNEL.DAT

;Title:Multi-Stage Tunnel Excavation
;*** Branch: Ground Reaction Curve ****
new
;... State: ms gr0.sav ....
config
grid 44,80
gen (0.0,-40.0) (0.0,-20.0) (12.0,-20.0) (12.0,-40.0) i 1 25 j 21 61
gen (0.0,-20.0) (0.0,0.0) (50.0,0.0) (12.0,-20.0) ratio 1.0,1.1 &
i 1 25 j 61 81
gen (0.0,-60.0) (0.0,-40.0) (12.0,-40.0) (50.0,-60.0) &
ratio 1.0,0.9090909 i 1 25 j 1 21
gen (12.0,-40.0) (12.0,-20.0) (50.0,0.0) (50.0,-60.0) &
ratio 1.1,1.0 i 25 45 j 21 61
gen arc 0.0,-33.93 5.5,-26.0 35
gen arc 0.0,-17.65 0.0,-34.65 20
ini x 7.9 y -28.9 i 17 j 44
gen arc 2.0,-31.2 7.9,-28.9 35
gen arc 4.6,-30.2 5.8,-33.55 90
gen arc 12.0,-30.0 5.5,-26.0 60
ini x 5.28 y -33.41 i 11 j 34
mark i 1 9 j 42
mark i 11 j 34
group rock notnull
model mohr notnull group rock
friction=33.0 dilation=0.0 tension=1000.0 notnull group rock
group top heading region 5 47
model mohr group top heading
friction=33.0 dilation=0.0 tension=1000.0 group top heading
group bench region 6 36
model mohr group bench
friction=33.0 dilation=0.0 tension=1000.0 group bench
group side region 13 41
FLAC Version 5.0
12 - 14
model mohr group side

friction=33.0 dilation=0.0 tension=1000.0 group side
group top heading i 11 j 48
group bench i 11 j 33
fix y i 1 25 j 1
fix y i 45 j 21
fix x i 45 j 21 61
fix x i 25 j 1
fix x i 1
set gravity=9.81
initial syy -1589000.0 var 0.0,1589000.0
initial sxx -794600.0 var 0.0,794600.0
initial szz -794600.0 var 0.0,794600.0
save ms gr0.sav
model null group side
model null group top heading
model null group bench
set echo off
call grc.fis
set relax value=1.0 isn=1
grc
apply xvelocity 0.0 from 1,32 to 1,52
apply yvelocity 0.0 from 1,32 to 1,52
cycle 1
set echo off
call relax tractions.fis
set ninc=1000 rstart=1.0 rstop=0.8 ib ap=1 ie ap=17 jb ap=32
set je ap=52
relax tractions
history 1 relax hist
fix x i 1 j 52
fix x i 1 j 32
solve
set echo off
FLAC Version 5.0
12 - 15
call grc.fis
grc
save ms gr1.sav
apply remove mech from 1,32 to 1,52
cycle 1
set echo off
set ninc=1000 rstart=1.0 rstop=0.8 ib ap=1 ie ap=17 jb ap=32 je ap=52
relax tractions
fix x i 1 j 52
fix x i 1 j 32
solve
set echo off
call grc.fis
grc
save ms gr2.sav
cycle 1
set echo off
relax tractions
fix x i 1 j 52
fix x i 1 j 32
solve
set echo off
call grc.fis
grc
save ms gr3.sav
cycle 1
FLAC Version 5.0
12 - 16
set echo off

relax tractions
fix x i 1 j 52
fix x i 1 j 32
solve
set echo off
call grc.fis
grc
save ms gr4.sav
cycle 1
set echo off
relax tractions
fix x i 1 j 52
fix x i 1 j 32
solve
set echo off
call grc.fis
grc
save ms gr5.sav
cycle 1
set echo off
relax tractions
fix x i 1 j 52
fix x i 1 j 32
solve
set echo off
call grc.fis
FLAC Version 5.0
12 - 17
grc
save ms gr6.sav
cycle 1
set echo off
relax tractions
fix x i 1 j 52
fix x i 1 j 32
cycle 5000
set echo off
call grc.fis
grc
save ms gr7.sav
;*** Branch: Multi-Stage Tunnel ****
new
;... State: ms 0.sav ....
config
grid 44,80
gen (0.0,-40.0) (0.0,-20.0) (12.0,-20.0) (12.0,-40.0) i 1 25 j 21 61
gen (0.0,-20.0) (0.0,0.0) (50.0,0.0) (12.0,-20.0) ratio 1.0,1.1 &
i 1 25 j 61 81
gen (0.0,-60.0) (0.0,-40.0) (12.0,-40.0) (50.0,-60.0) &
ratio 1.0,0.9090909 i 1 25 j 1 21
gen (12.0,-40.0) (12.0,-20.0) (50.0,0.0) (50.0,-60.0) &
ratio 1.1,1.0 i 25 45 j 21 61
gen arc 0.0,-33.93 5.5,-26.0 35
gen arc 0.0,-17.65 0.0,-34.65 20
ini x 7.9 y -28.9 i 17 j 44
gen arc 2.0,-31.2 7.9,-28.9 35
gen arc 4.6,-30.2 5.8,-33.55 90
gen arc 12.0,-30.0 5.5,-26.0 60
FLAC Version 5.0
12 - 18
ini x 5.28 y -33.41 i 11 j 34

mark i 1 9 j 42
mark i 11 j 34
group rock notnull
model mohr notnull group rock
prop density=2700.0 bulk=5.5550003E8 shear=4.166E8 cohesion=10000.0
friction=33.0 dilation=0.0 tension=1000.0 notnull group rock
group top heading region 5 47
group bench region 6 36
group side region 13 41
model mohr group side
friction=33.0 dilation=0.0 tension=1000.0 group side
group top heading i 11 j 48
group bench i 11 j 33
fix y i 1 25 j 1
fix y i 45 j 21
fix x i 45 j 21 61
fix x i 25 j 1
fix x i 1
set gravity=9.81
initial syy -1589000.0 var 0.0,1589000.0
initial sxx -794600.0 var 0.0,794600.0
initial szz -794600.0 var 0.0,794600.0
save ms 0.sav
;... State: ms Ia.sav ....
model null group side
apply xvelocity 0.0 long from 11,34 to 11,34
apply yvelocity 0.0 long from 11,34 to 11,34
cycle 1
set echo off
FLAC Version 5.0
&
&
&
&
&
&
12 - 19

set je ap=49
relax tractions
history 1 relax hist
history 4 sxx i=17, j=40
solve
save ms Ia.sav
;... State: ms Ib.sav ....
struct node 8 11.75,-32.25
struct node 10 12.12,-30.0
struct node 13 7.47,-27.95
struct node 14 10.75,-25.95
struct rockbolt begin node 1 end node 2 seg 5 prop 4001
struct prop 4001
struct prop 4001 e 2.04999999E11 area 5.0E-4 cs sstiff 1.50000005E10 &
cs scoh 800000.0 yield 500000.0 perimeter 0.0785
apply remove mech long from 11,34 to 11,34
cycle 1
FLAC Version 5.0
12 - 20
set echo off

set je ap=49
relax tractions
solve
save ms Ib.sav
;... State: ms Ic.sav ....
cycle 1
set echo off
set je ap=49
relax tractions
FLAC Version 5.0
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
node 83 grid 16,46

node 84 grid 15,46
node 85 grid 15,47
node 86 grid 14,47
node 87 grid 14,48
node 88 grid 13,48
node 89 grid 13,49
node 90 grid 12,49
node 91 grid 12,48
node 92 grid 11,48
node 93 grid 11,47
node 94 grid 11,46
node 95 grid 10,46
node 96 grid 10,45
node 97 grid 10,44
node 98 grid 10,43
liner begin node 55
liner begin node 56
liner begin node 57
liner begin node 58
liner begin node 59
liner begin node 60
liner begin node 61
liner begin node 62
liner begin node 63
liner begin node 64
liner begin node 65
liner begin node 66
liner begin node 67
liner begin node 68
liner begin node 69
liner begin node 70
liner begin node 71
liner begin node 72
liner begin node 73
liner begin node 74
liner begin node 75
liner begin node 76
liner begin node 77
liner begin node 78
liner begin node 79
liner begin node 80
liner begin node 81
liner begin node 82
liner begin node 83
liner begin node 84
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
end
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
node
12 - 21
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
FLAC Version 5.0
12 - 22

struct prop 5001
struct prop 5001 e 5.5000003E9
solve
save ms Ic.sav
node
node
node
node
node
node
node
node
node
node
node
node
node
node
86
87
88
89
90
91
92
93
94
95
96
97
98
55
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
area 0.1 I 8.333E-5 thickness 0.1 pratio 0.2
;... State: ms IIa.sav ....

model null group top heading
struct liner delete 81
cycle 1
set echo off
relax tractions
fix x i 1 j 52
fix x i 1 j 42
solve
save ms IIa.sav
FLAC Version 5.0
12 - 23
;... State: ms IIb.sav ....

struct node 100 6.8,-21.85
struct node 101 3.44,-24.9
struct node 103 2.13,-24.5
struct node 104 3.05,-20.4
cycle 1
set echo off
relax tractions
fix x i 1 j 52
fix x i 1 j 42
solve
save ms IIb.sav
;... State: ms IIc.sav ....
cycle 1
set echo off
relax tractions
fix x i 1 j 52
fix x i 1 j 42
FLAC Version 5.0
12 - 24

solve
save ms IIc.sav
;... State: ms IIIa.sav ....
model null group bench
cycle 1
FLAC Version 5.0
12 - 25
set echo off

relax tractions
fix x i 1 j 32
solve
save ms IIIa.sav
;... State: ms IIIb.sav ....
cycle 1
set echo off
relax tractions
fix x i 1 j 32
solve
save ms IIIb.sav
;... State: ms IIIc.sav ....
cycle 1
set echo off
relax tractions
fix x i 1 j 32
FLAC Version 5.0
12 - 26
struct liner begin node

solve
save ms IIIc.sav
138
139
140
141
142
143
144
145
146
147
148
end
end
end
end
end
end
end
end
end
end
end
node
node
node
node
node
node
node
node
node
node
node
139
140
141
142
143
144
145
146
147
148
149
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
5001
;... State: ms settle.sav ....

set echo off
call settle.fis
settle
save ms settle.sav
;plot name: rockbolt forces
plot hold struct rockbolt axial fill max 1000000.0 mark
;plot name: shotcrete forces
plot hold struct liner axial fill max -2000000.0 mark
;plot name: tractions
plot hold apply max 170000.0 bound liner iwhite
;plot name: stress histories
plot hold history 2 3 line 4 line
;plot name: SE numbers
plot hold bound struct liner node struct liner element
;plot name: settlement profile
label table 2
Vert. Dist. vs Dist.
;plot name: y-displacement
plot hold ydisp fill inv displacement bound
;plot name: ground reaction curve
label table 1
Relax Factor vs Vert. Disp.
FLAC Version 5.0
12 - 27
12.6 FISH Functions

12.6.1 RELAX TRACTIONS.FIS
;Name:relax tractions
;Diagram:
;Input:ninc/int/1000/relaxation steps
;Input:rstart/float/1.0/beginning relaxation ratio
;Input:rstop/float/0.5/ending relaxation ratio
;Input:ib ap/int/10/beginning i gridpoint for traction range
;Input:ie ap/int/17/ending i gridpoint for traction range
;Input:jb ap/int/34/beginning j gridpoint for traction range
;Input:je ap/int/49/ending j gridpoint for traction range
; FISH functions to control relaxation of tunnel tractions
def relax ini
nstart = step
nstop = nstart + ninc
end
def relax hist
if step < nstop
step inc = float(step - nstart)
relax hist = rstart - ((rstart - rstop)/ float(ninc)) * step inc
else
relax hist = rstop
endif
end
def apply rf
loop ii (ib ap,ie ap)
loop jj (jb ap,je ap)
if and(flags(ii,jj), 2) = 2 then
if and(flags(ii,jj), 4) = 4 then
xftmp = -xforce(ii,jj)
yftmp = -yforce(ii,jj)
command
apply xforce=xftmp hist relax hist i=ii j=jj
apply yforce=yftmp hist relax hist i=ii j=jj
end command
endif
endif
endloop
endloop
end
FLAC Version 5.0
12 - 28
def relax tractions

relax ini
relax hist
apply rf
end
12.6.2 GRC.FIS
;Name:grc
;Diagram:
;Input:relax value/float/1.0/relaxation factor
;Input:isn/int/1/step number
def grc
xtable(1,isn) = -ydisp(1,52)
ytable(1,isn) = relax value
end
12.6.3 SETTLE.FIS
;Name:settle
;Diagram:
; generate a settlement profile plot
def settle
loop i1 (1,25)
xtable(2,i1) = x(i1,jgp)
ytable(2,i1) = ydisp(i1,jgp)
end loop
end
FLAC Version 5.0
Embankment Loading on a Cam-Clay Foundation
13 - 1
13 Embankment Loading on a Cam-Clay Foundation

The response of a saturated soil foundation to loading by an embankment is studied in this example.
The foundation is 10 meters deep and the groundwater free surface is at the ground level. The
embankment is 8 meters wide. The soil behavior corresponds to a Cam-clay material. The initial
stress and pore pressure states correspond to equilibrium under gravity with a ratio of horizontal
to vertical effective stress of 6/13. The weight of the embankment is simulated by an applied
surcharge, and drainage occurs at the soil surface.
The soil has the following properties:
drained Poissons ratio ()
soil constant (M)
slope of normal consolidation line ()
slope of elastic swelling line ()
reference pressure (p1 )
specific volume at reference pressure (v )
porosity
dry density ()
0.3
0.888
0.161
0.062
100.0 Pa
2.858
0.3
2 103 kg/m3
The clay is lightly over-consolidated, and the initial value of the cap pressure, pc , is equal to 1.6
105 Pa in the example. (Note that for a normally consolidated soil, the value for pc is equal to
1.579 105 Pa at the base of the clay layer, where p = 8.33 104 Pa and q = 7.0 104 Pa.) The
drained Poissons ratio of the material is assumed to remain constant during the simulation.
The foundation has a permeability, k, of 1012 (m/s)/(Pa/m). The soil moduli are functions of the
mean effective pressure and the soil specific volume, quantities which vary in space and evolve
during the simulation. The average value of K + 4/3G stays, however, in the order of 106 Pa, or
two orders of magnitude lower than the water bulk modulus (Kw is 2 108 Pa). The diffusivity,
c, is thus controlled by the soil material in this example, its magnitude can be estimated from the
formula c = k(K + 4/3G), and is of the order of 106 m2 /s. The time scale for the diffusion
process can be estimated using tc = L2 /c, where L is the model height. Using L = 10 m, we
have that tc is on the order of 3 years. Compared to that time, construction of the embankment
may be assumed to occur instantaneously. An undrained analysis is first conducted to evaluate the
foundation settlement in the short-term after building of the embankment; the long-term response
is then monitored after allowing drainage from the soil surface.
FLAC Version 5.0
13 - 2

The model, represented in Figure 13.1, takes advantage of half symmetry. The size is 20 meters
wide and 10 meters deep. Note that the width of the model is not necessarily large enough to
accurately represent an extensive soil layer; the model is intended for illustrative purposes only.
The mechanical boundary conditions correspond to roller boundaries along the symmetry line and
the far boundary of the model, and to fixed displacements in the x- and y-direction at the model base.
The maximum bulk modulus of the clay (bulk) is set to 5 106 Pa, a value that is approximately
twice the initial value of the actual bulk modulus (bulk current) at the bottom of the clay layer.
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
14-Jan-05 14:45
step
6645
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
8
7
6
0.800
Grid plot
5E 0
Net Applied Forces

0.400
1E 5
History Locations
0.000
-0.400

0.200
0.600
1.000
(*10^1)
1.400
1.800
Figure 13.1 Model geometry

The first stage of the simulation corresponds to the short time response of the system in which no
flow is assumed to take place. The command SET ow off is specified. Loading by the embankment
is simulated by progressive application of a pressure of 50 kPa on a 4 meter section of the model
top boundary. This procedure avoids overshoots related to transient loading. Once the full load is
attained, the model is cycled to equilibrium. During this stage, pore pressures develop as a result
of volumetric deformations, but do not dissipate.
FLAC Version 5.0
13 - 3
In the second stage, fluid flow is allowed to develop by issuing the command SET ow on. Water
then drains through the top of the model where the pore pressure is fixed at zero, and additional
settlement takes place under the embankment. The SOLVE auto on command, used to perform
the coupled simulation, requires parameters that determine the accuracy of the solution. These
parameters may need to be different if different properties or model conditions are used. Refer to
Section 1.8.6 in Fluid-Mechanical Interaction for a discussion on these topics.
Stresses, pore pressures and vertical displacements are monitored during the calculation. The data
file for this problem, EMC.DAT, is listed in Section 13.4.
Displacement vectors, vertical displacement contours, the pore pressure distribution and the plastic
state at the end of the undrained and drained numerical simulations are presented in Figures 13.2
to 13.9. The vertical displacement histories in Figure 13.10, recorded at four monitoring points
(locations 2, 3, 4 and 5 in Figure 13.1), indicate that the maximum settlement under the embankment
increases from approximately 0.14 cm to 0.19 cm as a result of drainage. Note that the displacement vectors in Figure 13.3, and vertical displacement contours in Figure 13.5, correspond to the
combined undrained and drained displacements.
In Figure 13.11, the graph of pore pressure evolution at two monitoring points (locations 8 and 9 in
Figure 13.1) confirms that a steady-state flow has been reached by the end of the drained simulation.
FLAC Version 5.0
13 - 4
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
14-Jan-05 14:45
step
6645
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Boundary plot
0
5E 0
0.400
2E -1
0.000
-0.400

0.200
0.600
1.000
(*10^1)
1.400
1.800
Figure 13.2 Displacement vectors undrained response
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
14-Jan-05 15:17
step 84237
Flow Time
1.0007E+08
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Boundary plot
0
5E 0
0
0.400
5E -1
0.000
-0.400

0.200
0.600
1.000
(*10^1)
1.400
1.800
Figure 13.3 Displacement vectors end of drained simulation
FLAC Version 5.0
13 - 5
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
14-Jan-05 14:45
step
6645
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Boundary plot
0
5E 0
-1.25E-01
-1.00E-01
-7.50E-02
-5.00E-02
-2.50E-02
0.00E+00
2.50E-02
0.400
0.000
-0.400

0.200
0.600
1.000
(*10^1)
1.400
1.800
Figure 13.4 Vertical displacement contours undrained response
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
14-Jan-05 15:17
step 84237
Flow Time
1.0007E+08
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Boundary plot
0
5E 0
-1.75E-01
-1.50E-01
-1.25E-01
-1.00E-01
-7.50E-02
-5.00E-02
-2.50E-02
0.00E+00
0.400
0.000

-0.400

0.200
0.600
1.000
(*10^1)
1.400
1.800
Figure 13.5 Vertical displacement contours end of drained simulation
FLAC Version 5.0
13 - 6
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
14-Jan-05 14:45
step
6645
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Boundary plot
0
5E 0

0.00E+00
2.00E+04
4.00E+04
6.00E+04
8.00E+04
1.00E+05
1.20E+05
0.400
0.000
-0.400

0.200
0.600
1.000
(*10^1)
1.400
1.800
Figure 13.6 Pore pressure contours undrained response
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
14-Jan-05 15:17
step 84237
Flow Time
1.0007E+08
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Boundary plot
0
5E 0

0.00E+00
2.00E+04
4.00E+04
6.00E+04
8.00E+04
1.00E+05
0.400
0.000
-0.400

0.200
0.600
1.000
(*10^1)
1.400
Figure 13.7 Pore pressure contours drained response
FLAC Version 5.0
1.800
13 - 7
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
14-Jan-05 14:45
step
6645
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Boundary plot
0
5E 0
0.400
0.000
-0.400

0.200
0.600
1.000
(*10^1)
1.400
1.800
Figure 13.8 Plastic state undrained response
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
1.200
14-Jan-05 15:17
step 84237
Flow Time
1.0007E+08
-1.111E+00 <x< 2.111E+01
-6.111E+00 <y< 1.611E+01
0.800
Boundary plot
0
5E 0
0.400
0.000
-0.400

0.200
0.600
1.000
(*10^1)
1.400
1.800
Figure 13.9 Plastic state drained response
FLAC Version 5.0
13 - 8
JOB TITLE : .
FLAC (Version 5.00)

(10
-01
LEGEND
14-Jan-05 15:17
step 84237
Flow Time
1.0007E+08
1.600
HISTORY PLOT
Y-axis :
Rev 2 Y displacement( 1, 11)
1.200
Rev
3 Y displacement( 3, 11)
0.800
Rev
Rev
0.400
X-axis :
10 Groundwater flow time
0.000
-0.400
9
(10
07

Figure 13.10 Vertical displacement histories drained response
JOB TITLE : .
FLAC (Version 5.00)

(10
04
LEGEND
14-Jan-05 15:17
step 84237
Flow Time
1.0007E+08
6.000
HISTORY PLOT
Y-axis :
8 Pore pressure ( 1, 10)
5.000
4.000
X-axis :
3.000
2.000
1.000
9
(10

Figure 13.11 Pore pressure histories drained response
FLAC Version 5.0
07
13 - 9
13.4 Data File EMC.DAT

;... State: emc.sav ....
;----------------------------------------------------------; emc.dat
; Loading by an embankment on a Cam-clay foundation
;----------------------------------------------------------config gw extra 3
grid 20 10
; --- model and properties --model cam
prop dens 2000 poiss .3 bu 5e6
prop mm 0.888 lambda 0.161 kappa 0.062 mpc 160e3 mp1 1e3 mv l 2.858
;
prop perm 1e-12 por .3
water bulk 2e8 tens 1e10 den 1000
; --- fish function --def ramp
ramp = min(1.0,float(step)/2000.0)
end
; --- boundary conditions --fix x i=1
fix x y j=1
fix x i=21
fix pp j=11
fix sat j=11
apply nstress=-5e4 his ramp from 1,11 to 5,11
; --- initial conditions --ini sxx -1.6e5 var 0 1.6e5
ini szz -1.6e5 var 0 1.6e5
ini syy -2.3e5 var 0 2.3e5
ini pp 1e5 var 0 -1e5
; --- settings --set flow off
set g 10
; --- histories --hist nstep 100
hist unbal
hist yd i=1 j=11
hist yd i=3 j=11
hist yd i=5 j=11
hist yd i=7 j=11
hist syy i=1 j=10
hist sxx i=1 j=10
FLAC Version 5.0
13 - 10
hist pp i=1 j=10

hist pp i=2 j=8
; --- undrained response --solve sratio 1e-3
save emc.sav
;... State: emc2.sav ....
; --- drained response --set flow on
set nmech 50
set for 50
set clock 100000 step 1000000
hist gwtime
solve auto on age 1.e8
save emc2.sav
;plot name: Model geometry
plot hold grid apply aforce history
plot hold bound displacement
;plot name: Vertical displacement contours
plot hold bound ydisp fill inv
;plot name: Pore pressure contours
plot hold bound pp fill
;plot name: Plastic state
;plot name: Vertical displacement histories
plot hold history -2 line -3 line -4 line -5 line vs 10
;plot name: Pore pressure histories
plot hold history 8 line 9 line vs 10
FLAC Version 5.0
Dewatered Construction of a Braced Excavation
14 - 1
14 Dewatered Construction of a Braced Excavation

A braced excavation is constructed in saturated ground. The excavation is dewatered during construction and is supported by diaphragm walls that are braced at the top by horizontal struts. The
purpose of the FLAC analysis is to evaluate the deformation of the ground adjacent to the excavation, and the performance of the wall and struts, throughout the construction stages. The analysis
starts from the stage after the wall has been constructed, but prior to any excavation. Dewatering,
excavation and installation of struts are simulated in separate construction stages. A calculation of
the factor of safety of the excavation is made after the final stage.
In practice, the construction may involve several stages of dewatering, excavation and adding of
support. For simplicity, in this example, only three construction stages are analyzed: (1) dewatering
to a 20 m depth in the region to be excavated; (2) excavation to a 2 m depth; and (3) installation
of a horizontal strut and excavation to a 10 m depth. Additional excavation stages can readily be
incorporated in the FLAC analysis, as required.
Figure 14.1 shows the geometry for this example. The excavation is 20 m wide and the final depth
is 10 m. The diaphragm walls extend to a 30 m depth and are braced at the top by horizontal struts
at a 2 m interval. The ground consists of two soil layers: a 20 m thick soft clay underlain by a stiff
sand layer that extends to a great depth. The initial water table is at the ground surface.
Figure 14.1 Geometry for braced excavation example
FLAC Version 5.0
14 - 2
The drained material properties associated with the two material types are summarized in Table 14.1.
The soil/wall interface is relatively smooth; strength properties are lower than those of the adjacent
soils. The interface friction angle is 12.5 and the interface cohesion is 2500 Pa.
Table 14.1 Drained properties for sand and clay layers
(kg/m3 )
Dry density
Youngs modulus (MPa)
Poissons ratio
Cohesion (Pa)
Friction angle (degrees)
Dilation angle (degrees)
Mobility coefficient (m2 /(Pa-sec))
Porosity
Sand layer
Clay layer
1700
40.0
0.3
1000
32
2
1010
0.3
1600
10.0
0.35
5000
25
0
1010
0.3
The properties selected in this example to simulate the behavior of the diaphragm wall and the struts
are listed in Tables 14.2 and 14.3. Note that for the two-dimensional FLAC analysis, the Youngs
modulus of the wall should be divided by (1 2 ) to represent the plane-stress formulation for the
structural elements in the plane-strain condition of a continuous wall. Thus, a value of 5.95 GPa is
input to FLAC for the wall elastic modulus.
The strut properties are listed in Table 14.3. Note that the spacing of the struts is 2 m. A simple
way to simulate the three-dimensional effect of the strut spacing in the FLAC model is with linear
scaling of the material properties of the struts by dividing by the strut spacing. For this example,
by using elastic beam elements, it is only necessary to scale the elastic modulus and the density of
the struts. This is done in FLAC automatically when the parameter, spacing, is specified.
Table 14.2 Properties of the diaphragm wall
Equivalent thickness (m)
Density (kg/m3 )
Youngs modulus (GPa)
Poissons ratio
Moment of inertia (m4 )
1.26
2000
5.712
0.2
0.167
Table 14.3 Properties of the strut

Area (m2 )
Spacing (m)
Density (kg/m3 )
Youngs modulus (GPa)
Moment of inertia (m4 )
FLAC Version 5.0
1.0
2.0
3000
4.0
0.083
14 - 3

The recommended procedure to simulate this type of problem with FLAC is illustrated by performing
the analysis in seven steps:
Step 1 Generate the model grid and assign material models and properties and boundary conditions to represent the physical system.
Step 2 Determine the initial in-situ stress state of the ground prior to construction.
Step 3 Determine the initial in-situ stress state of the ground with the diaphragm wall
installed.
Step 4 Lower the water level within the region to be excavated to a depth of 20 m
below the ground surface.
Step 5 Excavate to a depth of 2 m.
Step 6 Install the horizontal struts at the top of the wall and then excavate to a depth
of 10 m.
Step 7 Determine the factor of safety for the 10 m excavation.
The model is created using FLAC s graphical interface, the GIIC . Upon entering the GIIC, the
groundwater flow option, the adjust total stress option*, structural elements and factor-of-safety
calculation are activated from the Model Options dialog. The Project Tree Record format and SI
system of units are also selected for this example. The dialog is shown in Figure 14.2.
We set up a project file to save the model state at various stages of the simulation. We click on ?
in the Project File (*.prj) dialog to select a directory in which to save the project file. We assign
a title to our project and save the project as EXCAVATE.PRJ. (Note that the .PRJ extension
is assigned automatically.) This project file contains the project tree and allows direct access to
all the save (.SAV) files that we will create for the different stages of the analysis. We can stop
working on the project at any stage, save it and re-open it at a later time simply by opening the
project file (from the File / Open Project menu item); the entire project and associated save files
will be accessible in the GIIC.
A record of the FLAC commands used to create this model can also be saved separately using
the File / Export Record menu item. A listing of the record created for this model is given in
Section 14.3.
* The automatic adjustment of total stresses for external pore-pressure change is selected because we
will use this facility when we change from an unsaturated to a saturated state. See Section 1.9.7 in
Fluid-Mechanical Interaction.
FLAC Version 5.0
14 - 4
Figure 14.2 Model options selected for braced excavation example
14.2.1 Step 1: Model Generation

We begin the analysis by building the model grid using the Build tool. The braced excavation is
a common form of retaining structure used in geotechnical engineering. We can find this type of
geometry in the grid library available from the Build / Library tool. We click on the Retaining wall,
2 interfaces library item to access this grid type. The grid library dialog for this tool is shown in
Figure 14.3.
We click OK to begin manipulating this grid to fit our problem geometry. Note that it is only necessary
to consider half of the problem region shown in Figure 14.1 because of the symmetric geometry.
The grid corners are selected to correspond to the right half of the excavation with the axes origin
at the centerline of the excavation. A uniform mesh density is assigned with a zone size of 1 m.
By using this library item, the wall is created automatically as a set of beam elements, connected
to the grid on both sides by interfaces. The FLAC commands are automatically generated to create
this grid when the Execute button is pressed. The grid created for the braced excavation example is
shown in Figure 14.4. The figure also shows the location of the wall beams, and identifies the two
interfaces by ID numbers 101 and 102.
FLAC Version 5.0
14 - 5
Figure 14.3 Grid library tool for retaining wall grid
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

101
4.000
LEGEND
14-Jan-05 9:39
step
0
-5.044E+00 <x< 5.004E+01
-7.544E+00 <y< 4.754E+01
3.000
Grid plot
0
1E 1
interface id#s
Beam plot
2.000
102
1.000
0.000

0.000
1.000
2.000
(*10^1)
3.000
4.000
Figure 14.4 Grid created for braced excavation
FLAC Version 5.0
14 - 6
Next, we assign material properties for the soils and soil/wall interface. The interface properties are
prescribed using the Alter / Interface tool. We click on the Property radio button in this pane and then
click on the circled number at one end of the interface highlighted in the model plot. An Interface
Properties dialog will appear, as shown in Figure 14.5. We select the Unbonded button and then enter
the interface properties.
Figure 14.5 Interface material properties dialog

It is usually reasonable to select the interface normal- and shear-stiffness properties such that the
stiffness is approximately ten times the equivalent stiffness of the stiffest neighboring zone. By
doing this, the deformability at the interface will have minimal influence on both the compliance
of the total model and the calculational speed. The equivalent stiffness of a zone normal to the
interface is

max
K + 43 G
zmin
(14.1)
where:K & G are the bulk and shear moduli, respectively; and
zmin is the smallest width of an adjoining zone in the normal direction
The max [ ] notation indicates that the maximum value over all zones adjacent to the interface is
to be used (e.g., there may be several materials adjoining the interface).
In this example, the smallest grid width adjacent to the interface is 1 m, and the maximum equivalent
stiffness is approximately 55 MPa. Therefore, we select a representative value of 550 MPa/m for
the normal and shear stiffnesses.
FLAC Version 5.0
14 - 7
The soil properties are assigned in the Material / Assign tool. We store the material properties in the
GIIC material database. This database can then be utilized by other projects if needed. We click
on the Database button to open the Material list dialog. We then click on the Edit tab to create two
materials and assign properties. Figure 14.6 shows the Material list dialog that is edited to contain
the properties for stiff sand. After the properties for both soils are added to the database, they can
be saved in a file (e.g., EXCAVATE.GMT) that can be loaded for other analyses. We click OK to
select these materials for assignment in the present model.
Figure 14.6 Material database dialog

We highlight Excavation: sand in the Assign tool, check the Layer radio button, and click-and-drag
the mouse over the zone layers we wish to assign as stiff sand, then repeat the process for the zones
assigned as soft clay. Figure 14.7 shows the model with the materials assigned.
FLAC Version 5.0
14 - 8
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:39
step
0
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
User-defined Groups
Excavation:sand
Excavation:clay
Grid plot
2.000
1E 1
1.000
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.7 Materials assigned to the model

The groundwater properties, porosity and permeability, are assigned in the Material / GWProp tool. We
click on the SetAll button to open the Model Groundwater properties dialog to enter these properties.
Note that the permeability required by FLAC is actually the mobility coefficient (i.e., the coefficient
of the pore pressure term in Darcys law). When we click OK , these properties are assigned to all
zones in the model.
We also need to specify the material properties for the diaphragm wall because the initial model
grid includes the structural elements representing the wall. We click on the Structure / SEProp tool and
then click on one of the beam elements to open the Beam Element Properties dialog. The dialog
is divided into two panes, as shown in Figures 14.8 and Figure 14.9. We enter the area, Youngs
modulus and moment of inertia from Table 14.2. (The area is 1.26 m2 because the two-dimensional
model assumes a 1 m dimension out of the analysis plane.) We do not assign the density of the wall
at this stage because we will first calculate the equilibrium stress state before the wall is constructed;
this is done by neglecting the weight of the wall.
FLAC Version 5.0
14 - 9
Figure 14.8 Beam properties assigned for the diaphragm wall

mechanical properties
Figure 14.9 Beam properties assigned for the diaphragm wall

geometric properties
It is now necessary to prescribe the boundary conditions for the model. We use the In Situ / Fix tool
to assign the mechanical boundary conditions. We press the X & Y radio button, then hold the left
mouse button and drag the mouse along the bottom boundary of the grid. Gridpoints are marked
and, when the button is released, a letter denoting the fixity condition is printed at the selected
gridpoints. In addition, the corresponding FIX command is listed in the Changes sub-menu. We
then click on the X radio button and drag the mouse along the left and right boundaries of the grid.
We press Execute to send these commands to FLAC. The resulting boundary conditions are shown
in Figure 14.10.
FLAC Version 5.0
14 - 10
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

LEGEND
14-Jan-05 9:39
step
0
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
Grid plot
0
1E 1
Fixed Gridpoints
X X-direction
B Both directions
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
B BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB
B BBBBBBBBBB
4.000
3.000
2.000
1.000
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.10 Fixed boundary conditions

This completes the first step of our analysis: generation of the model grid and assignment of material
models and properties and boundary conditions. We save the model state at this stage by clicking
on the Save button at the bottom of the record pane. We name the saved state EX01.SAV; a new
branch with this name appears in the project tree shown in the record pane. See Figure 14.11.
Note that the commands associated with this branch have now been grayed out. If we find we have
made a mistake or wish to modify these commands, we can press the Edit button at the bottom of
this pane. We can then edit the commands in this pane and re-execute them in FLAC by pressing the
Rebuild
button at the bottom of the pane. The state must be saved again if modifications are made.
FLAC Version 5.0
14 - 11
Figure 14.11 Project tree at completion of step 1
14.2.2 Step 2: Initial Stress State

The initial stress state for this example corresponds to a saturated soil with the phreatic surface at
the ground surface. Pore pressure and total (and effective) stress gradients must be compatible at
the initial state. We can either start from an unsaturated state and then include pore pressures (i.e.,
raise the water table), or we can determine the saturated stress state directly. In this example, we
will start from an unsaturated state. Then, we will show how different calculational approaches can
be followed to achieve the same initial stress state for the saturated condition.
14.2.2.1 Unsaturated State
We first determine the unsaturated stress state before the phreatic surface is introduced. We specify gravity using the Settings / Gravity tool. We select a gravitational magnitude of 10.0 m/sec2 to
simplify this example. We also turn off the groundwater-flow mode from the Settings / GW tool. The
recommended approach to determine the initial unsaturated stress state is to use the In Situ / Initial
tool to set the initial stresses to conform to the gravitational loading and to the expected horizontal
stress state. Typically, this is defined by the ratio of horizontal to vertical stresses, Ko . For this
example, we assume Ko = 0.5. The initial stress components are entered by checking the Zones
radio button, and then clicking on the syy menu item to set the vertical stress distribution, and on
the sxx and szz menu items to set the horizontal stress distributions.
FLAC Version 5.0
14 - 12
The vertical stresses, yy , are equal to -gz (i.e., gravitational acceleration times mass density times
depth). For this example with two soil layers of different densities, and the coordinate axes origin
at the bottom of the model, the vertical stresses are defined by the following equations:
(y ys )
(ye ys )
(0 y 20)
(14.2)
(y ys )
(ye ys )
(20 y 40)
(14.3)
yy = 660, 000 + 340, 000
yy = 320, 000 + 320, 000
where ys and ye are the starting and ending y-coordinates for each layer.* We use these equations to
enter the vertical stress variation in the yy-component of stress dialog. After clicking on syy in the
Zone Info/Stress menu, we click and drag the mouse over the zones corresponding to (0 y 20)
and then press the Assign button. The yy-component of stress dialog opens, and we enter the values
from Eq. (14.2). We repeat the process for Eq. (14.3). Figure 14.12 shows the dialog for Eq. (14.3).
Figure 14.12 Dialog to enter vertical stress variation

We repeat this procedure for the xx and zz stress variations, recognizing that these stresses are
equal to half the vertical stresses. Finally, we press Execute to send these commands to FLAC.
* See Eq. (3.2) in the Users Guide for a description of the general equation to assign stress gradients
in a FLAC model.
FLAC Version 5.0
14 - 13
The initial stress state should now be in equilibrium for the applied gravitational loading. We
confirm this by using the Run / Solve tool to begin the calculation process. When we click on Solve ,
a Solve options dialog opens and informs us that a mechanical-only calculation will be performed.
We select the Solve initial equilibrium as elastic model radio button* and press Execute . The model takes a
few steps and stops. (A few steps are needed to make a small adjustment to the equilibrium state,
to account for the effect of the stiffnesses of the beam elements and the interfaces.) We save this
state as EX02.SAV in our project tree. The vertical stress contours for the initial unsaturated state
are shown in Figure 14.13. The output plots for this model, such as that shown in Figure 14.13,
are generated from the Plot / Model tool. This opens a Plot items dialog in which we can select one
or more plot items to add to a plot view. For example, Figure 14.13 contains the yy total stress
zone-contour plot item and the boundary plot item. Note that the contour interval is changed to
5 104 in this figure by using the Edit button in the Plot items dialog. The grayscale for the filled
contour plot is also reversed by selecting the Inverse switch in the Plot Item Switches dialog.
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:39
step
51
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
YY-stress contours
-6.50E+05
-5.50E+05
-4.50E+05
-3.50E+05
-2.50E+05
-1.50E+05
-5.00E+04
2.000

Boundary plot
0
1.000
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.13 Vertical stress contours for initial unsaturated state
* This option performs the initial calculation in two steps: first the calculation uses high values for
strength properties to provide an elastic solution; and then it uses the actual strength values. By
following this approach, a uniform stress distribution is ensured. See Section 3.4.6 in the Users
Guide for further discussion on this topic.
FLAC Version 5.0
14 - 14
14.2.2.2 Saturated State

For this example, we illustrate three different approaches to achieve the initial saturated stress
state. Each approach results in the same distribution for pore pressures, total stresses and effective
stresses. The first approach, a coupled analysis, is a general approach for coupled calculations and
can be used in most problem settings. However, this approach usually requires more calculation
effort and can be quite time-consuming. The second approach is an uncoupled analysis and is
more efficient than the coupled approach. The uncoupled approach is not suitable for all problem
situations; the conditions required for applying an uncoupled analysis are described in Section 1.8.2
in Fluid-Mechanical Interaction. The final approach uses a FISH function to calculate the initial
saturated stress state automatically. This is the most efficient approach, but the function is limited
to horizontally layered media.
1. Coupled Analysis FLAC can perform a coupled analysis in which fluid-flow and
mechanical calculations are performed alternately: first, one flow calculation step is
taken; then one mechanical calculation step is taken; and then the process is repeated. If
the Solve button is pressed, this coupled calculation stepping is performed automatically
until a steady-flow state is achieved.
Beginning with the unsaturated state (EX02.SAV), we turn on the groundwater flow
calculation mode from the Settings / GW tool. We also need to specify the water bulk
modulus and water density. We specify a value of 10,000 Pa for the water bulk modulus;
this low value will speed convergence to steady-state flow. We can do this because we
are not interested in the transient behavior. (Note that there is a lower limit for the water
bulk modulus to satisfy numerical stability see Section 1.4.2.1 in Fluid-Mechanical
Interaction.) We set the water density to 1000 kg/m3 .
For this example, the phreatic surface is at the ground surface. Therefore, we can easily
calculate the steady-state pore-pressure distribution. We initialize the pore pressures
throughout the model to provide this distribution using the In Situ / Initial tool. We click
on the GP Info/Groundwater/pp menu item and then press the Assign button. This opens
the Pore Pressure dialog and allows us to assign the pore-pressure distribution to the
entire model. We input the distribution based on the following equation:*
p = 400, 000 400, 000y
(14.4)
which corresponds to the given gravitational acceleration, water density and model depth.
We enter the value and variation from this equation in the dialog, as shown in Figure 14.14.
* See Eq. (3.2) in the Users Guide for a description of the general equation to assign pressure
gradients in a FLAC model.
FLAC Version 5.0
14 - 15
Figure 14.14 Dialog to enter pore-pressure variation with the

In Situ Initial
/
tool
Whenever pore pressure is changed without performing a flow calculation, the total
stresses in the affected zones must also be adjusted. This is required because an imposed
instantaneous change in pore pressure does not affect the inter-granular forces in a material i.e., the effective stress does not change instantaneously with the pore pressure
change. Therefore, the total stresses must be corrected to satisfy the condition that the
effective stresses do not change. This change to the total stresses is made automatically when the Adjust Tot. Stress box is checked in the Model Options dialog, as shown in
Figure 14.2.
The top boundary of this model is a free surface. The pore pressure is fixed at zero to
satisfy this condition. The pore pressures are also fixed along the side boundaries to
maintain the steady-flow condition.
By default, the saturation = 1 (fully saturated) for this case. We fix the saturation at the
top boundary to ensure that the grid will remain fully saturated during this stage.
The cohesion and tensile strength properties of the soils are set to high values, (the materials are cloned in the Material / Assign tool and re-assigned with high values for cohesion
and tensile strength), and the cohesion and the friction of the interface elements are set to
zero during this stage. This is done to allow a uniform stress gradient to develop during
the coupled calculation.
We are now ready to perform the coupled mechanical-flow calculation. When we click on
Solve
, the Solve options dialog now informs us that a mechanical-groundwater analysis
will be performed. We press Execute , and FLAC begins the calculation automatically,
alternating between one mechanical calculation step and one flow calculation step until
a steady-state flow condition has been reached.
When steady state is achieved (as determined by the equilibrium fluid flow ratio limit of
103 ), we save this state in our project tree with the name EX03.SAV. Figure 14.15
displays the pore pressure contours at steady state.
FLAC Version 5.0
14 - 16
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:40
step
2158
Flow Time
1.2642E+08
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000

0.00E+00
5.00E+04
1.00E+05
1.50E+05
2.00E+05
2.50E+05
3.00E+05
3.50E+05
4.00E+05
2.000
1.000

Boundary plot
0
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.15 Pore pressure contours for initial saturated state

At this stage, the pore pressure, total vertical stress and effective vertical stress distributions are compatible for the change to a fully saturated state. However, the horizontal
stress state must still be adjusted to satisfy the Ko = 0.5 condition. Note that Ko is the
ratio of effective horizontal to vertical stresses. In order to set this condition in the model,
we write a simple FISH function to prescribe the effective horizontal stress components
to all zones in the model as a function of Ko times the effective vertical stress component.
The FISH function is written in the Fish Editor, as shown in Figure 14.16. Note that FISH
only accesses total stress components. Therefore the pore pressure variable is used to
compute the effective stress. k0x and k0z are the ratios of the effective xx to effective
yy , and the effective zz to effective yy , respectively.
FLAC Version 5.0
14 - 17
Figure 14.16 INI HORSTR.FIS FISH function to initialize effective horizontal stresses
We also change the strength properties of the soils and interfaces back to their actual
values and continue the mechanical calculation to determine the final initial stress state.
We use the Material / Assign tool to re-assign the actual clay and sand materials, and we
use the Alter / Interface tool to reset the cohesion and friction angle for the interfaces. We
also turn off the fluid flow calculation and set the fluid bulk modulus to zero using the
Settings GW
/ tool. We set the modulus to zero because we do not wish to generate pore
pressures as a result of mechanical deformation. We solve for the mechanical equilibrium
state with the Solve tool. The saturated stress state determined by this coupled approach
is illustrated by the contour plots of total vertical stress in Figure 14.17, effective vertical
stress in Figure 14.18, and effective horizontal stress in Figure 14.19. We save this state
in our project tree with the name EX04.SAV.
FLAC Version 5.0
14 - 18
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:41
step
3572
Flow Time
1.2642E+08
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
YY-stress contours
-7.50E+05
-6.50E+05
-5.50E+05
-4.50E+05
-3.50E+05
-2.50E+05
-1.50E+05
-5.00E+04
2.000
1.000

Boundary plot
0
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.17 Total vertical stress contours for initial saturated state
coupled analysis
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:41
step
3572
Flow Time
1.2642E+08
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
Effec. SYY-Stress Contours

-3.50E+05
-3.00E+05
-2.50E+05
-2.00E+05
-1.50E+05
-1.00E+05
-5.00E+04
0.00E+00
2.000
1.000

Boundary plot
0
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.18 Effective vertical stress contours for initial saturated state
coupled analysis
FLAC Version 5.0
14 - 19
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:41
step
3572
Flow Time
1.2642E+08
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
Effec. SXX-Stress Contours

-1.75E+05
-1.50E+05
-1.25E+05
-1.00E+05
-7.50E+04
-5.00E+04
-2.50E+04
0.00E+00
2.000
1.000

Boundary plot
0
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.19 Effective horizontal stress contours for initial saturated state
coupled analysis
The total number of calculational steps reported at this stage of the analysis is approximately 3,600. Note that when the coupled analysis stage is being performed, only the
fluid-flow steps are reported; the mechanical steps are considered sub-steps. This is
because the fluid timestep is typically much larger than the mechanical timestep (in other
words, mechanical effects occur instantaneously when compared to fluid diffusion effects), and in this case, the flow step is considered the master step. It is possible that
even one fluid step can put the system considerably out of equilibrium and, hence, several
mechanical sub-steps may be required so that the system remains in equilibrium during
the diffusion process. If we believe that the system is considerably out of equilibrium
during the flow calculation (e.g., by monitoring the unbalanced forces), then we can
increase the number of mechanical sub-steps within each fluid step to keep the system in
equilibrium. This can either be done manually by increasing the number of mechanical
sub-steps (using the Settings / GW tool), or automatically (by clicking on the Auto check box
in the Run / Solve tool), in which case the sub-stepping for both mechanical and fluid steps
are adjusted automatically to keep the maximum unbalanced force ratio below a preset
value.
In our example, the default condition of alternating between one fluid step and one
mechanical step is reasonable because high unbalanced forces are not produced. For
comparison purposes with the other analysis approaches, it should be noted that the total
number of (fluid and mechanical) calculational steps at this stage is approximately 6,000.
FLAC Version 5.0
14 - 20
2. Uncoupled Analysis In an uncoupled analysis, the flow calculation and mechanical

calculation are performed independently. The modeling procedure that we follow here
is identical to that followed for the coupled analysis that we just performed, except that
we now turn off the mechanical-calculation mode before performing the calculation. We
begin again with the unsaturated state (by restoring EX02.SAV). Note that we return
to EX02.SAV by double-clicking on this branch in the project tree. When we begin
performing new operations, a new branch (branch B) is created, and the existing results
are listed under branch A. We right-click on the branch A node and rename the branch to
Saturated - Coupled, and then right click on the branch B node and rename this branch
to Saturated - Uncoupled
We turn off the mechanical calculation mode from the Settings / Mech tool, and turn on the
fluid-flow calculation mode and set water bulk modulus and water density, as we did for
the coupled analysis, from the Settings / GW tool.
We initialize the pore pressures throughout the model by using the In Situ/Initial tool
in the same way as we did previously for the coupled analysis. We click on the GP
Info/Groundwater/pp menu item and then press the Assign button. This opens the Pore
Pressure dialog and allows us to assign the pore-pressure distribution to the entire model.
We assign the pore-pressure distribution as defined by Eq. (14.4). Finally, we use the
In Situ Fix
/
tool to fix the pore pressures along the top and side boundaries, and fix the
saturation along the top boundary, to satisfy the flow conditions, as we did for the coupled
analysis.
We now solve for the steady-flow state ( Solve with fluid-flow only) and a few steps are
taken to satisfy the steady-flow criterion. We save this state as EX05.SAV.
Next, we turn off the groundwater flow calculation and set the water bulk modulus to zero
(using the Settings / GW tool) and turn on the mechanical calculation (using the Settings / Mech
tool). We set the cohesion and tensile strength properties of the soils to high values and
the interface cohesion and friction to zero, as we did for the coupled analysis, in order
to produce a uniform stress gradient. We solve for the new mechanical equilibrium state
using the RUN / Solve tool. We save this state in the project tree as EX06.SAV.
Finally, to complete the uncoupled analysis, we invoke the INI HORSTR.FIS FISH
function to reset the effective horizontal stresses to 0.5 times the effective vertical stresses,
we change the strength properties of the soils and interfaces back to their actual values,
and we continue the mechanical calculation to determine the final initial stress state. The
saturated stress state determined by the uncoupled approach is illustrated by the contour
plots of total vertical stress in Figure 14.20, effective vertical stress in Figure 14.21, and
effective horizontal stress in Figure 14.22. We save this state in our project tree with the
name EX07.SAV. Note that a total of approximately 3600 steps are required to reach
this state.
FLAC Version 5.0
14 - 21
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:43
step
3633
Flow Time
1.2000E+05
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
YY-stress contours
-7.50E+05
-6.50E+05
-5.50E+05
-4.50E+05
-3.50E+05
-2.50E+05
-1.50E+05
-5.00E+04
2.000
1.000

Boundary plot
0
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
uncoupled analysis
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:43
step
3633
Flow Time
1.2000E+05
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000

-3.50E+05
-3.00E+05
-2.50E+05
-2.00E+05
-1.50E+05
-1.00E+05
-5.00E+04
0.00E+00
2.000
1.000

Boundary plot
0
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
uncoupled analysis
FLAC Version 5.0
14 - 22
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:43
step
3633
Flow Time
1.2000E+05
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000

-1.75E+05
-1.50E+05
-1.25E+05
-1.00E+05
-7.50E+04
-5.00E+04
-2.50E+04
0.00E+00
2.000
1.000

Boundary plot
0
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
uncoupled analysis
3. Automatic Initialization with the ININV.FIS FISH Function A FISH function can
be written to automatically calculate the pore pressures and stresses that are compatible
for a model containing a phreatic surface. The calculation can become quite complicated,
however, if there are multiple, non-uniform layers of materials. An example FISH
function that applies for the specific case of horizontally layered media is provided in the
FISH library (see ININV.FIS in Section 3 in the FISH volume). We use this function
here to calculate the saturated stress state directly.
We begin at the unsaturated stress state (EX02.SAV). We set the flow calculation mode
on and specify the fluid bulk modulus and density as before, using the Settings / GW tool.
We now call in the FISH function ININV.FIS from the FISH library by clicking on
the Utility / FishLib tool. We click on the Library/Groundwater/ininv menu item in the
FISH/Library dialog, and press OK . This opens the FISH Call Input dialog, as shown in
Figure 14.23. We enter the phreatic surface height (wth = 40) and the Ko ratios (k0x =
0.5 and k0z = 0.5) in the dialog, and press OK . The FISH function will then be called
into FLAC and executed. The pore pressure distribution and total stress adjustment is
calculated automatically. After the pore pressure distribution is calculated, we use the
In Situ Fix
/
tool to fix the pore pressures along the top and side boundaries, and saturation
along the top boundary, to satisfy the flow conditions, as we did for the coupled and
uncoupled analyses. We now solve for the new equilibrium state, using Run / Solve .
FLAC Version 5.0
14 - 23
The saturated stress state at equilibrium is illustrated by the contour plots of total vertical
stress in Figure 14.24, effective vertical stress in Figure 14.25, and effective horizontal
stress in Figure 14.26. The total step count at this stage is 700 steps (5,300 steps fewer
than the coupled analysis). We save the model at this state in EX08.SAV.
Figure 14.23 FISH library function ININV.FIS input dialog
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:43
step
696
Flow Time
3.8700E+07
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
YY-stress contours
-7.50E+05
-6.50E+05
-5.50E+05
-4.50E+05
-3.50E+05
-2.50E+05
-1.50E+05
-5.00E+04
2.000
1.000

Boundary plot
0
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
automatic initialization with ININV.FIS
FLAC Version 5.0
14 - 24
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:43
step
696
Flow Time
3.8700E+07
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000

-3.50E+05
-3.00E+05
-2.50E+05
-2.00E+05
-1.50E+05
-1.00E+05
-5.00E+04
0.00E+00
2.000
1.000

Boundary plot
0
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:43
step
696
Flow Time
3.8700E+07
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000

-1.75E+05
-1.50E+05
-1.25E+05
-1.00E+05
-7.50E+04
-5.00E+04
-2.50E+04
0.00E+00
2.000
1.000

Boundary plot
0
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
FLAC Version 5.0
14 - 25
By comparing Figures 14.17, 14.20 and 14.24, we see that the total stress results are nearly identical
with less than a 0.2% difference in values. The same is true if we compare effective stress profiles.
The slight difference may be attributed to the normal stiffness of the interfaces. In the first two
approaches, the interfaces have to deform to develop stresses; in the last approach, the stresses are
initialized. In this example, it is obvious that the last approach is the most efficient. However,
for different problem conditions, the coupled or uncoupled approaches may be more suitable. The
guidelines given in Section 1.8.2 in Fluid-Mechanical Interaction should be reviewed to help
select the appropriate analysis approach.
14.2.2.3 Step 3: Install Diaphragm Wall
The next stage of the analysis is the installation of the diaphragm wall. This is simulated by adding
the weight of the wall in the model. We begin at EX08.SAV, and include the weight of the wall
by specifying a mass density for the beam elements. We use the Structure / SEProp tool to enter the
density in the Beam Elements Properties dialog. We also turn off the flow mode and set the water
bulk modulus to zero because we do not wish to generate pore pressures during this stage. We press
Run Solve
/
again to find the equilibrium state with the wall weight included. We save this state as
EX09.SAV. Figure 14.27 shows the total vertical stress distribution at this stage.
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:45
step
4016
Flow Time
3.8700E+07
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
YY-stress contours
-8.00E+05
-7.00E+05
-6.00E+05
-5.00E+05
-4.00E+05
-3.00E+05
-2.00E+05
-1.00E+05
0.00E+00
Boundary plot
0
2.000
1.000
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
weight of wall included
FLAC Version 5.0
14 - 26
14.2.2.4 Step 4: Dewater to a Depth of 20 m

For the dewatering stage, we assume, for simplicity, that the water level is dropped instantaneously
within the excavation region.* We set the saturation and pore pressure to zero using the In Situ / Initial
tool. We click on the GP Info/Groundwater/saturation menu item and drag the mouse over the
gridpoints within the dewatered region (0 x 10, 20 y 40). The affected gridpoints
will be highlighted. We click on Assign to open the dialog to specify a zero saturation value for
these gridpoints. We click OK to create this command, then click on the GP Info/Groundwater/pp
menu item and press Assign . A dialog will open to assign a zero pore pressure to the same region.
Figure 14.28 shows the GwFlow Pore Pressure dialog with the values selected for the variation
from 0 x 10, 20 y 40. The affected region is also highlighted on the model plot.
Figure 14.28 GwFlow Pore Pressure dialog in the
In Situ/Initial
tool
The total stress is adjusted automatically when we impose this change in the pore pressures. This
is a result of selecting the Adjust Tot. Stress box in the Model Options dialog. We can check that
this adjustment to total stress has been made by plotting effective stresses before and after these
commands are issued: the effective stresses are unchanged in the model when the instantaneous
pore pressure change is imposed.
We can now solve for the coupled response that results from the dewatering. In the Settings / GW tool,
we set groundwater flow on, and set the water bulk modulus to 10,000 Pa. This is an unsaturated flow
analysis, so we can use the fast-unsaturated flow scheme to speed the calculation to steady state. (See
* For a more realistic solution, FLAC can calculate the time-dependent evolution of the phreatic
surface and stress state, due to pumping.
FLAC Version 5.0
14 - 27
Section 1.4.1 in Fluid-Mechanical Interaction.) We check the

box to turn on this scheme.
<funsat>Fast unsaturated flow calculation?
We free the saturation condition for gridpoints along the top boundary outside the excavation,
and along the left boundary below the excavation. The region near the ground surface can now
become unsaturated if the water level drops. We initialize the displacements in the model to zero
so that we can monitor the displacement change that occurs due only to the dewatering. Press the
Displmt & Velocity
button in the In Situ / Initial tool to initialize displacements and velocities. Then click
Run/Solve
on
to solve for the equilibrium state with dewatering.
The steady-state pore pressure distribution after dewatering is shown in Figure 14.29. Figure 14.30
plots the vertical displacement contours and displacement vectors at equilibrium. This indicates
the amount of settlement induced by the dewatering. We save the model state as EX10.SAV at
this stage.
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:49
step 11191
Flow Time
4.3050E+08
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000

0.00E+00
5.00E+04
1.00E+05
1.50E+05
2.00E+05
2.50E+05
3.00E+05
3.50E+05
4.00E+05
2.000
1.000

Boundary plot
0
1E 1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.29 Pore pressure distribution following dewatering
FLAC Version 5.0
14 - 28
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:49
step 11191
Flow Time
4.3050E+08
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
-6.00E-02
-5.00E-02
-4.00E-02
-3.00E-02
-2.00E-02
-1.00E-02
2.000

0
1.000
2E -1
0.000

0.500
1.500
2.500
(*10^1)
Figure 14.30 Displacements induced by dewatering
FLAC Version 5.0
3.500
4.500
14 - 29
14.2.2.5 Step 5: Excavate to 2 m Depth

We are now ready to begin the excavation. We set flow off and set the water bulk modulus to zero for
this mechanical-only calculation. We again initialize the displacements, using the Displmt & Velocity
button, in order to evaluate the deformation induced by the excavation.
We use the MaterialAssign tool to perform the excavation. We excavate by assigning the null model
to the material to be removed. We click on the zones in the region 0 x 10, 38 y 40.
These zones are then removed from the model plot, and the corresponding MODEL null commands
are created for sending to FLAC. See Figure 14.31.
We press Run/Solve to calculate the equilibrium state with this first excavation. This is the long-term
response (with water bulk modulus set to zero). We save this state as EX11.SAV.
The deformation induced by this excavation is illustrated in Figure 14.32. We can also calculate
the response of the wall. For example, the moment distribution in the wall after the first excavation
is shown in Figure 14.33. Note that various results for the wall response (e.g., wall displacements,
axial forces, shear forces) can be plotted using the Plot items dialog in the Plot/Model tool.
Figure 14.31 Excavated zones in the
MaterialAssign
tool
FLAC Version 5.0
14 - 30
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:50
step 14529
Flow Time
4.3050E+08
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
-5.00E-03
5.00E-03
1.00E-02
1.50E-02
2.00E-02
2.50E-02
3.00E-02
3.50E-02
2.000
1.000

0
0.000
1E -1

0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.32 Displacements induced by excavation to 2 m depth
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:50
step 14529
Flow Time
4.3050E+08
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
Boundary plot
0
1E 1
2.000
Beam Plot
Moment
on
Structure
Max. Value
# 1 (Beam )
4.056E+05
Beam Plot
1.000
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.33 Moment distribution in wall after excavation to 2 m depth
FLAC Version 5.0
14 - 31
14.2.2.6 Step 6: Install Strut and Excavate to 10 m Depth

For the final excavation step, we install a horizontal strut at the top of the wall and then excavate to
a 10 m depth. We use the StructureBeam tool to install the strut. We press the Add radio button in the
Modes menu, then move the mouse on the model view to one end position of the strut, hold the left
button and move the mouse to the other end position. A line will be drawn indicating the location
of the strut. We can position the strut more precisely by right-clicking the mouse over each end
location. A dialog opens and we enter the endpoint coordinates. For example, Figure 14.34 shows
the Node Parameters dialog that opens when we right-click on the left node. Note that the left node
is free (the Free radio button is active in the dialog). We position the right node at the same location
as the top node of the wall and, consequently, the right node is the existing node of the wall.*
Figure 14.34 Node Parameters dialog in the
StructureBeam
tool
The strut is not rigidly connected to the wall in this exercise. We define a pin connection (which
permits free rotation at the strut/wall connection) by selecting the Pin radio button in the Modes list
and clicking the mouse over the connecting node. An arrow is drawn at the node, denoting this as
a pin connection (see Figure 14.35).
* When the mouse is positioned to create a new node at the same location as an existing node, the
existing node is used to define the end-node of the new structural element. If two separate nodes
are required at the same position, first offset the mouse slightly to create the new node and then
reposition the new node at the same location as the existing node, using the Move mode.
FLAC Version 5.0
14 - 32
Figure 14.35 Selecting a pin connection in the
StructureBeam
tool
We also prescribe a different material property number to the strut in the Beam tool so that we can
assign the strut properties. We click on the PropID radio button in the Modes list, and the identification
number B1 appears over the beam elements in the model plot. We click on the strut element, and a
dialog opens to allow us to rename the property ID to B2.
We now press Execute to send these commands to FLAC to create the strut, pin the strut to the wall,
and assign the property number. Two nodes (1531 and 1532) are created, connected as a single
beam element and assigned property number 1002. A pin connection is defined between node 1532
and wall node 1501.
We enter the Structure / Node tool to assign fixity conditions for the strut. Node 1531 is located along
the centerline of the excavation. We click on this node to open a Node: 1531 dialog, as shown in
Figure 14.36. We fix this node from movement in the x-direction, and from rotating (which are
appropriate conditions for a node located along a line of symmetry), by clicking on the X-velocity
and Rotation check boxes in the dialog. We click OK and then Execute to send the node condition
commands to FLAC.
FLAC Version 5.0
Figure 14.36 Node 1531 dialog in the
14 - 33
StructureNode
tool
We assign the strut properties using the Structure/SEProp tool. We click on the strut element in this
tool and open the Beam Element Properties dialog, as we did previously for the wall properties, to
enter the strut properties as listed in Table 14.3.
We are now ready to perform the second excavation step. We use the MaterialAssign tool and change
the zones within the range 0 x 10, 30 y 38 to null material. We press Run / Solve to
calculate the equilibrium state with this second excavation. We save this state as EX12.SAV.
The total deformation induced by the excavation to the 10 m depth is illustrated in Figure 14.37;
the moment distribution in the wall and axial force in the strut are shown in Figure 14.38.
FLAC Version 5.0
14 - 34
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:53
step 20620
Flow Time
4.3050E+08
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
-2.50E-02
2.50E-02
5.00E-02
7.50E-02
1.00E-01
1.25E-01
1.50E-01
2.000

0
1.000
5E -1
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.37 Displacements induced by excavation to 10 m depth
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 9:53
step 20620
Flow Time
4.3050E+08
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000
Boundary plot
0
1E 1
2.000
Beam Plot
Moment
on
Structure
Max. Value
# 1 (Beam )
1.806E+06
Beam Plot
Axial Force on
Structure
Max. Value
# 2 (Beam )
3.887E+05
1.000
0.000

0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.38 Moment distribution in wall and axial force in strut after excavation to 10 m depth
FLAC Version 5.0
14 - 35
14.2.2.7 Step 7: Determine Factor of Safety at 10 m Depth Excavation

Finally, we wish to determine the factor of safety for the model state at the end of Step 6. After
the EX12.SAV state is saved, we press the SolveFoS button in the Run tool bar. This opens the
Factor-of-Safety parameters dialog to begin the factor-of-safety calculation. We use the default
parameters and assign a failure state save file name of EX12.FSV. We then press OK to begin
the calculation. The calculated factor of safety is 2.96. Figure 14.39 displays a failure-state plot,
containing shear-strain contours and velocity vectors, which defines the failure surface. Note that
the factor-of-safety calculation can be made at any stage in the simulation.
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.000
LEGEND
14-Jan-05 11:12
step 43384
Flow Time
4.3050E+08
-4.167E+00 <x< 4.917E+01
-6.667E+00 <y< 4.667E+01
3.000

Max. shear strain-rate
0.00E+00
5.00E-06
1.00E-05
1.50E-05
2.00E-05
2.50E-05
3.00E-05
3.50E-05
4.00E-05
4.50E-05
2.000
1.000

Velocity vectors
0.000
0
2E -4
0.500
1.500
2.500
3.500
4.500
(*10^1)
Figure 14.39 Failure surface plot for Step 6; failure surface defined by shearstrain contours and velocity vectors
FLAC Version 5.0
14 - 36
14.3 Data File EXCAVATE.DAT

;Title:Braced Excavation
;... State: ex01.sav ....
config gwflow ats
grid 46,40
gen 0.0,0.0 0.0,10.0 10.0,10.0 10.0,0.0 i 1 11 j 1 11
gen 10.0,0.0 10.0,10.0 45.0,10.0 45.0,0.0 i 12 47 j 1 11
gen 0.0,10.0 0.0,40.0 10.0,40.0 10.0,10.0 i 1 11 j 11 41
gen 10.0,10.0 10.0,40.0 45.0,40.0 45.0,10.0 i 12 47 j 11 41
struct node 1501 10.0 40.0
struct node 1500 10.0 10.0
struct beam beg node 1500 end node 1501 prop 1001 seg 30
struct prop 1001
int 101 as from 11,11 to 11,41 bs from node 1500 to node 1501
int 102 as from 12,41 to 12,11 bs from node 1501 to node 1500
group Excavation:sand notnull j 1 20
model mohr notnull group Excavation:sand
friction=32.0 dilation=2.0 tension=0.0 notnull group Excavation:sand
group Excavation:clay notnull j 21 40
model mohr notnull group Excavation:clay
prop density=1600.0 bulk=1.1111111E7 shear=3703703.8 cohesion=5000.0 &
friction=25.0 dilation=0.0 tension=0.0 notnull group Excavation:clay
struct prop 1001 e 5.95E9 area 1.26 I 0.167
prop por=0.3 perm=1.0E-10 notnull
fix x y j 1
fix x i 47
fix x i 1
save ex01.sav
set gravity=10.0
set flow=off
initial syy -660000.0 var 0.0,340000.0 j 1 20
FLAC Version 5.0
initial syy -320000.0 var

initial sxx -330000.0 var
initial sxx -160000.0 var
initial szz -330000.0 var
initial szz -160000.0 var
solve elastic
save ex02.sav
0.0,320000.0
0.0,170000.0
0.0,160000.0
0.0,170000.0
0.0,160000.0
14 - 37
j
j
j
j
j
21 40
1 20
21 40
1 20
21 40
;*** Branch: Saturated - coupled ****

set flow=on
water bulk=10000.0
group Excavation:sand - high strength notnull j 1 20
model mohr notnull group Excavation:sand - high strength
prop density=1700.0 bulk=3.3333336E7 shear=1.5384616E7 cohesion=1.0E10 &
friction=32.0 dilation=2.0 tension=1.0E10 notnull group &
Excavation:sand - high strength
group Excavation:clay - high strength notnull j 21 40
model mohr notnull group Excavation:clay - high strength
prop density=1600.0 bulk=1.1111111E7 shear=3703703.8 cohesion=1.0E10 &
Excavation:clay - high strength
initial pp 400000.0 var 0.0,-400000.0
fix pp j 41
fix pp i 1
fix pp i 47
fix saturation j 41
solve
save ex03.sav
set echo off
call ini horstr.fis
set k0x=0.5 k0z=0.5
ini horstr
FLAC Version 5.0
14 - 38

set flow=off
water bulk=0.0
solve
save ex04.sav
;*** Branch: Saturated - uncoupled ****
restore ex02.sav
set mechanical=off
set flow=on
water bulk=10000.0
initial pp 400000.0 var 0.0,-400000.0
fix pp j 41
fix pp i 1
fix pp i 47
fix saturation j 41
solve
save ex05.sav
set flow=off
water bulk=0.0
set mechanical=on
group Excavation:sand - high strength notnull j 1 20
model mohr notnull group Excavation:sand - high strength
prop density=1700.0 bulk=3.3333336E7 shear=1.5384616E7 cohesion=1.0E10 &
Excavation:sand - high strength
group Excavation:clay - high strength notnull j 21 40
model mohr notnull group Excavation:clay - high strength
prop density=1600.0 bulk=1.1111111E7 shear=3703703.8 cohesion=1.0E10 &
Excavation:clay - high strength
FLAC Version 5.0
14 - 39

solve
save ex06.sav
set echo off
call ini horstr.fis
set k0x=0.5 k0z=0.5
ini horstr
solve
save ex07.sav
;*** Branch: Saturated - ININV.FIS ****
restore ex02.sav
set flow=on
water bulk=10000.0
set echo off
call Ininv.fis
set wth=40 k0x=0.5 k0z=0.5
ininv
fix pp j 41
fix pp i 1
fix pp i 47
fix saturation j 41
solve
save ex08.sav
struct prop 1001 density 2000.0
set flow=off
water bulk=0.0
FLAC Version 5.0
14 - 40
solve
save ex09.sav
initial pp 0.0 i 1 11 j 21 41
fix saturation i 1 11 j 21 41
free pp i 1 j 1 20
free pp i 13 46 j 41
free saturation i 13 46 j 41
set flow=on
water bulk=10000.0
history 1 gpp i=3, j=19
history 5 gwtime
set gwtime=0.0
set funsat=on
solve
save ex10.sav
set flow=off
water bulk=0.0
group delete null
solve
save ex11.sav
group delete null
struct node 1532 10.0,40.0 pin slave x y 1501
struct node 1531 fix x r
struct prop 1002 density 3000.0 e 4.0E9 area 1.0 I 0.083
struct chprop 1002 range 31 31
solve
FLAC Version 5.0
14 - 41
save ex12.sav
;... State: ex12.fsv ....
;FoS save state:
solve fos no restore file ex12.fsv
;plot name: total syy contours
plot hold syy fill int 50000.0 inv bound
;plot name: effective syy contours
plot hold esyy fill inv bound
;plot name: effective sxx contours
plot hold esxx fill inv bound
;plot name: pp contours
;plot name: structure moment & force
plot hold bound struct beam moment 1 fill white struct beam axial 2 fill &
lcyan
;plot name: displacements
plot hold ydisp fill zero displacement bound
;plot name: FOS
plot hold fos ssr fill velocity bound
;plot name: pp histories
plot hold history 1 line 2 line 3 line 4 line vs 5
FLAC Version 5.0
14 - 42
FLAC Version 5.0
Earthquake Loading of a Pile-Supported Wharf
15 - 1
15 Earthquake Loading of a Pile-Supported Wharf

A seismic hazard concern in the design of pile-supported wharves at port waterfronts is the structural
stability of the wharf if earthquake-induced liquefaction occurs in the soils supporting the piles. The
analysis of this type of problem is demonstrated using FLAC with the dynamic analysis option and
liquefaction modeling facility. Calculations can be made with FLAC for both the deformation of
the liquefiable soils and the displacements of the wharf structure that are induced by the earthquake
motion. It is also possible to monitor various problem conditions during the seismic excitation
including the development of excess pore pressures in the soils and moments in the piles.
This example analysis illustrates the recommended procedure to simulate this problem with FLAC.
The analysis is divided into three stages.
Stage 1: Determine the initial equilibrium state of the waterfront soils.
Stage 2: Determine the equilibrium state after the wharf is constructed.
Stage 3: Apply the earthquake motion and monitor the wharf and soil response during the shaking
period.
Figure 15.1 shows the problem conditions. The waterfront soils in this exercise consist of three
layered deposits, denoted Soils 1, 2 and 3: the upper two layers (Soils 2 and 3) are characterized as
liquefiable silty sands. The thicknesses of the Soil 1, Soil 2 and Soil 3 layers are 11 m, 6.5 m and
4.5 m, respectively. Soil 1 overlies a strong, stiff bedrock.
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.500
LEGEND
22-Sep-04 11:05
3.500
-4.500E+00 <x< 8.550E+01

-3.400E+01 <y< 5.600E+01
2.500
User-defined Groups
wharf:soil1
wharf:soil2
wharf:soil3
wharf
1.500
piles
0.500
Boundary plot
0
2E 1
-0.500
-1.500
-2.500

0.500
1.500
2.500
3.500
4.500
(*10^1)
5.500
6.500
7.500
Figure 15.1 Pile-supported wharf on layered embankment
FLAC Version 5.0
15 - 2
The wharf is constructed on a waterfront embankment that is 11 m high and has a slope angle of
approximately 27 degrees. The wharf is supported by two rows of piles that are 3 m apart and 16
m in length. The piles extend through Soils 2 and 3 and into Soil 1, as shown in Figure 15.1. Each
row of piles has a spacing of 2 m along the length of the wharf.
The following (drained) material properties are assigned to the soils.
Table 15.1 Drained properties for Soils 1, 2 and 3
Dry density (kg/m3 )
Youngs Modulus (MPa)
Poissons ratio
Cohesion (Pa)
Friction angle (degrees)
Soil 1
Soil 2
Soil 3
2009
610.9
0.3
4000
40
0
1813
163.7
0.3
1000
35
0
1715
163.7
0.3
1000
30
0
The liquefaction condition is estimated for the upper two layers, Soils 2 and 3, in terms of Standard
Penetration Test results. A normalized Standard Penetration Test value, (N1 )60 , of 20 is selected
as representative for Soil 2 and Soil 3. This value is used to determine the parameters C1 , C2
and C3 in the liquefaction model in FLAC (selected by setting the property ff switch = 1 for
the Finn-Byrne model). For a normalized SPT blow count of 20, the Byrne model parameters are
C1 = 0.2062, C2 = 1.940 and C3 = 0.0. See Byrne (1991) for a discussion on the derivation of these
parameters.
The dynamic characteristics of all of the soils in this model are assumed to be governed by the
modulus reduction factor (G/Gmax ) and damping ratio () curves, as shown in Figures 15.2 and
15.3, and denoted by the Shake91 legend. These curves are considered to be representative of
sandy soils with an average mass density of 1800 kg/m3 , and an average shear modulus of 150
MPa; the data are derived from the input file supplied with SHAKE91 (for more information see
http://nisee.berkeley.edu/software/shake91/).
FLAC Version 5.0
15 - 3
Figure 15.2 Modulus reduction curve for sandy soils (from SHAKE91 data)
Figure 15.3 Damping ratio curve for sandy soils (from SHAKE91 data)
FLAC Version 5.0
15 - 4
The structural properties for the wharf are listed in Tables 15.2 and 15.3. The properties listed in
Table 15.2 are assigned to the wharf beam and pile elements, and the properties listed in Table 15.3
are assigned to represent the behavior at the pile-soil interface.
Table 15.2 Structural properties for wharf

Elastic
Modulus
(GPa)
Moment of
Inertia
(m4 )
Cross Sect.
Area
(m2 )
Mass
Density
(kg/m2 )
Pile
Perimeter
(m)
2.0
2.0
2.364103
1.302106
0.305
0.025
2000
2000
0.63
Beams
Piles
Table 15.3 Coupling spring properties for pile-soil interface
Soil 1
Soils 2 & 3
Normal
Stiffness
(GPa/m)
Shear
Stiffness
(GPa/m)
Normal
Cohesion
(Pa/m)
Shear
Cohesion
(Pa/m)
Normal
Friction
(degrees)
Shear
Friction
(degrees)
1.0
1.0
1.0
1.0
4000
1000
4000
1000
40
30
40
30
The wharf is subjected to an earthquake motion with a peak acceleration of approximately 0.3 g
and duration of 40 sec. Figure 15.4 shows the acceleration time history. This history is assumed to
be recorded near the wharf site. A reduction factor of 0.5 is applied to the acceleration to correlate
the motion to that at a depth of 11 m beneath the toe of the waterfront embankment.* A Fast
Fourier Transform analysis of the acceleration record (using FFT.FIS in Section 3 in the FISH
volume) results in a power spectrum as shown in Figure 15.5. This figure indicates that the highest
frequency is less than approximately 15 Hz, and that the majority of the frequencies are less than
10 Hz. The data file INPUT.DAT listed in Section 15.5 contains the commands to generate the
power spectrum.
* Note that a FLAC or SHAKE analysis can be used to determine an input acceleration at the depth it
is applied in a model, accounting for propagation of the wave from the location where it is recorded.
FLAC Version 5.0
15 - 5
JOB TITLE : .
FLAC (Version 5.00)

(10
-01
LEGEND
10-Jun-04 11:46
step
0
Acceleration Record
(gs vs sec)
2.000
1.500
1.000
0.500
0.000
-0.500
-1.000
-1.500
-2.000
-2.500
10
15
20
25
30
35

Figure 15.4 Horizontal acceleration time history
JOB TITLE : .
FLAC (Version 5.00)

(10
-02
LEGEND
10-Jun-04 11:46
step
0
1.000
Power Spectrum
(power vs frequency in Hz)
0.800
0.600
0.400
0.200
10
15
20
25

Figure 15.5 Power spectrum of input acceleration
FLAC Version 5.0
15 - 6

The model is created using FLAC s graphical interface, the GIIC. Upon entering the GIIC, the
dynamic option, groundwater flow and adjust total stress options, structural elements and advanced
constitutive models facilities are activated, and two extra grid variables are selected in the FLAC
Options dialog. A project file, WHARF.PRJ, is created to save the model state at various stages
of the simulation. (We click on ? in the Project File dialog to select a directory in which to save
the project files.) The input commands to create and solve this problem are listed in the data file
WHARF.DAT, given in Section 15.4.
The problem geometry is created in FLAC using the Build/Library tool for a simple slope, uniform
grid. Figure 15.6 shows the grid created by this tool after the mesh has been manipulated to fit
the given geometry. The figure also shows the marked gridpoints, created with the Alter/Mark tool to
define the boundaries of the three soil layers.
The dynamic calculation phase is performed using the large-strain mode in FLAC. By using
the selected meshing tool, the model will consist of only quadrilateral-shaped zones. This will
help prevent the development of badly distorted zones along the slope face during the large-strain
calculation.
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.500
LEGEND
22-Sep-04 10:25
step
1993
-4.500E+00 <x< 8.550E+01
-3.400E+01 <y< 5.600E+01
3.500
2.500
User-defined Groups
wharf:soil1
wharf:soil2
wharf:soil3
Grid plot
0
1.500
2E 1
0.500
Marked Gridpoints
-0.500
-1.500
-2.500

0.500
1.500
2.500
3.500
4.500
(*10^1)
5.500
6.500
7.500
Figure 15.6 FLAC model of layered embankment

The mesh size for the model should be selected to provide accurate wave transmission. Based upon
the elastic properties listed in Table 15.1, Soil 2 has the lowest shear wave speed (186 m/sec). If
the largest zone in the model is set to approximately 1.5 m (in order to provide reasonable runtimes
for this example), then the maximum frequency that can be modeled accurately is
FLAC Version 5.0
f =
15 - 7
Cs
12 Hz
10 l
(15.1)
Before applying the seismic record, it is filtered to remove high frequencies. The FISH function
FILTER.FIS, described in Section 3 in the FISH volume, is used to remove frequencies above
10 Hz (which is selected slightly lower than the value in Eq. (15.1) to account for reduction in shear
wave speed that may occur in the liquefiable materials during the dynamic loading). The acceleration
history, converted to m/sec2 versus seconds, and filtered at 10 Hz, is shown in Figure 15.7, and the
power spectrum for the filtered wave is shown in Figure 15.8. Note that the original acceleration
record is converted to m/sec2 versus seconds, using the FISH function CONVERT.FIS, before
filtering. This is the form of the record that will be applied as the dynamic input at the base of the
model.
The properties in Table 15.1 are entered into a material database by clicking the Materials button in
the lower-right corner of the Material/Assign tool. The three material types, Soil1, Soil2 and Soil3, are
created in a material class named wharf, and their properties are assigned by editing the dialog for
each material. The materials are then stored in a separate database file, named WHARF.GMT,
which can be accessed at any time in subsequent analyses. The three materials are made available
for the present model by clicking the OK button in the Material List dialog; the materials will then
be listed when the Material radio button is pressed in the Material/Assign tool. By clicking on the Region
radio button, then highlighting each material and clicking on one zone in each of the three regions of
the model plot, the selected region of zones will change color, corresponding to that of the selected
material. Once all three materials have been assigned, the Execute button is pressed to send the
commands to FLAC. The resulting model with the assigned materials is shown in Figure 15.6.
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
23-Sep-04 10:16
step
0
Acceleration Record
(m/sec2 vs sec)
2.000
1.500
1.000
0.500
0.000
-0.500
-1.000
-1.500
-2.000
-2.500
10
15
20
25
30
35

Figure 15.7 Horizontal acceleration time history with 10 Hz filter
FLAC Version 5.0
15 - 8
JOB TITLE : .
FLAC (Version 5.00)

(10
-02
LEGEND
23-Sep-04 10:16
step
0
1.000
Power Spectrum
0.800
0.600
0.400
0.200
10
15
20
25

Figure 15.8 Power spectrum of input acceleration with 10 Hz filter

Stage 1
Stage 1 of the analysis is performed in two steps. First, the model is brought to an equilibrium
state, assuming dry material conditions. After assigning boundary conditions via the InSitu/Fix tool,
gravity is prescribed in the Settings/Mechanical tool, groundwater flow is turned off in the Settings/GW
tool, and the dynamic analysis mode is turned off in the Settings/Dynamic tool. The model is then
brought to a static mechanical-equilibrium state by selecting the Run/Solve tool and then checking
the Solve initial equilibrium as elastic model box.
Next, the water level is raised and the equilibrium state of the submerged embankment is calculated. The steady-state flow condition is achieved by applying a pore-pressure gradient through
the InSitu/Apply and InSitu/Initial tools. The gradient corresponds to a water level at y = 20 m. After
assigning groundwater properties through the Material/Assign tool, groundwater flow is turned on and
the fluid density is assigned in the Settings/GW tool, the mechanical calculation is turned off in the
Settings/Mechanical
tool, and the model is solved to ensure that the model is at steady-state flow. The
pressure of the reservoir water above the embankment is included via the InSitu/Apply tool. Then,
the groundwater flow mode is turned off, the mechanical calculation is turned on, and the model is
solved for the mechanical equilibrium state of the submerged embankment.
FLAC Version 5.0
15 - 9
Stage 2
The wharf structure is added in Stage 2, using the Structure/Beam tool to create the wharf deck and the
Structure/Pile
tool to create the supporting piles. Note that each pile is divided into 16 segments. This
ensures that at least one pile node is located within each zone along the length of the pile. The wharf
beam and piles share the same nodes at their intersection. This provides a rigid connection between
the wharf and piles. The structural properties for the wharf structure, as listed in Tables 15.2 and
15.3, are specified using the Structure/Prop tool. Note that the 2 m spacing is entered with the pile
properties. The different pile-soil interface properties listed in Table 15.3 are assigned by specifying
a structural property ID number for the pile segments within Soil 1 different from that for segments
within Soils 2 and 3.
The model is brought to an equilibrium state with the wharf in place. Figure 15.9 shows the model
geometry with the wharf structure. The initial pore pressure contours are also plotted.
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.500
LEGEND
13-Oct-04 13:26
step
6759
Flow Time
3.2501E+06
-4.500E+00 <x< 8.550E+01
-3.400E+01 <y< 5.600E+01
3.500
2.500

0.00E+00
2.50E+04
5.00E+04
7.50E+04
1.00E+05
1.25E+05
1.50E+05
1.75E+05
2.00E+05
1.500
0.500
-0.500

Boundary plot
0
-1.500
2E 1
wharf
-2.500
piles
0.500
1.500
2.500
3.500
4.500
(*10^1)
5.500
6.500
7.500
Figure 15.9 Pore-pressure contours at equilibrium state, incl. wharf structure
FLAC Version 5.0
15 - 10
Stage 3
For Stage 3, it is necessary to specify a realistic value for the water bulk modulus in order to evaluate
the liquefaction condition of the embankment. To avoid excessive runtimes, this value is selected
using the formula
4
Kw = 20n(K + G)
3
(15.2)
where n is the porosity, and K and G are the bulk and shear moduli of the unsaturated soil. For this
problem, Kw is calculated to be approximately 1.0 GPa. (Note that the actual water bulk modulus
of pure water is 2.0 GPa.)
We begin the dynamic loading stage by turning on the dynamic calculation mode from the Settings/Dyna
tool. The original acceleration record (shown in Figure 15.4) is designated as ACC1.HIS, and
is loaded into FLAC using the Read button in the Utility/History tool. The Execute button should be
pressed to execute the command and load the record into FLAC. The record is converted into m/sec2
versus seconds via CONVERT.FIS, which is executed from the Fish Editor. The record is then
filtered at 10 Hz using FILTER.FIS, which is executed from the Utility/FishLib tool. The filtered
record is stored in table 111.
We monitor histories of selected model variables during the dynamic calculation. These are chosen
using the Utility/History tool. In particular, pore pressures are monitored at two locations: at the toe
of the slope; and in the interior of the embankment. These histories, along with the dynamic time,
are recorded to evaluate the potential for generation of pore pressure (and liquefaction) during the
dynamic calculation.
The dynamic boundary conditions are now assigned in the In Situ/Apply tool. The free-field boundary
is set for the side boundaries by selecting the Free-Field button. The filtered acceleration record is
applied along the bottom boundary by first selecting the Dynamic/xacc boundary condition type and
then dragging the mouse along the bottom boundary. The Apply value dialog opens when the
Assign
button is pressed, an x-acceleration value of 0.5 is entered as a reduction factor for the input
acceleration, and the record in table 111 is applied by selecting the Multiplier/Table button. The
acceleration record will now be multiplied by 0.5 and applied along the bottom boundary when the
OK
button is pressed. The factor of 0.5 corresponds to the adjusted acceleration at the bottom of the
model.
The bottom of the model is fixed from movement in the y-direction to represent the presence of
the bedrock base. This boundary condition is assigned by selecting the Velocity/yvelocity boundary
condition type and dragging the mouse along the bottom boundary. The Apply value dialog opens
when the Assign button is pressed, and a y-velocity value of 0.0 is entered to fix the boundary from
movement in the y-direction.
Before performing the liquefaction calculation, runs are made with Mohr-Coulomb material assigned to all three soils in order to evaluate the amount of additional mechanical damping required
for the wharf model. We first determine the dominant natural frequencies associated with this
model; these are required for Rayleigh damping. An undamped run is performed for a dynamic
FLAC Version 5.0
15 - 11
loading time period of 10 seconds. Velocities are monitored at gridpoints in the different materials
and at structural element nodes. Fast Fourier Transform analyses are then performed on these
velocity records in order to determine the dominant frequencies. Figure 15.10 presents one FFT
result recorded at a gridpoint in Soil 2. For this example, all of the recorded velocities in the grid
and wharf structure indicate that the dominant frequency is approximately 0.8 Hz, as shown in Figure 15.10. In general, values may vary in different materials, requiring different Rayleigh damping
parameters for different regions in a model.
JOB TITLE : .
FLAC (Version 5.00)

(10
-03
LEGEND
7-Oct-04 4:26
step 136381
Flow Time
3.2501E+06
Dynamic Time 1.0000E+01
Table Plot
Table 206
1.000
0.800
0.600
0.400
0.200
10
12
14
16
18

Figure 15.10 Power spectrum for x-velocity at gridpoint (31,10) in soil 2

Three dynamic simulations are now made with different types of damping:
1. Rayleigh damping applied for both the grid and the wharf structure;
2. Rayleigh damping applied for the grid; and
3. hysteretic damping applied for the grid.
For the first two damping cases, a center frequency of 0.8 Hz, identified from the undamped run,
is specified for the Rayleigh damping. The fraction of critical damping is determined by running
simple element tests with the different soil types for the dynamic loading as defined in Figure 15.7.
The fraction of critical damping is adjusted in the simple element tests to approximate the cyclic
strain response for sandy soils, as defined by the SHAKE91 data in Figures 15.2 and 15.3. The
maximum cyclic strain is roughly 0.002% for this loading, and the fraction of critical damping is
estimated to be 0.01.
FLAC Version 5.0
15 - 12
The same Rayleigh damping (mass- and stiffness-proportional damping of 1% at the dominant
frequency of 0.8 Hz) is applied for the wharf structure and the grid in the first damping case.
Rayleigh damping is assigned in the Settings/Dyna tool. A dynamic run is made for 5 seconds while
monitoring the displacements of the grid and wharf structure. The limiting timestep for this case is
approximately 3.7 106 seconds.
The dynamic run of 5 seconds is repeated for the second damping case, with Rayleigh damping
applied only for the grid. The displacement results are nearly identical to the first case; this time
the timestep is approximately 2.1 105 seconds. The agreement between the two cases indicates
that it is not necessary to include structural damping in this analysis. Damping for the grid alone is
sufficient to damp the natural oscillation modes identified from the undamped simulation.
In the third damping case, hysteretic damping corresponding to the dynamic characteristics represented by the (G/Gmax ) and () curves shown in Figures 15.2 and 15.3 is applied. These figures
show a comparison of the (G/Gmax ) and () variations to those computed using the default hysteretic model in FLAC. The selected parameters (L1 = -3.325 and L2 = 0.823) for the default model
produce the FLAC curves shown in these figures. Hysteretic damping is assigned in the In Situ/Initial
tool. Hysteretic damping does not completely damp high frequency components, so a small amount
of stiffness-proportional Rayleigh damping is also applied, at 0.5% at the dominant frequency (0.8
Hz). The limiting timestep for the hysteretic damping case is approximately 4.0 105 seconds.
The results from the hysteretic damping case and the Rayleigh damping case are comparable,
as indicated by the plots in Figures 15.11 through 15.14. Rayleigh damping is shown to induce
somewhat greater material damping over the given range of cyclic strains, as indicated by the smaller
displacements and lower pile moments for the Rayleigh-damping case compared to the hystereticdamping case. In both cases, though, the wharf structure and slope are stable, and motion essentially
ceases after roughly 20 seconds of shaking.
FLAC Version 5.0
15 - 13
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
13-Oct-04 15:44
step 500413
Flow Time
3.2501E+06
0.000
-0.400
HISTORY PLOT
Y-axis :
X Displacement (Nd
-0.800
1)
X-axis :
Dynamic time
-1.200
-1.600
-2.000
10
12
14
16
18
20

Figure 15.11 x-displacements at gridpoint (17,14) and pile node (node 1)

Mohr-Coulomb material and hysteretic damping
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
13-Oct-04 11:49
step 959337
Flow Time
3.2501E+06
0.000
-0.400
HISTORY PLOT
Y-axis :
X Displacement (Nd
-0.800
1)
X-axis :
Dynamic time
-1.200
-1.600
-2.000
10
12
14
16
18
20


Mohr-Coulomb material and Rayleigh damping
FLAC Version 5.0
15 - 14
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.500
LEGEND
13-Oct-04 15:44
step 500413
Flow Time
3.2501E+06
-4.500E+00 <x< 8.555E+01
-3.400E+01 <y< 5.605E+01
3.500
2.500

0.00E+00
5.00E-02
1.00E-01
1.50E-01
2.00E-01
1.500
0.500

Pile Plot
Moment
Structure
# 2 (Pile )
# 3 (Pile )
# 4 (Pile )
# 5 (Pile )
wharf
-0.500
on
Max. Value
5.084E+04
5.084E+04
-5.422E+04
4.644E+04
-1.500
-2.500

0.500
1.500
2.500
3.500
4.500
(*10^1)
5.500
6.500
7.500
Figure 15.13 Shear-strain increment contours and pile moments at 20 seconds

JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.500
LEGEND
13-Oct-04 11:49
step 959337
Flow Time
3.2501E+06
-4.500E+00 <x< 8.554E+01
-3.400E+01 <y< 5.604E+01
3.500
2.500

0.00E+00
2.50E-02
5.00E-02
7.50E-02
1.00E-01
1.500
0.500

Pile Plot
Moment
Structure
# 2 (Pile )
# 3 (Pile )
# 4 (Pile )
# 5 (Pile )
wharf
-0.500
on
Max. Value
3.706E+04
3.703E+04
-3.301E+04
2.893E+04
-1.500
-2.500

0.500
1.500
2.500
3.500
4.500
(*10^1)
5.500
6.500
7.500
Figure 15.14 Shear-strain increment contours and pile moments at 20 seconds

FLAC Version 5.0
15 - 15
The liquefaction simulation is performed by changing Soils 2 and 3 to Finn model materials, using
the Material/Model tool. The Byrne (1991) liquefaction model is assigned to these soils, and properties
are prescribed corresponding to a normalized SPT blow count of 20. For example, Figure 15.15
displays the dialog to enter properties for Soil 2. Note that the latency property is set to a high value
at this stage. This is done to check that the model is still at equilibrium when changing Soils 2 and
3 from Mohr-Coulomb to Finn material. When Run/Solve is issued, only a few steps are taken, which
ensures the model is still in equilibrium.
Figure 15.15 Model finn properties dialog w/ properties for soil 2

The dynamic calculation is now run in the same manner as described previously. The latency value
is set to 50, and hysteretic damping is imposed. Note that the free-field boundary condition must
be applied after these changes are made, to ensure that these changes to the grid are transferred to
the free field.
This time the embankment slope fails as a result of the earthquake motion. The progressive development of the slope failure is displayed by the plots in Figures 15.16 and 15.17. At 9.35 seconds
of shaking (the state shown in Figure 15.17), the run stops because the geometry distortion reaches
the bad geometry limit.
There is a pronounced increase in pore pressure at the measurement location inside the slope (20,8),
as shown in Figure 15.18. As a comparison, the pore pressure history plot for the non-liquefiable
(Mohr-Coulomb material) run is shown in Figure 15.19. The pore-pressure generation is evident
when the Finn model is used. Failure of the slope is also indicated by the displacement history plot
given in Figure 15.20. Compare this result to that in Figure 15.11.
FLAC Version 5.0
15 - 16
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.500
LEGEND
14-Oct-04 16:34
step 75218
3.500

-4.520E+00 <x< 8.551E+01
-3.402E+01 <y< 5.600E+01
2.500

0.00E+00
1.50E-01
3.00E-01
4.50E-01
6.00E-01
7.50E-01
1.500
0.500
-0.500

Pile Plot
Moment
on
Structure
Max. Value
# 2 (Pile )
8.667E+04
# 3 (Pile )
8.885E+04
# 4 (Pile ) -1.026E+05
# 5 (Pile )
9.192E+04
wharf
-1.500
-2.500
0.500
1.500
2.500
3.500
4.500
(*10^1)
5.500
6.500
7.500
Figure 15.16 Shear-strain increment contours and pile moments

at 5 seconds soils 2 and 3 can liquefy
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

4.500
LEGEND
15-Oct-04 9:57
step 143410
3.500

-4.494E+00 <x< 8.554E+01
-3.400E+01 <y< 5.603E+01
2.500

0.00E+00
5.00E-01
1.00E+00
1.50E+00
2.00E+00
2.50E+00
1.500
0.500
-0.500

Pile Plot
Moment
on
Structure
Max. Value
# 2 (Pile )
1.848E+05
# 3 (Pile )
1.764E+05
# 4 (Pile ) -1.604E+05
# 5 (Pile )
1.186E+05
wharf
-1.500
-2.500
0.500
1.500
2.500
3.500
4.500
(*10^1)
5.500
6.500
7.500
Figure 15.17 Shear-strain increment contours and pile moments

at 9.35 seconds soils 2 and 3 can liquefy
FLAC Version 5.0
15 - 17
JOB TITLE : .
FLAC (Version 5.00)

(10
05
LEGEND
14-Oct-04 16:34
step 75218
Flow Time
3.2501E+06
1.000
HISTORY PLOT
Y-axis :
0.950
0.900
X-axis :
Dynamic time
0.850
0.800
10
15
20
25
30
35
40
45
(10
-01

Figure 15.18 Pore pressure histories at toe of slope in soil 1, zone (9,6), and
within slope in soil 2, zone (20,8) soils 2 and 3 can liquefy
JOB TITLE : .
FLAC (Version 5.00)

(10
05
LEGEND
13-Oct-04 15:05
step 129980
Flow Time
3.2501E+06
HISTORY PLOT
Y-axis :
X-axis :
Dynamic time
1.100
1.050
1.000
0.950
0.900
0.850
0.800
0.750
0.700
0.650
10
15
20
25
30
35
40
45
50
(10
-01

Figure 15.19 Pore pressure histories at toe of slope in soil 1, zone (9,6), and
within slope in soil 2, zone (20,8) soils 2 and 3 cannot liquefy
FLAC Version 5.0
15 - 18
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
14-Oct-04 16:34
step 75218
Flow Time
3.2501E+06
0.000
-0.400
HISTORY PLOT
Y-axis :
X Displacement (Nd
-0.800
1)
X-axis :
Dynamic time
-1.200
-1.600
-2.000
10
15
20
25
30
35
40
45
(10
-01


soils 2 and 3 can liquefy
15.3 References
Byrne, P. A Cyclic Shear-Volume Coupling and Pore-Pressure Model for Sand, in Proceedings:
Second International Conference on Recent Advances in Geotechnical Earthquake Engineering
and Soil Dynamics (St. Louis, Missouri, March, 1991), Paper No. 1.24, 47-55.
FLAC Version 5.0
15 - 19
15.4 Data File WHARF.DAT

;Title:Earthquake Loading of Wharf
;... STATE: WF1 ....
config gwflow ats dynamic extra 2
grid 40,18
gen 0.0,0.0 0.0,11.0 26.73,11.0 26.73,0.0 i 1 17 j 1 7
gen 26.73,0.0 26.73,11.0 54.27,11.0 54.27,0.0 i 17 29 j 1 7
gen 54.27,0.0 54.27,11.0 81.0,11.0 81.0,0.0 i 29 41 j 1 7
gen 26.73,11.0 48.0,22.0 68.0,22.0 54.27,11.0 i 17 29 j 7 19
gen 54.27,11.0 68.0,22.0 81.0,22.0 81.0,11.0 i 29 41 j 7 19
save wf1.sav
;... STATE: WF2 ....
mark i 17 41 j 7
mark i 17 41 j 14
group wharf:soil1 region 33 2
model mohr notnull group wharf:soil1
friction=40.0 dilation=0.0 tension=0.0 notnull group wharf:soil1
fix x y j 1
fix x i 41
fix x i 1 j 1 7
set gravity=10.0
set flow=off
set dyn=off
solve elastic
save wf2.sav
;... STATE: WF3 ....
FLAC Version 5.0
15 - 20
prop por=0.3 perm=1.0E-10 notnull

apply pp 200000.0 var 0.0 -200000.0 from 1,1 to 17,17
apply pp 200000.0 var 0.0 -200000.0 from 41,1 to 41,17
initial pp 200000.0 var 0.0,-200000.0 j 1 17
set mechanical=off
set flow=on
water bulk=10000.0
solve
save wf3.sav
;... STATE: WF4 ....
apply pressure 90000.0 var 0.0 -90000.0 from 1,7 to 17,17
set flow=off
water bulk=0.0
set mechanical=on
solve
save wf4.sav
;... STATE: WF5 ....
struct prop 1001
struct pile begin node 3 end node 5 seg 16 prop 3001
struct prop 3001
struct prop 1001 density 2000.0 e 2E9 area 0.305 I 0.002364
struct prop 3001 density 2000.0 spacing 2.0 e 2E9 area 0.025 I 0.0011 &
cs sstiff 1E9 cs scoh 1000.0 cs nstiff 1E9 cs sfric 30.0 cs ncoh 1000.0 &
perimeter 0.28 cs nfric 30.0
struct prop 3002 density 2000.0 spacing 2.0 e 2E9 area 0.025 I 0.0011 &
cs sstiff 1E9 cs scoh 4000.0 cs nstiff 1E9 cs sfric 40.0 cs ncoh 4000.0 &
perimeter 0.28 cs nfric 40.0
FLAC Version 5.0
15 - 21
solve
save wf5.sav
;*** BRANCH: M-C MODEL ****
;... STATE: WF6 ....
set echo off
call savepp.fis
savePP
water bulk=1e9
set dyn=on
set =large
hist 100 read acc1.his
hist write 100 table 100
set echo off
call convert.fis
set tab in=100 tab out=101 npnts=4000 cfactor=10.0
convert
set echo off
call Filter.fis
set filter in=101 filter out=111 fc=10
filter
history 1 dytime
history 5 xvel i=9, j=3
history 9 node 19 xvelocity
history 12 node 1 xdisplace
history 13 ssi i=32, j=4
history 16 sxy i=32, j=4
history 19 xaccel i=13, j=1
set step=10000000
save wf6.sav
;*** BRANCH: UNDAMPED ****
FLAC Version 5.0
15 - 22
;... STATE: WF7 ....

apply ffield
apply xacc 0.5 hist table 111 from 1,1 to 41,1
solve dytime 10.0
save wf7.sav
;*** BRANCH: SOIL 1 ****
;... STATE: WF7 A ....
set echo off
call tab ind.fis
set fft inp1=106 fft inp2=206
tab ind
set echo off
call Fft.fis
fftransform
save wf7 a.sav
restore wf7.sav
;... STATE: WF7 B ....
set echo off
call tab ind.fis
tab ind
set echo off
call Fft.fis
fftransform
save wf7 b.sav
restore wf7.sav
;... STATE: WF7 C ....
set echo off
call tab ind.fis
tab ind
set echo off
call Fft.fis
FLAC Version 5.0
15 - 23
fftransform
save wf7 c.sav
;*** BRANCH: PILE IN SOIL1 ****
restore wf7.sav
;... STATE: WF7 D ....
set echo off
call tab ind.fis
tab ind
set echo off
call Fft.fis
fftransform
save wf7 d.sav
;*** BRANCH: PILE IN SOIL2 ****
restore wf7.sav
;... STATE: WF7 E ....
set echo off
call tab ind.fis
tab ind
set echo off
call Fft.fis
fftransform
save wf7 e.sav
;*** BRANCH: W/ RAYLEIGH DAMP GRID AND STRUCTURE ****
restore wf6.sav
;... STATE: WF8S ....
set dy damping rayleigh=0.01 0.8
set dy damping struct rayleigh=0.01 0.8
apply ffield
solve dytime 5.0
save wf8s.sav
;*** BRANCH: W/ RAYLEIGH DAMP GRID ONLY2 ****
restore wf6.sav
FLAC Version 5.0
15 - 24
;... STATE: WF8 ....

apply ffield
solve dytime 5.0
save wf8.sav
;... STATE: WF9 ....
solve dytime 10.0
save wf9.sav
;... STATE: WF10 ....
solve dytime 15.0
save wf10.sav
;... STATE: WF11 ....
solve dytime 20.0
save wf11.sav
;*** BRANCH: M-C MODEL W/ HYST DAMP ****
restore wf5.sav
;... STATE: WF6H ....
set echo off
call savepp.fis
savePP
water bulk=1e9
set dyn=on
set =large
set echo off
call convert.fis
convert
set echo off
call Filter.fis
filter
history 1 dytime
FLAC Version 5.0
15 - 25

set step=10000000
save wf6h.sav
;... STATE: WF8H ....
ini dy damp hyst default -3.325 0.823
apply ff
solve dytime 5.0
save wf8h.sav
;... STATE: WF9H ....
solve dytime 10.0
save wf9h.sav
;... STATE: WF10H ....
solve dytime 15.0
save wf10h.sav
;... STATE: WF11H ....
solve dytime 20.0
save wf11h.sav
;*** BRANCH: FINN MODEL ****
restore wf5.sav
;... STATE: WF6F1 ....
group Soil 2 (Finn) region 32 9
model finn group Soil 2 (Finn)
friction=35.0 ff c1=0.2057 ff c2=1.946 ff switch=1 ff latency 1000000 &
FLAC Version 5.0
15 - 26
group Soil 2 (Finn)

group Soil 3 (Finn) region 25 16
model finn group Soil 3 (Finn)
friction=30.0 ff c1=0.2057 ff c2=1.946 ff switch=1 ff latency 1000000 &
group Soil 3 (Finn)
solve
save wf6f1.sav
;... STATE: WF6F2 ....
water bulk=5.0E8
set dyn=on
set =large
set echo off
call convert.fis
convert
set echo off
call Filter.fis
filter
history 1 dytime
set step=10000000
prop ff latency 50 region 32 10
FLAC Version 5.0
15 - 27
group wharf:soil1 i 40 j 7 18
model mohr group wharf:soil1
friction=40.0 dilation=0.0 tension=0.0 group wharf:soil1
ini dy damp hyst default -3.325 0.823
apply ffield
save wf6f2.sav
;... STATE: WF8F ....
solve dytime 5.0
save wf8f.sav
;... STATE: WF9F ....
solve dytime 10.0
save wf9f.sav
;plot name: shear strain
plot hold ssi fill int 0.15 struct pile moment fill iwhite beam alias &
wharf white
;plot name: input gs
plot hold history 100
;plot name: fft
plot hold table 206 line end 200
;plot name: pp hist
;plot name: pp
;plot name: plasticity
plot hold displacement bound plasticity no past
;plot name: xdisp hist
;plot name: input acc
plot hold table 111 line
;plot name: xacc base
plot hold history 19 line 20 vs 1
FLAC Version 5.0
15 - 28
15.5 Data File INPUT.DAT

;Title:Input wave
;... STATE: INPUT ....
config
call convert.fis
set tab in 100 tab out 101 npnts 4000 cfactor 10.0
convert
set echo off
call Filter.fis
set cd back
filter
set echo off
call INT.FIS
set int in=101 int out=102
integrate
call INT.FIS
integrate
save input.sav
;*** BRANCH: UNFILTERED FFT ****
;... STATE: FFT1 ....
set echo off
call tab ind.fis
tab ind
set echo off
call Fft.fis
set cd back
fftransform
save fft1.sav
;*** BRANCH: FILTERED FFT ****
restore input.sav
;... STATE: FFT2 ....
set echo off
call tab ind.fis
FLAC Version 5.0
15 - 29
tab ind
set echo off
call Fft.fis
set cd back
fftransform
save fft2.sav
;plot name: input acc. (hist 100)
;plot name: input acc gs (table 100)
label table 100
(gs/sec2 vs sec)
plot hold table 100 line alias Acceleration Record
;plot name: input acc m/s2 (table 101)
label table 101
(m/sec2 vs sec)
plot hold table 101 line alias Acceleration Record
;plot name: vel m/s (table 102)
label table 102
uncorrected velocity
label table 104
corrected velocity
plot hold table 102 line alias Velocity Record
;plot name: disp m (table 103)
label table 106
uncorrected disp.
label table 105
corrected disp.
label table 103
uncorrected displacement
plot hold table 103 line alias Displacement Record
;plot name: fft
label table 110
label table 201
plot hold table 201 line alias Power Spectrum
FLAC Version 5.0
15 - 30
FLAC Version 5.0
Pile-Supported Highway Embankment
16 - 1
16 Pile-Supported Highway Embankment

End-bearing piles can be used to support highway embankments constructed over soft foundation
materials. This method of support can reduce the potential for excessive deformations and failure
during the undrained stage of construction when excess pore pressures are induced in the foundation
materials by the embankment loading.
This example presents a FLAC analysis of the initial (undrained) construction stage for a highway
embankment built over soft saturated foundation clay and muck, using timber piles to support
the embankment.* The piles extend through the soft materials and into underlying silty sands.
The embankment includes foamed concrete engineered fill as part of the embankment materials.
The light-weight foamed concrete is placed in lifts of approximately 0.6 m thickness. The first
lift is placed over a wire mesh directly in contact with the top of the timber piles. Earth fill and
pavement material are placed as cover over the foamed concrete. The analysis also includes a
traffic surcharge of 11,500 Pa (240 psf). Figure 16.1 shows a section view of the embankment and
foundation materials. The groundwater surface is at the top of the foundation materials.
Figure 16.1 Half-section view of foamed concrete embankment on timber

piles
* This analysis is based on information provided by K. J. Kim of the North Carolina Department of
Transportation on the design of a foamed concrete embankment supported on timber piles for the
U.S. 64 widening project in Tyrrell County, North Carolina.
FLAC Version 5.0
16 - 2
The properties assumed for the foundation and embankment materials are listed in Tables 16.1
and 16.2. Note that both saturated and dry densities are shown for the foundation materials. The
embankment materials are assumed to remain dry. The FLAC simulation is an undrained analysis
using the groundwater configuration mode. Consequently, the drained material bulk modulus and
strength properties and the dry mass densities are input for this calculation mode, because the effect
of water is incorporated in the FLAC calculation.
Table 16.1 Properties for foundation soils
Saturated unit weight (N/m3 )
Porosity (%)
Drained Youngs modulus (MPa)
Drained Poissons ratio
Drained bulk modulus (MPa)
Shear modulus (MPa)
Drained cohesion (Pa)
Drained friction angle (degrees)
Horizontal permeability (m/day)
Vertical permeability (m/day)
muck
very soft clay
silty sand
11,100
90
231
0.3
0.49
5.0
0.1
3500
0
0
0.003
0.001
13,560
80
582
0.5
0.45
1.67
0.17
5000
0
0
0.0003
0.0001
18,840
30
1620
15.0
0.3
12.5
5.77
0
32
0
2.4
0.8
Table 16.2 Properties for embankment materials

Porosity (%)
Drained Youngs modulus (MPa)
Drained Poissons ratio
Drained bulk modulus (MPa)
Shear modulus (MPa)
Drained cohesion (Pa)
Drained friction angle (degrees)
Horizontal permeability (m/day)
Vertical permeability (m/day)
foamed concrete
earth fill
640
30
600.0
0.15
286.0
261.0
50,000
0
0
1.2
0.4
1920
30
10.0
0.3
8.33
3.85
2400
30
0
1.2
0.4
Treated timber piles are located on a 2.5 m by 2.5 m rectangular spacing beneath the embankment
materials. The length of each pile is 12.8 m (42 ft), and the average pile diameter is 0.3048 m (12
in). The properties of the timber piles are listed in Table 16.3.
FLAC Version 5.0
16 - 3
Table 16.3 Properties for treated timber piles

Elastic modulus (GPa)
End-bearing capacity (KN)
10.0
250.0

This analysis is performed as a parametric study to compare the deformation of an unsupported
embankment to that of a pile-supported embankment. In both cases, we first determine the initial
equilibrium state of the saturated foundation soils. Then, for the unsupported case, we add the
embankment materials and monitor the vertical displacement along the foundation surface directly
beneath the embankment. For the pile-supported case, we install the timber piles and then add the
layers of embankment materials while monitoring the vertical displacements in the same locations
as those for the unsupported case.
The model is created using FLAC s graphical interface, the GIIC. Upon entering the GIIC, the
groundwater flow option, structural elements and factor-of-safety calculation are activated in the
Model Options dialog. The Save Project As menu item is then selected from the File menu in
order to set up a project file to save the model state at various stages of the simulation. We click on
?
in this menu dialog to select a directory in which to save the project file. A record of the FLAC
commands used to create this model is saved after the analysis is complete, using the File / Export
Record menu item. A listing of the record created for this model is given in Section 16.3.
We generate the grid using the Build/Block tool to create a two block by two block grid. Then, we use
the Alter/Shape tool to generate lines defining the excavation slope and the excavation and foundation
material boundaries. The embankment is 3 m high and the pavement half-width is 12 m. The
resulting grid is shown in Figure 16.2 and coincides with the half-section shown in Figure 16.1.
The grid before alteration is saved as state P1.SAV, and after alteration as P2.SAV.
FLAC Version 5.0
16 - 4
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

1.500
LEGEND
17-Jan-05 16:23
step
0
-4.444E+00 <x< 8.444E+01
-6.294E+01 <y< 2.594E+01
0.500
-0.500
Grid plot
0
2E 1
-1.500
-2.500
-3.500
-4.500
-5.500

0.500
1.500
2.500
3.500
(*10^1)
4.500
5.500
6.500
7.500
Figure 16.2 FLAC grid for highway embankment analysis

The different materials and their associated properties are assigned by group names using the
Material/Assign
tool. Three foundation soil groups are created: silty sand, very soft clay and muck.
The embankment consists of four lifts of 0.6 m thick foamed concrete and the earth fill layer.
The groups defined for the embankment and foundation materials are shown in Figure 16.3. The
groundwater properties are assigned using the Material/GWProp tool. The model is saved at this stage
as P3.SAV.
FLAC Version 5.0
16 - 5
Figure 16.3 Groups defined for embankment and foundation materials

After all the material groups are assigned, the foamed concrete lifts and earth fill groups are excavated using the Material/Cut&Fill tool. The initial stress state for the saturated foundation soils is then
calculated. We use the ININV.FIS FISH function provided in the FISH library (see ININV.FIS
in Section 3 in the FISH volume). This function automatically calculates the pore pressures and
total stresses that are compatible for a model containing a phreatic surface. The groundwater density
and water bulk modulus are specified before applying this FISH function. We use the Settings/GW
tool to set the groundwater density to 1000 kg/m3 and the groundwater bulk modulus to 10,000
Pa (to speed convergence to steady-state flow). We then use the Utility/FishLib tool to access the
ININV.FIS FISH function. We enter the phreatic surface elevation (wth = 0) and the Ko ratios
(k0x = 1.0 and k0z = 1.0) in the dialog, and press OK . The FISH function is called into FLAC and
executed. The pore pressure distribution and total stress adjustment are then calculated automatically. We now solve for the new equilibrium state, using the Run/Solve tool and running in coupled
mechanical-groundwater flow mode. The pore-pressure distribution is shown in Figure 16.4. The
model is saved at this state as P4.SAV.
FLAC Version 5.0
16 - 6
JOB TITLE : .
(*10^1)
FLAC (Version 5.00)

1.500
LEGEND
17-Jan-05 16:25
step
2210
Flow Time
1.9620E+06
-4.444E+00 <x< 8.444E+01
-6.433E+01 <y< 2.456E+01
0.500
-0.500

0.00E+00
5.00E+04
1.00E+05
1.50E+05
2.00E+05
2.50E+05
3.00E+05
3.50E+05
-1.500
-2.500
-3.500

-4.500
-5.500

0.500
1.500
2.500
3.500
(*10^1)
4.500
5.500
6.500
7.500
Figure 16.4 Initial pore-pressure distribution in foundation soils

The embankment construction is analyzed assuming undrained conditions. This is accomplished
by setting the groundwater flow calculation mode off, and increasing the water bulk modulus to
approximate a nearly incompressible fluid. We increase the water bulk modulus to 0.2 GPa and
set flow off using the Settings/GW tool. Also, because we anticipate large deformations during the
construction, we perform this stage in large-strain mode and set this option on in the Settings/Mech
tool.
The unsupported embankment construction is simulated by adding each embankment-lift group
individually (via the Material/Cut&Fill tool), and then solving for the equilibrium state with this lift
in place. As each group is added, the saturation values of the gridpoints in the group are fixed
at zero (using the In Situ/Initial and In Situ/Fix tools) to simulate the unsaturated condition of the
embankment materials. Figure 16.5 shows the Material/Cut&Fill tool with the foamed concrete1 lift
added (filled). (We note that if the Show Excavations? box is checked, then the excavated groups are
shown grayed-out in this tool.) These steps are repeated for each of the three remaining foamed
concrete lifts and the earth-fill lift. Finally, the traffic surcharge is applied along the top of the
embankment, using the In Situ/Apply tool. Each of the unsupported construction stages are saved as
separate save states in P5.SAV through P10.SAV.
FLAC Version 5.0
16 - 7
Figure 16.5 Addition of first embankment lift foamed concrete1

Vertical displacement histories are recorded at four locations along the base of the embankment, at
(x = 0, y = 0), (x = 5, y = 0), (x = 11, y = 0) and (x = 16, y = 0). The displacements are monitored
throughout the embankment construction; the results are shown in Figure 16.6. The extent of the
displacements induced by the unsupported construction is shown in Figure 16.7. The maximum
vertical displacement beneath the embankment is approximately 0.6 m (2 ft).
The displacements are associated with excess pore pressures that develop in the muck and very
soft clay. This is evident from the pore pressure histories recorded along the centerline of the
embankment at y = 0 and y = -6 (in the muck), and at y = -10 (in the very soft clay). The plots of
pore-pressure histories are given in Figure 16.8.
A factor-of-safety calculation is performed at this stage by selecting the Run/SolveFoS tool. The safety
factor for the unsupported embankment is calculated to be 1.06. Figure 16.9 shows the failure
surface that develops if cohesion and friction of the embankment and foundation materials are
reduced by this factor.
FLAC Version 5.0
16 - 8
JOB TITLE : .
FLAC (Version 5.00)

(10
-01
LEGEND
17-Jan-05 17:47
step 76054
Flow Time
1.9620E+06
0.000
HISTORY PLOT
Y-axis :
-1.000
-2.000

-3.000
X-axis :
Number of steps
-4.000
-5.000
10
20
30
40
50
60
70
(10
03

Figure 16.6 Vertical displacements along base of embankment for unsupported embankment construction
JOB TITLE : .
(*10^1)
2.000
FLAC (Version 5.00)

LEGEND
1.000
17-Jan-05 17:47
step 76054
Flow Time
1.9620E+06
-3.588E+00 <x< 6.343E+01
-4.508E+01 <y< 2.194E+01
0.000
0.00E+00
1.00E-01
2.00E-01
3.00E-01
4.00E-01
5.00E-01
6.00E-01
7.00E-01
8.00E-01
-1.000
-2.000

0
-3.000
2E 0
-4.000

0.500
1.500
2.500
3.500
4.500
5.500
(*10^1)
Figure 16.7 Displacement vectors and x-displacement contours for unsupported embankment construction
FLAC Version 5.0
16 - 9
JOB TITLE : .
FLAC (Version 5.00)

(10
05
LEGEND
17-Jan-05 17:47
step 76054
Flow Time
1.9620E+06
1.200
HISTORY PLOT
Y-axis :
5 Grid-point pp ( 1, 40)
1.000
0.800
X-axis :
Number of steps
0.600
0.400
0.200
10
20
30
40
50
60
70
(10
03

Figure 16.8 Pore pressures beneath center of embankment for unsupported

embankment construction
JOB TITLE : .
(*10^1)
2.000
FLAC (Version 5.00)

LEGEND
1.000
18-Jan-05 8:45
step 181132
Flow Time
1.9620E+06
-3.588E+00 <x< 6.343E+01
-4.508E+01 <y< 2.194E+01
0.000

0.00E+00
1.00E-01
2.00E-01
3.00E-01
4.00E-01
5.00E-01
-1.000
-2.000

Velocity vectors
-3.000
5E -6
-4.000

0.500
1.500
2.500
3.500
4.500
5.500
(*10^1)
Figure 16.9 Factor of safety and failure surface plot for unsupported embankment
FLAC Version 5.0
16 - 10
The pile-supported embankment construction is simulated by first installing pile elements in the
FLAC model. The model state at P4.SAV (the initial equilibrium state) is restored, and seven
piles of 12.8 m length are positioned at a 2.5 m spacing within the foundation soils. Before the
piles are placed in the model, the first foamed concrete lift is added. This is done so that the top of
the piles can be connected to the embankment materials. Then, the piles are positioned as shown
in Figure 16.10.
In order to represent the three-dimensional effect of the 2.5 m pile spacing, we scale the pile
properties by dividing by the pile spacing. This is done automatically by specifying the spacing
property when assigning pile properties. In this analysis only the elastic modulus and the endbearing capacity are scaled to account for the spacing. Note that we neglect the weight of the piles;
the pile density would also be scaled if this weight is included. (See Section 1.9.4 in Structural
Elements for additional information on scaling properties to simulate the three-dimensional effect.)
The properties of the pile coupling springs are selected to simulate an end-bearing capacity and zero
skin friction. The cohesive strengths of the shear coupling springs at the top and bottom elements
of each pile are set to 2.5 MN/m, while all other shear and normal coupling-spring strength values
are set to zero. The value for cohesive strength is derived from a simulation of axially loaded
piles at 2.5 m spacing to produce an end-bearing ultimate capacity of 250 KN in the silty-sand
foundation material. The value for coupling-spring shear stiffness is selected at approximately ten
times the equivalent stiffness of the stiffest neighboring zone. By doing this, the deformability at
the pile/soil interface will have minimal influence on both the compliance of the total model and
the calculational speed (see Section 4.4.1 in Theory and Background). The properties used for
the pile elements in this model are summarized in Table 16.4. The model is saved at this stage as
P11.SAV.
Figure 16.10 Location of piles in FLAC model
FLAC Version 5.0
16 - 11
Table 16.4 Properties for pile elements
Elastic modulus (GPa)

Radius (m)
Perimeter (m)
Spacing (m)
Shear coupling spring stiffness (GN/m/m)
Shear coupling spring cohesion (MN/m)
Shear coupling spring friction (degrees)
Normal coupling spring stiffness (GN/m/m)
Normal coupling spring cohesion (N/m)
Normal coupling spring friction (degrees)
middle segments
top & bottom segments
10.0
0.1524
0.976
2.5
0.0
0.0
0.0
0.0
0.0
0.0
10.0
0.1524
0.976
2.5
1.0
2.5
0.0
0.0
0.0
0.0
The embankment construction steps are now performed following the same sequence as for the
unsupported case. Each of the pile-supported stages are saved as separate save states in P12.SAV
through P17.SAV.
The vertical displacements are monitored as before; the histories are shown in Figure 16.11. The
maximum vertical displacement beneath the embankment is now approximately 0.03 m (1 in).
Also, we note that for this case there is an insignificant change in pore pressures in the muck and
very soft clay, as seen in Figure 16.12.
FLAC Version 5.0
16 - 12
JOB TITLE : .
FLAC (Version 5.00)

(10
-01
LEGEND
17-Jan-05 20:15
step 63680
Flow Time
1.9620E+06
1.000
0.000
HISTORY PLOT
Y-axis :
-1.000
-2.000

X-axis :
Number of steps
-3.000
-4.000
-5.000
10
20
30
40
50
60
(10
03

Figure 16.11 Vertical displacements along base of the embankment for pilesupported embankment construction
JOB TITLE : .
FLAC (Version 5.00)

(10
05
LEGEND
17-Jan-05 20:15
step 63680
Flow Time
1.9620E+06
1.000
HISTORY PLOT
Y-axis :
0.800
0.600
X-axis :
Number of steps
0.400
0.200
10
20
30
40
50
60
(10
03

Figure 16.12 Pore pressures beneath center of embankment for pile-supported

embankment construction
FLAC Version 5.0
16 - 13
We are also interested in the axial loading that develops in the piles. When the spacing property is
assigned, the axial force values that are printed and plotted output are the actual values (i.e., they
account for the pile spacing). We plot the actual axial forces in the piles in Figure 16.13. (The 21 pile
numbers shown in the plot legend correspond to the top, middle and bottom pile segments, which
are assigned different material property numbers.) The maximum pile loading is approximately
210 KN.
A factor-of-safety calculation is also performed at this stage. The calculated factor is 1.46, and the
failure surface is shown by the plot in Figure 16.14. Note that the critical failure surface for the
supported embankment is now at the toe of the earthfill berm. The safety factor for the foundation
material beneath the embankment is greater than 1.46, as a result of the support provided by the
piles.
Finally, we note that this project can be re-created by importing the data file PEMBANK.DAT
listed in Section 16.3, using the File / Import Record menu item. After the record is imported to
the GIIC, each save state can be created by first clicking on that state in the Record pane, and then
clicking on the restore state button at the top of the pane. The commands associated with that state
will then be called into FLAC. Note that the Project Tree Record format must be enabled (from the
Model Options dialog) to import this record.
Figure 16.13 Actual loads in piles for pile-supported embankment construction
FLAC Version 5.0
16 - 14
JOB TITLE : .
(*10^1)
2.000
FLAC (Version 5.00)

LEGEND
1.000
18-Jan-05 13:00
step 166424
Flow Time
1.9620E+06
-3.588E+00 <x< 6.343E+01
-4.508E+01 <y< 2.194E+01
0.000

0.00E+00
2.50E-01
5.00E-01
7.50E-01
1.00E+00
1.25E+00
1.50E+00
-1.000
-2.000

Velocity vectors
0
-3.000
5E -5
-4.000

0.500
1.500
2.500
3.500
4.500
5.500
(*10^1)
Figure 16.14 Factor of safety and failure surface plot for supported embankment
FLAC Version 5.0
16 - 15
16.3 Data File PEMBANK.DAT

Project Record Tree export
;Title:Piled Embankment
;... State: p1.sav ....
config gwflow
grid 60,45
gen 0.0,-40.0 0.0,-15.0 30.0,-15.0 30.0,-40.0 ratio 1.0,0.9 i 1 31 j 1 16
gen 0.0,-15.0 0.0,3.0 30.0,3.0 30.0,-15.0 i 1 31 j 16 46
gen 30.0,-40.0 30.0,-15.0 80.0,-15.0 80.0,-40.0 ratio 1.02,0.9 &
i 31 61 j 1 16
gen 30.0,-15.0 30.0,3.0 80.0,3.0 80.0,-15.0 ratio 1.02,1.0 i 31 61 j 16 46
model elastic i 1 60 j 1 45
save p1.sav
;... State: p2.sav ....
gen line 11.75,3.0 20.75,0.0
gen line 0.0,-9.0 20.0,-5.0
gen line 0.0,-12.0 20.0,-10.5
mark i 21 61 j 41
mark i 20 61 j 32
mark i 20 61 j 23
mark i 1 20 j 41
model null i 14 j 45
gen line 0.0,2.5 11.0,2.5
ini x 12.543801 y 2.0714378 i 13 j 44
ini x 13.681461 y 1.5236754 i 14 j 43
ini x 15.619698 y 0.84950686 i 16 j 42
ini x 14.187088 y 1.3551335 i 15 j 43
ini x 16.336002 y 0.63882875 i 17 j 42
save p2.sav
;... State: p3.sav ....
group silty sand region 55 5
model mohr group silty sand
friction=32.0 dilation=0.0 tension=0.0 group silty sand
FLAC Version 5.0
16 - 16
group very soft clay region 59 29

model mohr group very soft clay
friction=0.0 dilation=0.0 tension=0.0 group very soft clay
group muck region 58 36
model mohr group muck
friction=0.0 dilation=0.0 tension=0.0 group muck
group foamed concrete1 i 1 17 j 41
model mohr group foamed concrete1
friction=0.0 dilation=0.0 tension=0.0 group foamed concrete1
group earth fill i 20 j 41
model mohr group earth fill
friction=30.0 dilation=0.0 tension=0.0 group earth fill
FLAC Version 5.0
16 - 17
prop density=1920.0 bulk=8330000.0 shear=3850000.0 cohesion=2400.0

group earth fill i 14 15 j 43
group earth fill notnull i 13 14 j 44 45
group earth fill i 1 12 j 45
prop por=0.3 k11=2.83E-9 k22=9.44E-10 group foamed concrete1
prop por=0.3 k11=2.83E-9 k22=9.44E-10 group earth fill
prop por=0.3 k11=2.83E-9 k22=9.44E-10 group silty sand
prop por=0.8 k11=3.54E-13 k22=1.17E-13 group very soft clay
prop por=0.9 k11=3.53E-12 k22=1.18E-12 group muck
save p3.sav
&
&
&
&
&
;... State: p4.sav ....

model null group earth fill
model null group foamed concrete4
fix x y i 1 61 j 1
fix x i 61 j 1 41
fix x i 1 j 1 41
set gravity=9.81
water bulk 2e4
set echo off
call Ininv.fis
set wth=0.0 k0x=0.5 k0z=0.5
ininv
solve
FLAC Version 5.0
16 - 18
save p4.sav
;*** Branch: Unsupported ****
;... State: p5.sav ....
set flow=off
water bulk=2.0E8
set =large
fix x i 1 j 41 42
initial xdis 0.0 ydis 0.0
solve
save p5.sav
;... State: p6.sav ....
fix x i 1 j 42 43
solve
save p6.sav
;... State: p7.sav ....
fix x i 1 j 43 44
solve
save p7.sav
FLAC Version 5.0
16 - 19
;... State: p8.sav ....

fix x i 1 j 44 45
solve
save p8.sav
;... State: p9.sav ....
fix x i 1 j 45 46
solve
save p9.sav
;... State: p10.sav ....
solve
save p10.sav
;... State: FoSmode.fsv ....
;FoS save state:
solve fos file FoSmode.fsv
;*** Branch: Pile-supported ****
restore p4.sav
;... State: p11.sav ....
struct node 10 11.25,-12.8
FLAC Version 5.0
16 - 20
struct node 11 13.751,0.01

struct node 12 13.75,-12.8
struct node 13 16.251,0.01
struct node 14 16.25,-12.8
struct prop 1001
struct prop 2001
struct prop 3001
struct prop 3001 e 1E10 radius 0.1524 perimeter 0.976 cs ncoh 0 cs nfric 0 &
cs nstiff 0 cs scoh 0.0 cs sstiff 0.0 cs sfric 0 spacing 2.5
struct prop 3002 e 1E10 radius 0.1524 cs scoh 2500000.0 cs sstiff 1e9 per &
0.976 spacing 2.5
struct prop 3002 cs ncoh 0 cs nstiff 0
water bulk 2e8
set flow off
fix x i 1 j 40 42
save p11.sav
;... State: p12.sav ....
history 5 gpp i=1 j=40
FLAC Version 5.0
16 - 21

set large
initial xdis 0.0 ydis 0.0
solve
save p12.sav
;... State: p13.sav ....
solve
save p13.sav
;... State: p14.sav ....
fix x i 1 j 43 46
solve
save p14.sav
;... State: p15.sav ....
fix x i 1 j 43 46
solve
save p15.sav
;... State: p16.sav ....
fix x i 1 j 43 46
solve
save p16.sav
FLAC Version 5.0
16 - 22
;... State: p17.sav ....

solve
save p17.sav
;... State: FoSmode2.fsv ....
;FoS save state:
solve fos no restore file FoSmode2.fsv
;plot name: pore pressure
plot hold pp fill
;plot name: y-disp histories
plot hold history 1 line 2 line 3 line 4 line
;plot name: displacements
plot hold xdisp fill displacement
;plot name: pp histories
plot hold history 5 line 8 line 9 line
;plot name: pile forces
plot hold group struct pile axial fill max 1000000.0
;plot name: FoS
plot hold fos ssi fill velocity
FLAC Version 5.0
Lined Tunnel Construction in Saturated Ground
17 - 1
17 Lined Tunnel Construction in Saturated Ground

A circular tunnel with an excavated radius of 6 m, and with its center at a depth of 65 m below
the surface, is constructed in saturated ground. The pre-construction water table is at a depth of
5 m below the surface. The ground is dewatered during the construction phase. The tunnel is
supported by a temporary shotcrete liner, which is installed while the tunnel excavation advances,
and a final cast-in-place 0.4 m-thick concrete liner. The analysis is required to determine the forces
and moments that develop in the concrete liner both when the water level returns to its original
elevation and when, at a later time, the temporary shotcrete liner loses its strength.
The construction sequence considered in this analysis is divided into six stages.
1. Before construction begins, the water table is lowered to approximately 10 m below the
tunnel invert using dewatering wells. The initial insitu stress ratio, Ko , is 0.5 for the
unsaturated state.
2. The excavation begins, and the tunnel advancement produces tunnel closure corresponding to a 30% relaxation of traction forces acting on the tunnel periphery before the
shotcrete liner is installed.
3. The shotcrete is installed, the tunnel is advanced to produce 100% relaxation, and load
develops in the shotcrete. Note that the shotcrete is sufficiently porous such that it
provides negligible resistance to fluid flow when the original water table is restored.
4. A permanent cast-in-place concrete liner is then placed inside the shotcrete lined tunnel.
A plastic waterproof membrane covers the concrete liner; the concrete plus membrane
is impermeable.
5. After the concrete liner is installed, the dewatering wells are stopped, and the water rises
to the original level. Pore pressures are re-established throughout the ground and the
outer liner, but fluid does not penetrate past the impermeable inner liner. The water
exerts a pressure in the gap between the two liners; this pressure causes axial forces to
develop within the liners.
6. Finally, the shotcrete liner degrades over time, and the ground relaxes into the inner liner.
Figure 17.1 illustrates the construction sequence that is simulated in the FLAC analysis. The figure
also lists the rock, shotcrete and concrete liner properties assumed for the analysis.
FLAC Version 5.0
17 - 2
Figure 17.1 Conditions and sequence for the lined tunnel construction
FLAC Version 5.0
17 - 3

The six construction stages are simulated as separate steps in the FLAC analysis. Each step is
described in the following sections. The model is created using FLAC s graphical interface, the
GIIC. Upon entering the GIIC, the groundwater flow option, adjust total stress option and structural
elements facility are activated from the Model Options dialog. The option for automatic adjustment
of total stresses is selected to facilitate the calculation for Stage 5, when the water table is raised in
the model. Also, 5 extra grid variables are selected; these are available to store grid variables when
creating FISH functions for this analysis. The Project Tree Record format and SI system of units
are also selected for this example. The Model Options dialog is shown in Figure 17.2.
Figure 17.2 Model options selected for lined tunnel example

After the model options are selected, a Project File (*.prj) dialog opens. A working directory is
selected for storing files related to this project, and a project title (Lined tunnel construction in
saturated ground) and project filename (LINER.PRJ) are assigned in the dialog.
A record of all FLAC commands used to create and run this model are saved to file after the project is
complete (using the File / Export Record menu item). A listing of the data record for this model
is given in Section 17.4.
FLAC Version 5.0
17 - 4
17.2.1 Step 1: Initial State with Lowered Water Table

The model creation takes advantage of the problem symmetry, as shown in Figure 17.1. The grid
is constructed using the Build / Library tool. The Single tunnel-refined region grid type is selected, as
shown in Figure 17.3. A fine mesh is created in the region of the tunnel, and a coarse grid extends
away from the tunnel region. ATTACH commands are assigned automatically to connect the coarse
mesh to the fine mesh. When this grid type is selected and a coordinate range is defined, a grid
tool opens to prescribe mesh density and adjust the grid to the problem dimensions. Figure 17.4
shows the grid after adjustments have been made to the corner locations and mesh density. The grid
extends from 0.0 to 120.0 in the x-direction and from -80.0 to 65.0 in the y-direction. The x = 0
coordinate is the line of symmetry. The fine mesh region extends from 0.0 to 20.0 in the x-direction
and from -10.0 to 10.0 in the y-direction, and contains 36 36 zoning. The surrounding mesh is
three times coarser than the fine mesh.
The grid in the fine-mesh region is then altered to match the periphery of the excavated tunnel,
using the Alter / Shape tool. The Circle mode is selected to locate the tunnel with its center at x = 0.0,
y = 0.0 and radius of 6.0. The third row of gridpoints from the top of the grid is adjusted slightly
to the location y = 60.0, using the Line mode in the Shape . This is done to provide a more precise
representation for the pore-pressure distribution (in Stage 5) when the water level is raised. The final
grid is shown in Figure 17.5, and a close-up view of the tunnel region is shown in Figure 17.6. Note
that the resources panes are turned off in the second figure (by checking off the Show / Resources
menu item) to show a full model-view of the grid.
Figure 17.3
FLAC Version 5.0
Build
/ Library tool
17 - 5
Figure 17.4 Grid tool for the Single tunnel-refined region grid type
Figure 17.5 Grid created for lined tunnel example
FLAC Version 5.0
17 - 6
Figure 17.6 Grid created for lined tunnel example close-up view of tunnel
region
The material properties for the rock are entered using the MATERIAL / Assign tool. The properties are
displayed in the Define Material dialog shown in Figure 17.7. The groundwater properties (porosity
= 0.3, and permeability (i.e., mobility coefficient) = 1010 m2 /(Pa-sec)) are also assigned, using the
MATERIAL GWProp
/
tool (see Figure 17.8). Units for groundwater properties in FLAC are discussed in
Section 1.7 in Fluid-Mechanical Interaction.
Figure 17.7 Define Material dialog with rock properties for lined tunnel example
FLAC Version 5.0
17 - 7
Figure 17.8 Assign groundwater properties

The boundary conditions are applied using the In Situ / FIX tool. The left boundary is a line of
symmetry. The right boundary is fixed from movement in the x-direction. The top boundary is a
free surface, and the bottom boundary is fixed from movement in the y-direction. The FIX tool with
the boundary conditions assigned is shown in Figure 17.9.
The initial stress state is specified assuming that no water is present. The rock density listed in
Figure 17.1 is considered to be the unsaturated value, and assuming a value of 10 m/sec2 for the
gravitational magnitude, the calculated vertical stress at the bottom of the model (145 m depth) is
then -3.248 MPa. A linear variation based on this value is entered for the yy-stress component,
using the In Situ / Initial tool. The xx-stress and zz-stress components are also entered, based on
the lateral stress ratio, Ko = 0.5. The input stresses are shown in the Initial tool displayed in
Figure 17.10.
Figure 17.9 Assigned boundary conditions in the
FIX
tool
FLAC Version 5.0
17 - 8
Figure 17.10 Initial stress state distribution entered in the
Initial
tool
The gravitational magnitude is entered using the Settings / Gravity tool, and the groundwater flow
calculation is turned off using the Settings / GW tool. The Run / Solve tool is now used to calculate the
initial stress state. The model should be very close to an equilibrium state. In order to ensure that
a uniform stress distribution is developed in the model, the Solve Initial Equilibrium as Elastic Model check
box is selected.
After the equilibrium state is calculated, the initial stress distribution is checked. Note that, by
default, stress contours do not extend to the model boundaries because stress is calculated at zone
centroids. However, the FISH function EXTRAP.FIS can be used to extrapolate the zone-based
variable to gridpoints. We first copy the stress values we wish to plot into an extended array. The
FISH function shown in Figure 17.11 is entered in the FISH Editor and executed to write the yystress components into grid variable ex 1. Then, the EXTRAP.FIS function is opened in the FISH
Editor from the ITASCA\FLAC500\FISH\3-LIBRARY directory. (The function is described
in Section 3 in the FISH volume.) This function is executed from the FISH Editor by clicking
on the Run / Execute menu item. The extrapolated yy-stress values are calculated and stored in
grid variable ex 2. The yy-stress contours are plotted from the Plot / Model tool using the ContourGP / Extra menu item and selecting the ex 2 plot item. Figure 17.12 displays the resulting plot of the
initial vertical stress distribution. The contour plot is generated for a simple average extrapolation
(gp avg = 0). Figure 17.13 shows the same result presented in a Postscript format. Different
output formats can be selected from the File / Print Plot Setup menu item.
FLAC Version 5.0
17 - 9
Figure 17.11 FILL EX1 SYY.FIS FISH function to copy stresses to an extended array (for use with EXTRAP.FIS)
FLAC Version 5.0
17 - 10
Figure 17.12 Vertical stress distribution at initial state
JOB TITLE : Lined tunnel construction in saturated ground
(*10^2)
FLAC (Version 5.00)

0.700
LEGEND
0.500
15-Sep-04 9:50
step
12
-3.667E+01 <x< 1.567E+02
-1.042E+02 <y< 8.917E+01
0.300
yy-stress contours (extrapolated)

-3.00E+06
-2.50E+06
-2.00E+06
-1.50E+06
-1.00E+06
-5.00E+05
0.00E+00
Grid plot
0
0.100
-0.100
-0.300
5E 1
-0.500
-0.700
-0.900

-0.200
0.000
0.200
0.400
0.600
(*10^2)
0.800
1.000
1.200
1.400
Figure 17.13 Vertical stress distribution at initial state Postscript format
FLAC Version 5.0
17 - 11
17.2.2 Step 2: Tunnel Excavation with 30% Relaxation of Tunnel Tractions

The tunnel region is excavated using the Material / Assign tool. The Region radio button is selected to
delineate the tunnel region, and then the null material is highlighted and the mouse is clicked within
the tunnel region to null the zones in this region. Figure 17.14 shows the null tunnel region in the
Assign
tool.
Figure 17.14 Null tunnel zones in the
Assign
tool
If the Solve tool is executed at this point, the calculation corresponds to instantaneous excavation of
the tunnel. In this exercise, we wish to simulate the installation of the temporary shotcrete liner after
some change in the tunnel load has occurred due to the tunnel advancement. The effect of tunnel
advancement is simulated by relaxing the tractions acting along the tunnel periphery incrementally.
The relaxation tractions can be related to tunnel closure, which, in turn, can be related to the distance
from the face of an advancing tunnel e.g., see Panet (1979).
For this example, the tunnel tractions are relaxed by 30% before the shotcrete liner is installed. A
relaxation FISH function, relax hist, is used to control the rate of relaxation. The function
is shown in Figure 17.15. Three FISH variables are input for this function: ninc specifies the
step increment over which the tractions are relaxed; rstart is the starting multiplier for traction
values; and rstop is the multiplier for traction values after ninc steps are completed. For a 30%
relaxation, rstart = 1.0, rstop = 0.7 and we reduce the tractions over an interval of ninc =
100. The increment value is chosen such that inertial effects are minimized. (See Section 3.5.4 in
the Users Guide for further discussion on minimizing inertial effects.)
FLAC Version 5.0
17 - 12
Figure 17.15 RELAX.FIS controls relaxation of tractions at the tunnelboundary gridpoints

The FISH function apply rf, listed in Figure 17.16, is used to apply tractions along the tunnel
boundary. Reaction forces are first calculated by fixing the boundary gridpoints, using APPLY
xvelocity=0 and APPLY yvelocity=0 commands, and taking a STEP 1 to perform the calculation.*
Then, the function recovers the x- and y-reaction forces at the selected tunnel-boundary gridpoints
and assigns these forces as tractions (with an opposite sign) at the same boundary gridpoints,
using the APPLY xforce and APPLY yforce commands. The history keyword is used to reduce the
tractions linearly between the traction-multiplied limits (rstart and rstop) over the interval
ninc specified in the FISH function relax hist.
The FISH input variables for apply rf are the bounding range i- and j gridpoints for the tunnel
boundary. In this case, ib ap = 45, ie ap = 56, jb ap = 8 and je ap = 30. These values can be
identified from a grid plot that includes the gridpoint numbers (PLOT grid gnum), generated from
the Plot / Model tool.
After apply rf is executed, we need to re-fix the x-velocity of the gridpoints at the top and bottom
of the tunnel; the fixity condition was removed during the execution of the function. We also select
the large-strain calculation mode, using the Settings / Mech tool, and monitor displacements at the
top of the model and around the tunnel periphery, using the Utility / History tool. The FISH function
VERT CLOSURE.FIS (see Figure 17.17) is executed from the Fish Editor to monitor the vertical
closure/opening calculated as the difference between the vertical displacement at the crown and
invert of the tunnel. Before cycling, the application of the applied forces around the tunnel should
be checked by creating a plot of applied forces using the PLOT / Model tool. See Figure 17.18.
* Note that the tunnel-boundary gridpoints are found by using the logical condition for identifying
marked gridpoints (and(ags(ii,jj), 128) = 128). The tunnel gridpoints are the only marked gridpoints
in this model.
FLAC Version 5.0
17 - 13
Figure 17.16 APPLY RF.FIS FISH function to apply tractions to tunnelboundary gridpoints
We solve for the equilibrium solution at 30% relaxation of the tunnel tractions by invoking the
Run Solve
/
tool. The APPLY commands implemented in apply rf relax the tunnel tractions during
the calculation, and a new equilibrium state is found. The reduction is indicated by the appliedforces plot shown in Figure 17.19 compare to Figure 17.18.
Figure 17.17 VERT CLOSURE.FIS FISH function
FLAC Version 5.0
17 - 14
Figure 17.18 Applied forces added to tunnel-boundary gridpoints from apply rf FISH function
Figure 17.19 Applied forces after 30% relaxation of tractions
FLAC Version 5.0
17 - 15
17.2.3 Step 3: Install Shotcrete and Relax Tunnel Tractions 100%

The shotcrete is installed by using the Structure / Liner tool. The structural nodes are attached directly
to the grid, as shown in Figure 17.20. The shotcrete properties are assigned in the Structure / SEProp
tool. The property ID for the shotcrete liner elements is identified in the tool as L1. By clicking
on this ID number, the Liner Element Properties dialog opens and properties are entered, as shown
is Figure 17.21. Note that the weight of the shotcrete is neglected for this example.
Figure 17.20 Shotcrete liner installed as liner elements attached to tunnelboundary gridpoints
Because only one-half of the liner is modeled, the structural nodes along the line of symmetry must
be fixed from translation in the x-direction and from rotation. The fixity conditions for these nodes
(node numbers 1 and 45) are set in the Structure / Node tool.
FLAC Version 5.0
17 - 16
Figure 17.21 Shotcrete properties assigned in the Liner Element Properties

dialog
The tunnel loads are now relaxed completely around the tunnel. The FISH function RELAX.FIS
is executed again from the FISH Editor with rstart = 1.0, rstop = 0.0, and ninc = 100. The
tractions are then reduced 100% by executing APPLY RF.FIS again over the same gridpoint range
as before. The new equilibrium solution at 100% relaxation is found using Run / Solve .
Figure 17.22 displays the vertical and horizontal closure histories for the tunnel. The total vertical
closure is approximately 11.5 cm, and the total horizontal closure approximately 2.5 cm, after
100% relaxation. The axial forces that develop in the shotcrete after total relaxation are shown in
Figure 17.23. The maximum force corresponds to an axial stress of approximately 20 MPa.
FLAC Version 5.0
17 - 17
JOB TITLE : .
FLAC (Version 5.00)

(10
-01
LEGEND
17-Jan-05 16:32
step
4568
-0.200
HISTORY PLOT
Y-axis :
-0.400
5 vert_closure (FISH)
X-axis :
Number of steps
-0.600
-0.800
-1.000
10
15
20
25
30
35
40
45
(10
02

Figure 17.22 Vertical and horizontal closures around tunnel
JOB TITLE : .
FLAC (Version 5.00)
8.000
LEGEND
6.000
17-Jan-05 16:32
step
4568
-2.000E+00 <x< 1.600E+01
-9.000E+00 <y< 9.000E+00
4.000
Liner Plot
2.000
Axial Force on
Structure
Max. Value
# 1 (Liner)
4.076E+06
Boundary plot
0
0.000
5E 0
-2.000
-4.000
-6.000
-8.000

-0.100
0.100
0.300
0.500
0.700
(*10^1)
0.900
1.100
1.300
1.500
Figure 17.23 Axial forces in shotcrete after 100% relaxation of tunnel loads
FLAC Version 5.0
17 - 18
17.2.4 Step 4: Install Concrete Liner

The cast-in-place concrete liner is installed in Step 4. The Structure / Liner tool is used, as shown
in Figure 17.24. However, this time, an interface provides the connection between the new liner
and the grid.* The liner elements for the concrete liner are created by first checking the Attach nodes
to an interface
radio button in the Structure / Liner tool. We also specify an initial gap of 1 cm between
the liner elements and the grid, after checking the Add a gap from grid? radio button. When the new
nodes are positioned, by dragging the mouse along the tunnel boundary, an Interface properties
dialog opens to assign the interface properties. The interface is assumed to only provide frictional
resistance, with a friction angle of 30 . The normal and shear stiffness values for the interface (4000
MPa/m) are selected to satisfy the condition that the deformability of the interface has minimal
influence on both the compliance of the total model and the calculational speed (see Section 4.4.1
in Theory and Background).
Figure 17.24 Concrete liner elements installed and connected to the grid using
an interface
* Note that the concrete liner is connected to the grid via an interface, and not to the shotcrete liner.
This is done because the shotcrete liner will be deleted later in the analysis (in Step 6). If the
concrete and shotcrete liners are connected with an interface, and the shotcrete liner is deleted,
then the interface is also deleted. A new interface would need to be created between the concrete
liner and grid at Step 6. Because the shotcrete liner and grid are rigidly attached, it is sufficient to
connect the concrete liner to the grid with the interface at Step 4.
FLAC Version 5.0
17 - 19
In order to assign material properties for the concrete liner different from those for the shotcrete,
a different property ID number must be prescribed for the new liner elements. The PropID radio
button is selected while still in the Structure / Liner tool. An L1 symbol will appear over the newly
created liner elements. By clicking on this symbol, a dialog opens to rename the symbol, as shown
in Figure 17.25. Select L2 and all the new segments are assigned L2.
The material properties for the concrete liner are assigned using the Structure / SEProp tool. Click on
one of the L2 symbols to open the dialog, as shown in Figure 17.26. (Note that L2 property
should be highlighted in the dialog.) The weight of the concrete liner is included for the concrete
liner by assigning the mass density in this dialog. It is only necessary to assign the properties for
one L2 element; all elements with L2 IDs will then be prescribed these properties.
The translation and rotation fixity conditions for the concrete-liner nodes along the centerline (nodes
46 and 90) must also be set, using the Structure / Node tool. This satisfies the symmetry condition as
done previously for the shotcrete liner.
Figure 17.25 Re-assigning liner property ID numbers to L2 for the concrete

liner elements
FLAC Version 5.0
17 - 20
Figure 17.26 Concrete liner properties assigned in Liner Element Properties

dialog
The model is stepped again to an equilibrium state using Run / Solve . The concrete liner settles onto
the grid at the invert of the tunnel, and normal stresses develop along the interface, as shown in
Figure 17.27. The alternate interface normal stress plot is used, to clearly show those interface
segments in contact. Note that stresses act mainly on alternate segments because slight shear
movement on the interface causes geometrical mismatch for the contacting nodes on one side of the
interface. This behavior is quite normal and is illustrated in Figure 4.16 in Section 4.5.5 in Theory
and Background.
FLAC Version 5.0
17 - 21
JOB TITLE : .
FLAC (Version 5.00)
8.000
LEGEND
6.000
17-Jan-05 16:40
step 11796
Flow Time
1.2066E+04
-2.000E+00 <x< 1.600E+01
-9.000E+00 <y< 9.000E+00
4.000
2.000
Boundary plot
0
5E 0
0.000
Normal Stress (alternate)

on Interface # 1
Max Value = 1.156E+05
-2.000
-4.000
-6.000
-8.000

-0.100
0.100
0.300
0.500
0.700
(*10^1)
0.900
1.100
1.300
1.500
Figure 17.27 Normal stresses along interface after concrete liner settles onto
grid at tunnel invert
17.2.5 Step 5: Reestablish Water Table

The water level is raised to the original elevation in Step 5. This stage of the simulation is performed
in two steps (as an uncoupled analysis). First, a flow-only calculation is performed to establish
the pore-pressure distribution. Then, a mechanical-only calculation is performed to establish the
change in the stress state and the loading in the liners due to the water pressure acting on the liners.
The water level is raised by using the In Situ / Initial tool to specify a pore pressure distribution. For
the water table located at y = 60 m in this model (5 m below the surface), and the bottom of the
model located at y = -80 m, the pore pressure varies linearly from zero at y = 60 to 1.4 MPa at
y = -80. Figure 17.28 shows the input for the pore-pressure distribution in the In Situ / Initial tool.
Note that the tool must be applied twice to cover the model because the highlighted region cannot
extend across attached gridpoints.
FLAC Version 5.0
17 - 22
Figure 17.28 Assign pore-pressure distribution corresponding to raising water

table
The mechanical-calculation mode is turned off in the Settings / Mech tool, and the flow calculation is
turned on in the Settings / GW tool. The water modulus and water density are also required for the
flow-only calculation. The modulus is set to 10,000 Pa; this low value speeds convergence for
steady-state flow (see Section 1.7.5.3 in Fluid-Mechanical Interaction). The water density is set
to 1000 kg/m3 .
The pore-pressure distribution should correspond to a steady-flow state. Some stepping is necessary
to satisfy steady flow due to the variation in zoning. The Run / Solve tool is executed to calculate this
state. The resulting pore pressure distribution is shown in Figure 17.29.
FLAC Version 5.0
17 - 23
JOB TITLE : .
(*10^2)
FLAC (Version 5.00)

0.700
LEGEND
0.500
17-Jan-05 16:40
step 11796
Flow Time
1.2066E+04
-3.667E+01 <x< 1.567E+02
-1.042E+02 <y< 8.917E+01
0.300

0.00E+00
2.50E+05
5.00E+05
7.50E+05
1.00E+06
1.25E+06
0.100
-0.100
-0.300

Boundary plot
-0.500
5E 1
-0.700
-0.900

-0.200
0.000
0.200
0.400
0.600
(*10^2)
0.800
1.000
1.200
1.400
Figure 17.29 Pore-pressure distribution after raising water table to y = 60 m

The mechanical response resulting from raising the water level is now calculated. The flowcalculation mode is turned off and the water bulk modulus is set to zero in the Settings / GW tool. The
mechanical-calculation mode is turned on in the Settings / Mech tool.
When the water level is raised, the water is re-established throughout the ground and the permeable
shotcrete liner, but not in the impermeable concrete liner. Thus, the water exerts a pressure in the
gap between the two liners. The pressure acts on both the inner concrete liner and the outer shotcrete
liner. The pressure applied to the shotcrete liner is applied using the in Situ / Apply tool. For the water
level raised to y = 60, the water pressure applied at the tunnel boundary varies from 660,000 Pa at
the invert to 540,000 Pa at the crown. The application of this pressure in the in Situ / Apply tool is
illustrated in Figure 17.30.
FLAC Version 5.0
17 - 24
Figure 17.30 Apply pressure to shotcrete liner

In order to apply a pressure to the inner concrete liner, it is necessary to write a FISH function
that calculates the forces at the liner nodes that correspond to the water pressure. A function,
APPLY GAP PRESS.FIS, listed in Section 17.5, performs this operation. This function accesses
liner-element variables associated with the concrete liner. The file STR.FIN is used to identify
and access the various structural data. (See Section 4 in the FISH volume for a description of the
use of .FIN files to access the FLAC data structure.) The x- and y-direction forces are calculated
corresponding to the water pressure at the depth of each liner node along the concrete liner. These
forces are then added to the nodal forces at the offset locations identified by the symbols $kndlo1
for the x-applied force, and $kndlo2 for the y-applied force.
APPLY GAP PRESS.FIS is executed from the FISH Editor. Three input parameters are required
for this function: y wtab, the y-coordinate of the water table; tuncen x, the x-coordinate of the
tunnel center; and tuncen y, the y-coordinate of the tunnel center.
After APPLY GAP PRESS.FIS is executed, the new equilibrium state is calculated using the
Run Solve
/
tool.
The tunnel moves upward and distorts slightly inward at the springline, as indicated by the magnifiedmesh plot shown in Figure 17.31. The vertical distance between the crown and invert increases by
approximately 2.5 cm due to the water pressure gradient, as shown in Figure 17.32. The resulting
axial forces in the liners are shown in Figure 17.33. The maximum stress is approximately 17 MPa
in the shotcrete liner and 10 MPa in the concrete liner. The moment distribution in the concrete
liner and the normal stress along the interface are displayed in Figure 17.34. Some yielding of the
rock occurs during the water table rise.
FLAC Version 5.0
17 - 25
JOB TITLE : .
FLAC (Version 5.00)
8.000
LEGEND
6.000
17-Jan-05 16:44
step 15639
Flow Time
1.2066E+04
-2.000E+00 <x< 1.600E+01
-9.000E+00 <y< 9.000E+00
4.000
2.000

Boundary plot
0
0.000
5E 0
-2.000
-4.000
-6.000
-8.000

-0.100
0.100
0.300
0.500
0.700
(*10^1)
0.900
1.100
1.300
1.500
Figure 17.31 Exaggerated grid distortion after raising water table
JOB TITLE : .
FLAC (Version 5.00)

(10
-02
LEGEND
17-Jan-05 16:44
step 15639
Flow Time
1.2066E+04
6.000
5.000
HISTORY PLOT
Y-axis :
4.000
5 vert_closure (FISH)
X-axis :
Number of steps
3.000
2.000
1.000
0.000
12
12
13
13
14
14
15
15
(10
03

Figure 17.32 Tunnel closure/opening after raising water table
FLAC Version 5.0
17 - 26
JOB TITLE : .
FLAC (Version 5.00)
8.000
LEGEND
6.000
17-Jan-05 16:44
step 15639
Flow Time
1.2066E+04
-2.000E+00 <x< 1.600E+01
-9.000E+00 <y< 9.000E+00
4.000
2.000
Liner Plot
Axial Force on
Structure
Max. Value
# 1 (Liner)
3.338E+06
# 2 (Liner)
3.824E+06
Boundary plot
0.000
-2.000
5E 0
-4.000
-6.000
-8.000

-0.100
0.100
0.300
0.500
0.700
(*10^1)
0.900
1.100
1.300
1.500
Figure 17.33 Axial forces in liners after raising water table
JOB TITLE : .
FLAC (Version 5.00)
8.000
LEGEND
6.000
17-Jan-05 16:44
step 15639
Flow Time
1.2066E+04
-2.000E+00 <x< 1.600E+01
-9.000E+00 <y< 9.000E+00
4.000
2.000
Boundary plot
0
5E 0
0.000
Liner Plot
Moment
on
Structure
Max. Value
# 1 (Liner) -1.568E+04
# 2 (Liner)
3.706E+05
Normal Stress (alternate)
on Interface # 1
Max Value = 8.340E+05
-2.000
-4.000
-6.000
-8.000

-0.100
0.100
0.300
0.500
0.700
(*10^1)
0.900
1.100
1.300
1.500
Figure 17.34 Moments in concrete liner and normal stress along interface after
raising water table
FLAC Version 5.0
17 - 27
17.2.6 Step 6: Delete Shotcrete Liner

In the last step of this analysis, the shotcrete liner is deleted to represent the degradation of the liner.
The shotcrete elements are deleted in the Structure / Segments tool. The shotcrete elements (elements 1
through 44) are highlighted with the mouse and deleted, as shown in Figure 17.35.
The final axial load in the concrete liner, after the shotcrete is deleted, is shown in Figure 17.36.
The maximum force corresponds to an axial stress of approximately 17 MPa.
Figure 17.35 Delete shotcrete-liner elements in
Structure
/ Segments tool
FLAC Version 5.0
17 - 28
JOB TITLE : .
FLAC (Version 5.00)
8.000
LEGEND
6.000
17-Jan-05 16:47
step 18251
Flow Time
1.2066E+04
-2.000E+00 <x< 1.600E+01
-9.000E+00 <y< 9.000E+00
4.000
2.000
Liner Plot
Axial Force on
Structure
Max. Value
# 1 (Liner)
6.692E+06
Boundary plot
0
0.000
5E 0
-2.000
-4.000
-6.000
-8.000

-0.100
0.100
0.300
0.500
0.700
(*10^1)
0.900
1.100
1.300
1.500
Figure 17.36 Axial force in concrete liner after deleting shotcrete liner
17.3 Reference
FLAC Version 5.0
17 - 29
17.4 Data File LINER.DAT

;Title:Lined tunnel construction
;... STATE: INITIAL ....
config gwflow ats extra 5
grid 80,60
gen 0.0,-80.0 0.0,-10.0 20.0,-10.0 20.0,-80.0 i 2 14 j 1 25
gen 20.0,-80.0 20.0,-10.0 120.0,-10.0 120.0,-80.0 i 14 44 j 1 25
gen 0.0,-10.0 0.0,10.0 20.0,10.0 20.0,-10.0 i 45 81 j 1 37
gen 20.0,-10.0 20.0,10.0 120.0,10.0 120.0,-10.0 i 14 44 j 25 37
gen 0.0,10.0 0.0,65.0 20.0,65.0 20.0,10.0 i 2 14 j 37 61
gen 20.0,10.0 20.0,65.0 120.0,65.0 120.0,10.0 i 14 44 j 37 61
gen circle 0.0,0.0 6.0
gen line 0.0,60.0 120.0 60.0
group ground notnull
model mohr notnull group ground
friction=20.0 dilation=0.0 tension=0.0 notnull group ground
prop por 0.3 perm 1e-10 notnull
fix y i 2 44 j 1
fix x i 44
fix x i 2 j 1 25
fix x i 45 j 1 37
fix x i 2 j 37 61
initial syy -3248000.0 var 0.0,3248000.0
initial sxx -1624000.0 var 0.0,1624000.0
initial szz -1624000.0 var 0.0,1624000.0
set gravity=10.0
set flow=off
solve elastic
save initial.sav
;*** BRANCH: BRANCH A ****
FLAC Version 5.0
17 - 30
;... STATE: SYYPLOT ....

set echo off
call fill ex1 syy.fis
fill ex1 syy
set echo off
call Extrap.fis
set gp avg=0
extrap to gp
save syyplot.sav
;*** BRANCH: BRANCH B ****
restore initial.sav
;... STATE: RELAX 30 ....
group delete null
set echo off
call relax.fis
set ninc=100 rstart=1.0 rstop=0.7
relax ini
set echo off
call apply rf.fis
set ib ap=45 ie ap=56 jb ap=8 je ap=30
apply rf
fix x i 45 j 30
fix x i 45 j 8
set =large
set echo off
call vert closure.fis
vert closure
history 5 vert closure
solve
save relax 30.sav
;... STATE:
struct node
struct node
struct node
struct node
struct node
struct node
FLAC Version 5.0
RELAX 100 ....

1 grid 45,8
2 grid 46,8
3 grid 47,8
4 grid 48,8
5 grid 48,9
6 grid 49,9
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
node 7 grid 50,9

node 8 grid 50,10
node 9 grid 51,10
node 10 grid 52,10
node 11 grid 52,11
node 12 grid 53,11
node 13 grid 53,12
node 14 grid 54,12
node 15 grid 54,13
node 16 grid 54,14
node 17 grid 55,14
node 18 grid 55,15
node 19 grid 55,16
node 20 grid 56,16
node 21 grid 56,17
node 22 grid 56,18
node 23 grid 56,19
node 24 grid 56,20
node 25 grid 56,21
node 26 grid 56,22
node 27 grid 55,22
node 28 grid 55,23
node 29 grid 55,24
node 30 grid 54,24
node 31 grid 54,25
node 32 grid 54,26
node 33 grid 53,26
node 34 grid 53,27
node 35 grid 52,27
node 36 grid 52,28
node 37 grid 51,28
node 38 grid 50,28
node 39 grid 50,29
node 40 grid 49,29
node 41 grid 48,29
node 42 grid 48,30
node 43 grid 47,30
node 44 grid 46,30
node 45 grid 45,30
liner begin node 1
liner begin node 2
liner begin node 3
liner begin node 4
liner begin node 5
liner begin node 6
liner begin node 7
end
end
end
end
end
end
end
node
node
node
node
node
node
node
2
3
4
5
6
7
8
17 - 31
seg
seg
seg
seg
seg
seg
seg
1
1
1
1
1
1
1
prop
prop
prop
prop
prop
prop
prop
5001
5001
5001
5001
5001
5001
5001
FLAC Version 5.0
17 - 32
struct liner begin node 8 end node 9 seg 1 prop 5001

struct prop 5001
struct prop 5001 e 5E9 area 0.2 I 6.667E-4 thickness 0.2 pratio 0.2
struct node 1 fix r
set echo off
call relax.fis
set ninc=100 rstart=1.0 rstop=0.0
relax ini
set echo off
FLAC Version 5.0
17 - 33
call apply rf.fis

set ib ap=45 ie ap=56 jb ap=8 je ap=30
apply rf
fix x i 45 j 30
fix x i 45 j 8
solve
save relax 100.sav
;... STATE:
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
struct node
CONCRETE LINER ....

46 -0.009999999,-5.934354
47 0.5403872,-5.909409
48 1.0677705,-5.83766
49 1.3934847,-5.7689834
50 1.7148663,-5.6822453
51 2.2164936,-5.507348
52 2.6691952,-5.3049097
53 2.9959185,-5.12955
54 3.3110964,-4.934328
55 3.6289055,-4.7082486
56 3.9312108,-4.4618177
57 4.221777,-4.1916423
58 4.4936547,-3.9028683
59 4.741998,-3.6024983
60 4.9692087,-3.2874517
61 5.1648974,-2.975049
62 5.3416204,-2.650305
63 5.543802,-2.2015164
64 5.71753,-1.7040727
65 5.803216,-1.384472
66 5.870646,-1.0615453
67 5.9404373,-0.53864527
68 5.9640603,-0.003078309
69 5.9402924,0.5325633
70 5.870383,1.0556496
71 5.8029413,1.3787708
72 5.717355,1.6985627
73 5.5437098,2.196513
74 5.3415146,2.645681
75 5.1648083,2.9705613
76 4.969033,3.2831426
77 4.741786,3.598347
78 4.4934635,3.8988612
79 4.221613,4.187811
80 3.9311185,4.4582114
81 3.6289046,4.70485
82 3.311182,4.931169
FLAC Version 5.0
17 - 34
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
struct
node 83 2.9960687,5.1266155
node 84 2.6694071,5.30218
node 85 2.2167227,5.504829
node 86 1.7150635,5.6798687
node 87 1.3936529,5.7666674
node 88 1.0678986,5.8353524
node 89 0.54044735,5.907067
node 90 -0.010000001,5.931989
liner begin node 46 end node 47
FLAC Version 5.0
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
seg
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
prop
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
5002
17 - 35

struct prop 5002
struct prop 5002 density 2500.0 e 2.5E10 area 0.4 I 0.005333 &
thickness 0.4 pratio 0.2
solve
save concrete liner.sav
;... STATE: WATER ....
initial pp 1400000.0 var 0.0,-1400000.0 i 2 44 j 1 59
initial pp 700000.0 var 0.0,-200000.0 i 45 81 j 1 37
set mechanical=off
set flow=on
water bulk=10000.0
solve
save water.sav
;... STATE: GAP PRESSURE ....
set flow=off
water bulk=0.0
set mechanical=on
apply pressure 660000.0 var 0.0 -120000.0 from 45,8 to 45,30
set echo off
call apply gap press.fis
set y wtab=60.0 tuncen x=0.0 tuncen y=0.0
apply gap press
history reset
history 5 vert closure
solve
FLAC Version 5.0
17 - 36
save gap pressure.sav

;... STATE: DELETE SHOTCRETE ....
FLAC Version 5.0
17 - 37

solve
save delete shotcrete.sav
;plot name: applied tractions
plot hold bound apply max 500000.0
;plot name: tunnel disp
plot hold history 4 line 5 line
;plot name: shotcrete axial forces
plot hold struct liner axial fill max 5.0E7 white bound
;plot name: interface stresses
plot hold bound struct liner moment 2 fill max 2000000.0 iface 1 altns &
fill white
;plot name: pp contours
;plot name: grid
plot hold grid magnify 10.0 bound
FLAC Version 5.0
17 - 38
17.5 Data File APPLY GAP PRESS.FIS

;Name:apply gap press
;Diagram:
;Input:y wtab/float/60.0/y-coordinate of water table
;Input:tuncen x/float/0.0/x-coordinate of tunnel center
;Input:tuncen y/float/0.0/y-coordinate of tunnel center
call str.fin
def Apply Gap Press
strp
= str pnt
sp=imem(strp+$ksels)
;pointer to struc element list
loop while sp # 0
;loop through all struc elements
if imem(sp+$kelcod2)=5
;check if it is a liner element
pn1=imem(sp+$keln1)
;node 1 of liner elem.
pn2=imem(sp+$keln2)
;node 2 of liner elem.
if imem(pn1+$kndcod) = 1 ;not linked to gridpoint means inner liner
n1id=imem(pn1+$kndid)
n2id=imem(pn2+$kndid)
n1xcord=fmem(pn1+$kndx)
n1ycord=fmem(pn1+$kndy)
n2xcord=fmem(pn2+$kndx)
n2ycord=fmem(pn2+$kndy)
nxdif=abs(n1xcord-n2xcord)
nydif=abs(n1ycord-n2ycord)
sslope=atan(nxdif/nydif)
;slope of liner
shlength=fmem(sp+$kell)/2. ;half-length of liner
if n1xcord>=tuncen x
unx =-cos(sslope)
else
unx =cos(sslope)
end if
if n1ycord>=tuncen y
uny =-sin(sslope)
else
uny =sin(sslope)
end if
;node 1
n1depth=y wtab-n1ycord
;depth of node
n1pp = -ygrav*n1depth*wdens ;pore pressure
n1fmag = shlength * n1pp
;equivalent force on node
;from half of liner element
n1xforce
= n1fmag * unx
n1yforce
= n1fmag * uny
fmem(pn1+$kndlo1)=fmem(pn1+$kndlo1)+n1xforce
;add xforce
fmem(pn1+$kndlo2)=fmem(pn1+$kndlo2)+n1yforce
;add yforce
;node 2
FLAC Version 5.0
n2depth=y wtab-n2ycord
n2pp = -ygrav*n2depth*wdens
n2fmag = shlength * n2pp
n2xforce
= n2fmag * unx
n2yforce
= n2fmag * uny
fmem(pn2+$kndlo1)=fmem(pn2+$kndlo1)+n2xforce
fmem(pn2+$kndlo2)=fmem(pn2+$kndlo2)+n2yforce
end if
end if
sp = imem(sp)
end loop
end
17 - 39
;add xforce
;add yforce
FLAC Version 5.0
17 - 40
FLAC Version 5.0
Seismic Analysis of an Embankment Dam
18 - 1
18 Seismic Analysis of an Embankment Dam

An analysis of the seismic performance of an embankment dam should consider static-equilibrium
and coupled groundwater conditions, as well as fully dynamic processes, with calculations for the
state of stress prior to seismic loading, the reservoir elevation and groundwater conditions, the
mechanical behavior of the foundation and embankment soils including the potential for liquefaction, and the site-specific ground motion response. This example presents a FLAC model for an
embankment dam that demonstrates a procedure to incorporate these processes and calculations in
the seismic analysis.
The example is a simplified representation of a typical embankment dam geometry. The dam is
130 ft. high and 1120 ft. long and is constructed above a layered foundation of sandstone and
shale materials. The crest of the dam is at elevation 680 ft. when the seismic loading is applied.
The embankment materials consist of a low permeability clayey sand core zone with upstream and
downstream shells of gravelly clayey sands. The materials in this analysis are defined as foundation
soils 1 and 2 and embankment soils 1 and 2, as depicted in Figure 18.1.
JOB TITLE : Embankment Dam
(*10^3)
1.500
FLAC (Version 5.00)

1.300
LEGEND
30-Mar-04 10:41
step
0
-8.793E+01 <x< 1.888E+03
-4.472E+02 <y< 1.529E+03
1.100
0.900
Embankment Dam Soils

foundation:soil 2
foundation:soil 1
embankment:soil 1
embankment:soil 2
0.700
0.500
0.300
0.100
-0.100
-0.300

0.100
0.300
0.500
0.700
0.900
(*10^3)
1.100
1.300
1.500
1.700
Figure 18.1 Embankment dam

The properties for the different soil types are listed in Table 18.1. The embankment soils are
considered to be susceptible to liquefaction during a seismic event. The liquefaction condition is
determined for embankment soils 1 and 2 in terms of standard penetration test (SPT) results. An
average normalized SPT blow count, (N1 )60 , of 15 is estimated for both of these soils.
FLAC Version 5.0
18 - 2
Table 18.1 Properties for foundation and embankment soils

Foundation
moist unit weight (pcf)
drained Youngs modulus (ksf)
drained Poissons ratio
drained cohesion (psf)
drained friction angle (degrees)
dilation angle (degrees)
porosity
hydraulic conductivity (ft/sec)
Embankment
Soil 1
Soil 2
Soil 1
Soil 2
125
12,757
0.3
83.5
40
0
0.3
3.3 106
125
12,757
0.3
160
40
0
0.3
3.3 107
113
6,838
0.3
120
35
0
0.3
3.3 106
120
6,838
0.3
120
35
0
0.3
3.3 107
The dynamic characteristics of all of the soils in this model are assumed to be governed by the
modulus reduction factor (G/Gmax ) and damping ratio () curves, as shown in Figures 18.2 and
18.3, and denoted by the Shake91 legend. These curves are considered to be representative
of clayey soils with an average unit weight of 125 pcf and an average shear modulus of 6270
ksf; the data are derived from the input file supplied with SHAKE91 (for more information see
http://nisee.berkeley.edu/software/shake91/).
The earth dam is subjected to seismic loading representative of the 1987 Loma Prieta earthquake
in California. The earthquake input motion for this model is taken from that recorded at the left
abutment of the Lexington Dam during the Loma Prieta earthquake and, for this analysis, the record
is assumed to correspond to the acceleration at elevation 400 ft, directly beneath the dam.* The
record is provided in the file named ACC.HIS. The estimated peak acceleration is approximately
5.5 ft/sec2 (or 0.17 g), and the duration is approximately 40 sec. The record is shown in Figure 18.4.
A Fast Fourier Transform analysis of the acceleration record (using FFT.FIS in Section 3 in the
FISH volume) results in a power spectrum as shown in Figure 18.5. This figure indicates that the
highest frequency is less than 10 Hz.
* It is assumed that the input acceleration record has been modified to account for the propagation
from rock, where the record is recorded, to the depth it is applied in the model. This can be obtained
from a FLAC or SHAKE analysis, for example.
FLAC Version 5.0
18 - 3
The purpose of this example is to demonstrate the recommended procedure to perform a seismic
analysis of this embankment dam. This analysis is divided into four stages:
1. Establish the initial state of stress of the embankment dam and foundation after initial
construction.
2. Establish the state of stress with the reservoir level raised to its full height.
3. Perform a seismic calculation assuming the soils do not liquefy (represented as a MohrCoulomb material).
4. Perform a seismic calculation assuming the soils can liquefy (represented as a Byrne
material).
Each stage is described separately in the modeling procedure in the next section. A listing of the
data record for this model, EARTHDAM.DAT, is given in Section 18.6.
FLAC Version 5.0
18 - 4
Figure 18.2 Modulus reduction curve for clayey soils (from SHAKE91 data)
Figure 18.3 Damping ratio curve for clayey soils (from SHAKE91 data)
FLAC Version 5.0
18 - 5
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
2-Apr-04 14:59
step
0
4.000
Acceleration Record
(ft/sec2 vs sec)
2.000
0.000
-2.000
-4.000
10
15
20
25
30
35
40

Figure 18.4 Horizontal acceleration time history
JOB TITLE : .
FLAC (Version 5.00)

(10
-02
LEGEND
26-Mar-04 10:13
step
0
Power Spectrum
4.000
3.500
3.000
2.500
2.000
1.500
1.000
0.500
10
15
20
25

Figure 18.5 Power spectrum of input acceleration
FLAC Version 5.0
18 - 6

The FLAC model options selected for this analysis are shown activated in the Model Options dialog
displayed in Figure 18.6. The dynamic analysis and groundwater flow options are selected, and the
adjusted total stress option is included because the water level in the reservoir will be raised during
the simulation. Advanced constitutive models are also included in order to access the Byrne model,
which will be used for the liquefaction calculation phase. The Imperial system of units is specified
for this analysis.
Figure 18.6 Model options selected for the embankment dam example
The mesh size for the FLAC model is selected to ensure accurate wave transmission. Based upon
the elastic properties listed in Table 18.1, embankment soil 2 has the lowest shear wave speed
(840 ft/sec). If the largest zone size in the FLAC model is set to 10 ft. in order to provide reasonable
runtimes for this example, then the maximum frequency that can be modeled accurately is
f =
Cs
8.4 Hz
10 l
(18.1)
Before applying the seismic input record, it is filtered to remove frequencies above 5 Hz (by using
the FISH function FILTER.FIS described in Section 3 in the FISH volume). This filtering value
is selected to account for the reduction in shear wave speed that may occur in some of the materials
during the dynamic loading stage, as indicated in Figure 18.2. The acceleration history filtered at 5
Hz is shown in Figure 18.7, and the power spectrum for the filtered wave is shown in Figure 18.8.
Note that the difference between the unfiltered and filtered waves is minor (compare Figures 18.4
and 18.7).
FLAC Version 5.0
18 - 7
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
2-Apr-04 14:59
step
0
4.000
Acceleration Record - 5 Hz filter

(ft/sec2 vs sec)
2.000
0.000
-2.000
-4.000
10
15
20
25
30
35
40

Figure 18.7 Horizontal acceleration time history with 5 Hz filter
JOB TITLE : .
FLAC (Version 5.00)

(10
-02
LEGEND
2-Apr-04 14:59
step
0
Power Spectrum
4.000
3.500
3.000
2.500
2.000
1.500
1.000
0.500
10
15
20
25

Figure 18.8 Power spectrum of horizontal acceleration time history (with 5

Hz filter)
FLAC Version 5.0
18 - 8
The filtered wave is also checked for baseline drift (i.e., continuing residual displacement after the
motion has finished). The FISH function INT.FIS, described in Section 3 in the FISH volume, is
used to integrate the filtered acceleration record to produce the velocity and displacement waveforms
related to this acceleration. The resulting residual displacement is found to be approximately 0.3 ft.
A baseline correction is performed by adding a low frequency sine wave to the velocity record;
the sine wave parameters are adjusted so that the final displacement is zero. The uncorrected and
corrected resultant displacement histories are shown in Figure 18.9. (See Section 18.9.)
The data file INPUT.DAT, listed in Section 18.8, includes the different steps performed in filtering
the input acceleration record, integrating this record to produce velocity and displacement histories,
and correcting for baseline drift. The resultant, corrected velocity record, stored in table 104 in this
data file, is the input motion for this analysis.
JOB TITLE : .
FLAC (Version 5.00)

(10
-01
LEGEND
27-May-04 10:24
step
0
Table Plot
uncorrected disp.
3.000
2.000
corrected disp.
1.000
0.000
-1.000
-2.000
10
15
20
25
30
35
40

Figure 18.9 Uncorrected and corrected displacement histories

The dynamic calculation phase is performed using the large-strain mode in FLAC. When significant
deformation and distortion of the grid is anticipated, as in this example, it is important to minimize the
number of triangular-shaped zones in the mesh and, in particular, those along slope faces. Triangular
zones along slope faces are prone to become badly distorted during large-strain calculations, because
triangular zones do not contain overlaid sets of sub-zones. Quadrilateral-shaped zones contain two
FLAC Version 5.0
18 - 9
overlaid sets of sub-zones, which provide a more accurate calculation for materials undergoing
plastic yield. (See Sections 1.3.2 and 1.3.3.2 in Theory and Background.)*
A special FISH tool is provided in the Utility/FishLib library to assist with the creation of a mesh for
this model. The gentabletop tool, shown in the Fish Library dialog in Figure 18.10, is used to
transform a grid to fit an irregular upper surface defined by a table. The grid beneath the table is
adjusted to provide a uniform spacing of quadrilateral zones below the table surface.
Figure 18.10 gentabletop FISH function in TABTOP.FIS
* If a badly distorted zone causes a calculation to stop prematurely in a mesh containing a few
triangular zones, it may be possible to prevent this by increasing the strength of the individual zone.
This should not significantly affect the model results, provided that the strengths of only a few zones
are changed.
FLAC Version 5.0
18 - 10
The grid zoning is defined first using the Build/Simple tool. The grid parameters selected for this
model are shown in the Edit Grid dialog of the Simple tool, in Figure 18.11. The number of zones
corresponds to a zone size of 10 ft.
Figure 18.11 Grid parameters entered in the
Simple
dialog
After this grid is created, the Utility/Table tool is used to define the slope of the embankment dam.
Figure 18.12 shows the pairs of x- and y-values that are entered in the Edit Table Points dialog to
define the surface.
After this table is defined, the gentabletop tool is executed from the Utility/FishLib library to
create the grid. The resulting mesh is shown in Figure 18.13. Note that the grid is totally composed
of quadrilateral-shaped zones.
Figure 18.12 Edit Table Points dialog showing x- and y-values defining embankment dam surface
FLAC Version 5.0
18 - 11
JOB TITLE : .
(*10^3)
1.500
FLAC (Version 5.00)

1.300
LEGEND
9-Apr-04 13:01
step
0
-8.800E+01 <x< 1.888E+03
-4.480E+02 <y< 1.528E+03
1.100
0.900
User-defined Groups
Grid plot
0
0.700
5E 2
0.500
0.300
0.100
-0.100
-0.300

0.100
0.300
0.500
0.700
0.900
(*10^3)
1.100
1.300
1.500
1.700
Figure 18.13 Mesh created with gentabletop tool

The foundation soil layers and the embankment core and shell regions are delineated by lines
generated using the Alter/Shape tool. Then, the different materials and properties, listed in Table 18.1,
are specified, corresponding to Mohr-Coulomb materials, and assigned using the Material/Assign and
Material/GwProp
tools. The resulting model is shown in Figure 18.14. Note that some triangular
zones are created within the mesh when the different soil regions are defined (see Figure 18.15).
Triangular zones are also created at the slope toe and crest. It is difficult to eliminate triangular
zones completely in this model. However, there are only a small number of these zones along the
slope face, and the strengths of these zones can be readily adjusted if there is a distortion problem.
The model state, after the geometry shaping is complete, is saved in the GIIC Project Tree with the
name EDAM1.SAV, and after the materials have been added, with the name EDAM2.SAV. The
model is now ready to begin the four analysis stages.
FLAC Version 5.0
18 - 12
JOB TITLE : .
(*10^3)
1.500
FLAC (Version 5.00)

1.300
LEGEND
7-Apr-04 10:00
step
0
-9.938E+01 <x< 1.899E+03
-4.594E+02 <y< 1.539E+03
1.100
0.900

foundation:soil 2
foundation:soil 1
embankment:soil 1
embankment:soil 2
Grid plot
0
0.700
0.500
5E 2
0.300
0.100
-0.100
-0.300

0.100
0.300
0.500
0.700
0.900
(*10^3)
1.100
1.300
1.500
1.700
Figure 18.14 Embankment dam model with foundation and embankment soils
assigned
JOB TITLE : .
(*10^2)
FLAC (Version 5.00)

7.750
LEGEND
7-Apr-04 10:00
step
0
5.628E+02 <x< 9.772E+02
3.978E+02 <y< 8.122E+02
7.250
6.750

foundation:soil 2
foundation:soil 1
embankment:soil 1
embankment:soil 2
Grid plot
6.250
5.750
1E 2
5.250
4.750
4.250

6.000
6.500
7.000
7.500
(*10^2)
8.000
8.500
Figure 18.15 Close-up view of embankment dam model
FLAC Version 5.0
9.000
9.500
18 - 13
18.2.1 Stage 1: Establish Initial State of Stress

The analysis is started from the state before the embankment is constructed. The construction
process may affect the stress state, particularly if excess pore pressures develop in the soils and
do not dissipate completely during the construction stages. The embankment can be constructed
in stages with a consolidation time specified in the FLAC model, if pore-pressure dissipation is a
concern. In this example, the excess pore pressures are assumed to dissipate before a new lift of
embankment material is placed.
It should be noted that staged modeling of the embankment lift construction also provides a better
representation of the initial static shear stresses in the embankment. This is important, particularly in a liquefaction analysis, because the initial static shear stresses can affect the triggering of
liquefaction. In this simplified example, the embankment is placed in one stage. However, it is
recommended that the lift construction stages be simulated as closely as practical in order to provide
a realistic representation of the initial stress state.
The embankment materials are temporarily removed from the model by using the Material/Cut&Fill
tool. These materials will be added back after the calculation for the initial equilibrium state of the
foundation. The boundary conditions are specified using the In Situ/Fix tool. Note that the bottom of
the model is fixed from movement in both the x- and y-directions. If a roller boundary is specified
along the bottom boundary, then the foundation is free to slide along the base when the embankment
is constructed, which may cause unrealistic failure modes.
The water density of 1.94 slugs/ft3 and gravitational magnitude of 32.2 ft/sec2 are assigned, and
fluid flow and dynamic analysis modes are turned off in the global Settings tools.
The most efficient way to achieve an equilibrium stress state in a saturated, horizontally layered soil
is to use the special FISH tool, ininv, provided in the Utility/FishLib library. This function calculates
the pore pressures and stresses automatically for a model containing a phreatic surface. The function
requires the phreatic surface height (wth = 550 in this example), and the ratios of horizontal to
vertical effective stresses (assumed to be k0x = k0z = 0.5 in this example). The pore pressure,
total stress and effective stress distributions are then calculated automatically, accounting for the
different soil unit weights, and the position of the water table. The equilibrium state is checked
(using the SOLVE elastic option in the Run/Solve tool). Figure 18.16 shows the initial pore-pressure
distribution in the foundation soils. This state is saved in the Project Tree as EDAM3.SAV.
The embankment materials can be added to the model in stages to simulate the construction processing, by using the Material/Cut&Fill tool. In this example, both embankment soils 1 and 2 are added
simultaneously, and pore pressures are assumed not to change. The displacements resulting from
adding the embankment in one step are shown in Figure 18.17. The saved state at this stage in
named EDAM4.SAV.
FLAC Version 5.0
18 - 14
JOB TITLE : .
(*10^3)
1.400
FLAC (Version 5.00)

LEGEND
31-Jan-05 12:33
step
1855
-1.000E+02 <x< 1.900E+03
-5.250E+02 <y< 1.475E+03
1.000

0.00E+00
1.00E+03
2.00E+03
3.00E+03
4.00E+03
5.00E+03
6.00E+03
7.00E+03
8.00E+03
9.00E+03
0.600
0.200

Boundary plot
0
-0.200
5E 2

0.200
0.600
1.000
1.400
1.800
(*10^3)
Figure 18.16 Pore pressure distribution in foundation soils
JOB TITLE : .
(*10^3)
FLAC (Version 5.00)
1.400
LEGEND
31-Jan-05 12:33
step
3502
-1.000E+02 <x< 1.900E+03
-4.600E+02 <y< 1.540E+03
1.000
-3.00E-01
-2.50E-01
-2.00E-01
-1.50E-01
-1.00E-01
-5.00E-02
0.00E+00
0.600
0.200
-0.200

0.200
0.600
1.000
1.400
1.800
(*10^3)
Figure 18.17 Displacements induced by embankment construction
FLAC Version 5.0
18 - 15
Note that the model is run in small-strain mode up to this stage and, consequently, the gridpoint
positions are not changed. This is done so that the embankment crest elevation (680 ft.) does not
change. If significant deformation occurs during embankment construction, making it necessary
to perform this stage in large-strain mode, then the initial embankment crest elevation for the
embankment zones (prior to construction) would need to be raised in order to obtain a specified
elevation after construction.
18.2.2 Stage 2: Establish the State of Stress with the Reservoir Level Raised
The earthquake motion is considered to occur when the reservoir level is at full pool (i.e., at its
full height at elevation 670 ft.). For Stage 2 of the analysis, the pore pressure distribution through
the embankment and foundation soils is calculated for the reservoir raised to this height. The
In Situ/Apply
tool is used to set the pore pressure distribution on the upstream side of the embankment,
corresponding to the reservoir elevation at 670 ft. The mouse is dragged in this tool along the
upstream boundary starting from the 670 elevation (at gridpoint i = 70, j = 29) and ending at the
400 elevation at the bottom left corner of the model (at gridpoint i = 1, j = 1). The distribution
parameters, shown in the Apply value dialog displayed in Figure 18.18, produce a pore-pressure
distribution along this boundary that ranges from zero at elevation 670 ft. to 16,866.36 psf at
elevation 400 ft.
Figure 18.18 Pore pressure distribution parameters corresponding to a reservoir elevation of 670 ft. on the upstream side of the embankment
This calculation is first run in flow-only mode. The groundwater-flow calculation is turned on,
and a water bulk modulus of 4.1 106 psf is assigned, in the Settings/GW tool. The water modulus
corresponds to water with entrained air. The fast unsaturated-flow calculation funsat and water
bulk scaling with permeability and porosity fastwb are set in order to speed the calculation to
steady state flow. The mechanical calculation mode is turned off in the Settings/Mech tool. In the
In Situ/Fix
tool, the pore pressures are fixed at gridpoints along the downstream slope to allow
flow across this surface, and the pore-pressure fixity condition is removed for gridpoints at the
FLAC Version 5.0
18 - 16
foundation/embankment interface inside the model. The porosity and permeability values are also
specified for the embankment materials, in the Material/GwProp tool.
Figure 18.19 plots pore pressure histories at different locations in the model, indicating that constant values are reached for the equilibrium ratio limit. Figure 18.20 displays the pore-pressure
distribution through the embankment and foundation at steady state. The saved state at steady-state
flow is named EDAM5.SAV.
JOB TITLE : .
FLAC (Version 5.00)

(10
03
LEGEND
31-Jan-05 12:51
step 13669
Flow Time
1.5168E+09
6.000
HISTORY PLOT
Y-axis :
5.000
4.000

X-axis :
3.000
2.000
1.000
10
12
14
(10

Figure 18.19 Pore-pressure histories
FLAC Version 5.0
08
18 - 17
JOB TITLE : .
(*10^3)
FLAC (Version 5.00)
1.400
LEGEND
31-Jan-05 12:51
step 13669
Flow Time
1.5168E+09
-1.000E+02 <x< 1.900E+03
-4.600E+02 <y< 1.540E+03
1.000

0.00E+00
2.50E+03
5.00E+03
7.50E+03
1.00E+04
1.25E+04
1.50E+04
0.600
0.200

Boundary plot
0
5E 2
-0.200

0.200
0.600
1.000
1.400
1.800
(*10^3)
Figure 18.20 Pore-pressure distribution at steady state flow for reservoir raised
to 670 ft.
The static equilibrium state is now calculated for the new pore-pressure distribution. A pressure
distribution is applied along the upstream slope to represent the weight of the reservoir water. This
time a mechanical pressure is assigned in the In Situ/Apply tool. The pressure ranges from zero at
elevation 670 ft. (at gridpoint i = 70, j = 29) to 8120.8 psf at elevation 550 ft. at the toe of the
slope (at gridpoint i = 1, j = 29); the dialog is displayed in Figure 18.21. The groundwater-flow
calculation is turned off, and the water bulk modulus is set to zero (in the Settings/GW tool). The
mechanical calculation is turned on (in the Settings/Mech tool). The model is now solved for this
applied condition, and the resulting total vertical-stress contour plot for the model at this stage is
shown in Figure 18.22. The saved state at this stage is named EDAM6.SAV.
FLAC Version 5.0
18 - 18
Figure 18.21 Mechanical pressure distribution parameters corresponding to a

reservoir elevation of 670 ft. on the upstream side of the embankment
JOB TITLE : .
(*10^3)
FLAC (Version 5.00)
1.400
LEGEND
31-Jan-05 12:52
step 15076
Flow Time
1.5168E+09
-1.000E+02 <x< 1.900E+03
-4.600E+02 <y< 1.540E+03
1.000
YY-stress contours
-3.50E+04
-3.00E+04
-2.50E+04
-2.00E+04
-1.50E+04
-1.00E+04
-5.00E+03
0.00E+00
0.600
0.200

Boundary plot
0
5E 2
-0.200

0.200
0.600
1.000
1.400
1.800
(*10^3)
Figure 18.22 Total vertical-stress distribution at steady state flow for reservoir
raised to 670 ft.
FLAC Version 5.0
18 - 19
18.2.3 Stage 3: Seismic Calculation Assuming No Liquefaction

For Stage 3, the foundation and embankment soils continue to be simulated as Mohr-Coulomb
materials. Pore pressures can change in these materials due to dynamic volume changes induced
by the seismic excitation. However, pore pressure generation due to particle rearrangement does
not occur. In order for pore pressures to change as a result of volume change, the actual value of
water bulk modulus must be prescribed. The value of 4.1 106 psf is respecified for the water
bulk modulus. Note that the groundwater-flow mode is not active, because it is assumed that the
dynamic excitation occurs over a much smaller time frame than that required for pore pressures
to dissipate. (FLAC can carry out the groundwater flow calculation in parallel with the dynamic
calculation, if dissipation is considered important.)
It is important in a seismic calculation of this type to set the dynamic conditions for the model in a
specific order. This is because the free-field boundaries that are used to represent the lateral extent
of the far-field are assigned properties and initial conditions that are transferred from the main-grid
zones adjacent to the free-field boundaries. Any changes to these zones or gridpoints after the free
field is assigned are not seen by the free field.
The following sequence is recommended to assign conditions for a dynamic calculation, to ensure
that these conditions are applied properly.
1. In the Settings tools, turn on the dynamic calculation mode ( Dyna tool) and the large-strain
calculation mode ( Mech tool).
2. In the Utility/History tool, read in the seismic record, and convert it to a table if the record
is to be filtered. Use the FILTER.FIS FISH function in the Utility/FishLib tool to filter
the record. The baseline correction, if needed, is also performed at this point, before the
dynamic input is applied to the boundary. Note that the dynamic input can be specified
as either an acceleration (or velocity) boundary condition, or a stress (or force) boundary
condition.
3. In the Utility/History tool, select various parameters to monitor during the dynamic simulation, such as gridpoint velocities and zone pore pressures and stresses.
4. In the In Situ/Initial tool, initialize the displacements and velocities in the grid, and specify
the type of mechanical damping and parameters.
5. In the In Situ/Apply tool, press the Free-Field button to assign the free-field boundaries. Then,
assign the dynamic boundaries for the seismic loading.
The following conditions are set in this first dynamic simulation. The dynamic calculation mode is
turned on, using the Settings/Dnya tool, and the large-strain mode is selected in the Settings/Mech tool.
The acceleration history (ACC.HIS) is read into FLAC via the Read button in the Utility/History tool.
The Execute button should be pressed to execute the command. The history is then converted into a
table by pressing the History -> Table button in the Utility/History tool. The dialog shown in Figure 18.23
appears, and the acceleration history (previously assigned ID number 100) is converted into a table
(designated by ID number 100).
FLAC Version 5.0
18 - 20
Figure 18.23 Copy history into a table

The FISH function FILTER.FIS is now accessed from the Tables menu item in the Utility/FishLib
tool to filter the acceleration at 5 Hz. The filtered table is given the ID number 101, as shown in the
dialog in Figure 18.24. OK is pressed to execute this FISH function and create the filtered record.
Figure 18.24 Input for filter FISH function

In order to perform the baseline correction, the filtered acceleration in table 101 is integrated to
produce a velocity record (using FISH function INT.FIS accessed from the Tables menu item in
the Utility/FishLib tool) and stored in table 102. Then, a low frequency sine wave is added to this
velocity record to produce a final displacement of zero. The sine wave is given in BASELINE.FIS
in Section 18.2.4. The corrected velocity wave is stored in table 104.
The displacements and velocities in the model are initialized by pressing the Displmt & Velocity button
in the In Situ/Initial tool. Damping is not prescribed for this initial dynamic simulation. Velocity
histories are recorded at several gridpoints throughout the model via the Utility/History tool. These
histories are used to evaluate the dominant natural frequency of the model, when no additional
damping is prescribed.
FLAC Version 5.0
18 - 21
The model state is saved at this point, with these dynamic conditions set, and named EDAM7.SAV.
This will provide a convenient starting state for the dynamic analysis when damping is added.
The dynamic boundary conditions are now applied in the In Situ/Apply tool. First, the free-field
boundary is set for the side boundaries, by selecting the Free-Field button.
Next, the dynamic input is assigned to the bottom boundary. In this model, the foundation materials
are assumed to extend to a significant depth beneath the dam. Therefore, it is necessary to apply a
quiet (viscous) boundary along the bottom of the model to minimize the effect of reflected waves
at the bottom.
In order to apply quiet boundary conditions along the same boundary as the dynamic input, the
dynamic input must be applied as a stress boundary, because the effect of the quiet boundary will
be nullified if the input is applied as an acceleration (or velocity) wave. The filtered and baselinecorrected velocity record (in table 104) is converted into a shear stress boundary condition using a
two-step procedure:
1. Convert the velocity wave into a shear stress wave using the formula
s = 2( Cs ) vs
where:
Cs
vs
=
=
=
=
(18.2)
applied shear stress;

mass density;
speed of s-wave propagation through medium; and
input shear particle velocity.
Note that the factor of two in Eq. (18.2) accounts for the input energy dividing into
downward and upward propagating waves.
2. Monitor the x-velocity at the base of the model during the dynamic run to compare this
velocity to the input velocity (in table 104). If the model is shallow, the free surface can
cause an increase in the velocity at the base.* Some adjustment to the input stress wave
may be required in order to produce a velocity at the base that corresponds to the input
velocity.
This two-step procedure is applied as follows to prescribe the dynamic wave as a shear stress
boundary condition along the base for this example.
* This is a result of the velocity-doubling effect of the free surface. Note that the effect of a free
surface on the velocity extends beneath the surface to, approximately, a depth of one-fourth of the
wavelength of the wave transmitted through the medium. (This can be shown simply by applying the
wave to the base of a column of zones with a free surface, and monitoring the maximum amplitude
experienced by each zone as the wave travels through the column.)
FLAC Version 5.0
18 - 22
First, the Stress/sxy boundary condition type is selected in the In Situ/Apply tool, and the mouse is
dragged from the bottom-left corner of the model (gridpoint i = 1, j = 1) to the bottom-right corner
(i = 180, j = 1). The Assign button is pressed, which opens the Apply value dialog. The velocity
record, in table 104, is considered a multiplier, vs , for the applied value. The velocity record is
applied by checking the Table radio button, and selecting table number 104 as the multiplier.
The applied value for sxy in the Apply value dialog is first set to 2 Cs (from Eq. (18.2)), in which
and Cs correspond to the properties for foundation soil 2. In this example, this value produces
x-velocities at the base that are greater than the input velocities in table 104. By reducing the
value to 1.75 Cs , the calculated peak velocities along the base are found to approximate the input
velocity more closely. The input selections for the Apply value dialog are shown in Figure 18.25.
The bottom of the model is considered to be a soft foundation. Quiet boundary conditions are
assigned in both the x- and y-directions by first selecting the xquiet button and dragging the mouse
along the bottom boundary, and then selecting the yquiet button and repeating the procedure.
The model state is saved again at this point and named EDAM8.SAV.
Figure 18.25 Apply shear stress boundary condition in Apply value dialog
The model calculation is performed for a dynamic time period of 20 seconds, which is sufficient to
identify the dominant natural frequency of the model. A representative plot of one of the x-velocity
histories is shown in Figure 18.26. This plot indicates that the dominant natural frequency is
approximately 1.25 Hz. This is also shown by performing a Fast Fourier Transform analysis of this
velocity record. The velocity history is first converted into a table (using the History -> Table button in
the Utility/History tool). Then the FFT.FIS function, described in Section 3 in the FISH volume,
is used to calculate the power spectrum. The result is shown in Figure 18.27. The undamped
calculation is saved as EDAM9.SAV.
FLAC Version 5.0
18 - 23
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
31-Jan-05 13:20
step 54454
Flow Time
1.5168E+09
HISTORY PLOT
Y-axis :
102 X velocity ( 80, 18)
X-axis :
12 Dynamic time
0.800
0.400
0.000
-0.400
-0.800
10
12
14
16
18

Figure 18.26 x-velocity time history at gridpoint (80,18) undamped
JOB TITLE : .
FLAC (Version 5.00)

(10
-02
LEGEND
31-Jan-05 13:20
step 54454
Flow Time
1.5168E+09
Power Spectrum - yvel(80,18)
2.000
1.600
1.200
0.800
0.400
10
12
14

Figure 18.27 Power spectrum of x-velocity time history at gridpoint (80,18)

undamped
FLAC Version 5.0
18 - 24
Additional damping is now prescribed for the model in order to damp the natural oscillation modes
identified from the undamped simulation. Hysteretic damping is applied corresponding to the
dynamic characteristics represented by the (G/Gmax ) and () curves shown in Figures 18.2 and
18.3. These figures also show a comparison of the (G/Gmax ) and () variations to those computed
using the default hysteretic model in FLAC. The selected parameters (L1 = -3.156 and L2 = 1.904)
for the default model produce the FLAC curves shown in these figures.
The damping is applied at the model state EDAM7.SAV by double-clicking on this state name
in the Project Tree. Hysteretic damping is assigned in the In Situ/Initial tool. The dialog shown in
Figure 18.28 is opened by selecting the Zones type, checking the Hysteretic Damping menu item, and
then Assign , to assign the same values for all zones in the model.
Hysteretic damping does not completely damp high frequency components, so a small amount of
stiffness-proportional Rayleigh damping is also applied. A value of 0.5% at the dominant frequency
(1.25 Hz) is assigned in the Rayleigh damping parameters dialog shown in Figure 18.29. Note that
Rayleigh damping is applied by selecting the GPs type, and then Dynamic Damping in the In Situ/Initial
tool.
Figure 18.28 Hysteretic damping parameters
Figure 18.29 Rayleigh damping parameters used with hysteretic damping
FLAC Version 5.0
18 - 25
After Execute is pressed in the In Situ/Initial tool, a new branch is created in the project tree to
perform a calculation including the additional hysteretic damping (and the small amount of stiffnessproportional Rayleigh damping). The dynamic boundary conditions must be applied again from the
In Situ/Apply
tool. The free-field is applied on the side boundaries, and sxy-stress history and quiet
boundaries are applied at the base, in the same way as for the undamped simulation. The model
state is saved at this stage as EDAM10.SAV.
A new simulation is now made for a dynamic time of 40 seconds. Note that the dynamic timestep
used for this calculation is approximately 3.1 104 seconds. Movement of the embankment on the
upstream slope is identified at 40 seconds in this simulation. This is shown in the x-displacement
contour plot, in Figure 18.31, and the shear-strain increment contour plot, in Figure 18.33. Note
that the maximum shear strain is less than 1% along the upstream slope face at 40 seconds. The
movement of gridpoint (62,29) along the upstream slope is shown in Figure 18.35.* The upstream
slope at this point settles approximately 2.5 in., and shifts horizontally upstream approximately 4.5
in.
The pore pressure and effective vertical stress histories in Figure 18.37, recorded at (i = 49, j =
23) near the upstream face, illustrate the minor pore-pressure change in the embankment materials
during the seismic loading. The model state is saved at this stage as EDAM11.SAV.
The dynamic simulation stage is now repeated using only Rayleigh damping. The center frequency
for Rayleigh damping is 1.25 Hz, as determined from the undamped analysis (see Figure 18.27).
The fraction of critical damping is determined by running simple element tests with the different
foundation and embankment soil properties for the dynamic loading, as defined by Figure 18.7. The
fraction of critical damping is adjusted in these tests to approximate the shear modulus and damping
ratio at the cyclic strain levels for clayey soils, as defined by the SHAKE91 data in Figures 18.2 and
18.3. For this case, the fraction of critical damping is estimated to be 0.05. The Rayleigh damping
parameters are specified as shown by the dialog in Figure 18.30.
Figure 18.30 Rayleigh damping parameters for Rayleigh damping simulation

* Note that this movement is a relative displacement referenced to the base of the model at gridpoint
(62,1). See RELDISPX.FIS in Figure 18.3.
FLAC Version 5.0
18 - 26
Note that one set of Rayleigh damping parameters are assumed for all the soils in this model. In
general, different damping parameters may be needed to represent different damping behavior of
the different materials and positions within the foundation and embankment. The spatial variation
in damping can be prescribed with the INITIAL dy damp command.
Also note that, for this case, with mass- and stiffness-proportional Rayleigh damping of 5% at the
natural frequency of 1.25 Hz, the limiting timestep is approximately 5.2 105 seconds. This
timestep is approximately six times smaller than that for hysteretic damping. The model state is
saved at this stage as EDAM12.SAV.
If Rayleigh damping alone is used, the results are very similar. Figure 18.32 plots the xdisplacement contours at 40 seconds for Rayleigh damping. Figure 18.34 shows the shear-strain
increment contours at this time. Both plots compare reasonably well with those using hysteretic
damping (compare to Figures 18.31 and 18.33). Figure 18.36 plots the relative movement at gridpoint (62,29). The relative vertical displacement component is approximately 1.5 in. downward,
and the horizontal component is approximately 2.5 in. in the upstream direction at this position,
which is somewhat less than that for hysteretic damping (compare to Figure 18.35).
Pore pressure and effective vertical stress histories for the Rayleigh damping run are also similar to
those for the hysteretic damping run (compare Figure 18.38 to Figure 18.37).
The Rayleigh damping run at 40 seconds is saved as EDAM13.SAV.
This comparison demonstrates the substantial benefit of hysteretic damping. The results are comparable to those using Rayleigh damping for similar damping levels, and the runtime with hysteretic
damping is greatly reduced.
FLAC Version 5.0
18 - 27
JOB TITLE : .
(*10^2)
FLAC (Version 5.00)

8.000
LEGEND
1-Feb-05 11:52
step 142640
Flow Time
1.5168E+09
2.205E+02 <x< 9.219E+02
1.851E+02 <y< 8.865E+02
7.000
6.000
-2.00E-01
-1.00E-01
0.00E+00
1.00E-01
2.00E-01
5.000

Boundary plot
0
4.000
2E 2
3.000

2.000
2.500
3.500
4.500
5.500
(*10^2)
6.500
7.500
8.500
Figure 18.31 x-displacement contours at 40 seconds

JOB TITLE : .
(*10^2)
FLAC (Version 5.00)

8.000
LEGEND
2-Feb-05 7:27
step 812615
Flow Time
1.5168E+09
2.205E+02 <x< 9.219E+02
1.851E+02 <y< 8.865E+02
7.000
6.000
-2.00E-01
-1.00E-01
0.00E+00
1.00E-01
2.00E-01
5.000

Boundary plot
0
4.000
2E 2
3.000

2.000
2.500
3.500
4.500
5.500
(*10^2)
6.500
7.500
8.500

FLAC Version 5.0
18 - 28
JOB TITLE : .
(*10^3)
FLAC (Version 5.00)
1.400
LEGEND
1-Feb-05 11:52
step 142640
Flow Time
1.5168E+09
-1.000E+02 <x< 1.900E+03
-4.600E+02 <y< 1.540E+03
1.000

0.00E+00
2.00E-03
4.00E-03
6.00E-03
8.00E-03
0.600
0.200
-0.200

0.200
0.600
1.000
1.400
1.800
(*10^3)
Figure 18.33 Shear-strain increment contours at 40 seconds

JOB TITLE : .
(*10^3)
FLAC (Version 5.00)
1.400
LEGEND
2-Feb-05 7:27
step 812615
Flow Time
1.5168E+09
-9.996E+01 <x< 1.900E+03
-4.599E+02 <y< 1.540E+03
1.000

0.00E+00
1.00E-03
2.00E-03
3.00E-03
4.00E-03
5.00E-03
0.600
0.200
-0.200

0.200
0.600
1.000
1.400
1.800
(*10^3)

FLAC Version 5.0
18 - 29
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
1-Feb-05 11:52
step 142640
Flow Time
1.5168E+09
0.000
-0.400
HISTORY PLOT
Y-axis :
5 reldispx
(FISH)
13 reldispy
-0.800
(FISH)
X-axis :
12 Dynamic time
-1.200
-1.600
-2.000
10
15
20
25
30
35
40

Figure 18.35 Relative displacements at gridpoint (62,29) along upstream slope

JOB TITLE : .
FLAC (Version 5.00)

LEGEND
2-Feb-05 7:27
step 812615
Flow Time
1.5168E+09
0.000
-0.400
HISTORY PLOT
Y-axis :
5 reldispx
(FISH)
13 reldispy
-0.800
(FISH)
X-axis :
12 Dynamic time
-1.200
-1.600
-2.000
10
15
20
25
30
35


FLAC Version 5.0
18 - 30
JOB TITLE : .
FLAC (Version 5.00)

(10
04
LEGEND
1-Feb-05 11:52
step 142640
Flow Time
1.5168E+09
0.800
0.600
HISTORY PLOT
Y-axis :
0.400
9 Effective SYY ( 49, 23)

X-axis :
12 Dynamic time
0.200
0.000
-0.200
10
15
20
25
30
35
40

Figure 18.37 Pore-pressure and effective vertical stress near upstream slope
JOB TITLE : .
FLAC (Version 5.00)

(10
04
LEGEND
2-Feb-05 7:27
step 812615
Flow Time
1.5168E+09
0.800
0.600
HISTORY PLOT
Y-axis :
0.400

X-axis :
12 Dynamic time
0.200
0.000
-0.200
10
15
20
25
30
35

FLAC Version 5.0
18 - 31
18.2.4 Stage 4: Seismic Calculation Assuming Liquefaction

For Stage 4, the embankment soils are changed to liquefiable materials. The liquefaction model by
Byrne (1991) is assigned to embankment soils 1 and 2, with parameters set to correspond to SPT
measurements. For a normalized SPT blow count of 15, the Byrne model parameters are C1 = 0.2954
and C2 = 1.354. See Byrne (1991) for a discussion on the derivation of these parameters.
The liquefaction simulation starts at the saved state EDAM6.SAV. The embankment soils are
changed at this state by using the Material/Model tool. (Note that this tool is activated when the
Include Advanced Constitutive Models?
box is checked in the Model Options dialog.) The Region range is
Dynamic
selected and the
models box is checked in this tool. The Finn model is then assigned to
each region of the embankment soils. When the mouse is clicked within one of the embankment
soil regions, a dialog opens to prescribe the model properties. Figure 18.39 shows the dialog
with the properties selected for embankment soil 1. Note that the Byrne radio button is checked
in order to prescribe the appropriate parameters for the Byrne model. Also, the value for Latency
is set to 1,000,000 at this stage. This is done to prevent the liquefaction calculation from being
activated initially. The model is first checked to make sure that it is still at an equilibrium state
when switching materials to the Byrne model, before commencing the dynamic simulation. The
model state is saved at this stage as EDAM14.SAV.
Figure 18.39 Model finn properties dialog w/ properties for embankment soil 1
The model is now ready for the dynamic analysis. The water bulk modulus is assigned as
4.1 106 psf using the Settings/GW tool. The value for latency of the embankment soils is reduced to
50 in the Material/Property tool. The dynamic conditions are now set again in the same manner as for
Stage 3. The model state is EDAM15.SAV before damping and dynamic boundaries are applied;
it is EDAM16.SAV after they are applied.
The model is now run for a dynamic time of 40 seconds. The results in Figures 18.40 through 18.43
show the effect of pore pressure generation in the embankment soils. There is now a substantial
movement along the upstream face, as shown by figures Figures 18.40 through 18.42. The relative
vertical settlement at gridpoint (62,29) is now approximately 8 in. and the relative shift upstream
is approximately 1.8 ft., as shown in Figure 18.42.
FLAC Version 5.0
18 - 32
A significant increase in pore pressure (and decrease in effective stress) is calculated in the upstream
region, as indicated in Figure 18.43. The location of the pore pressure/effective stress measurement
is at gridpoint (49,23), which is at a depth of approximately 45 ft. below the upstream slope face,
and 135 ft. from the toe of the upstream slope. Note that the normalized excess pore pressure ratio,
(or cyclic pore pressure ratio) ue / c ,* at this location is approximately 0.80 at 40 seconds. (ue =
1.58 ksf and c = 1.98 ksf)
The final state is EDAM17.SAV.
18.3 Comments
This simple example assumes that the shear strength parameters of the liquefiable soils do not
change. It has been shown (e.g., Olson et al., 2000) that if the effective stress goes to zero, the shear
strength reduces to a strain-mobilized (liquefied) shear strength, which implies a residual cohesion.
There are several ways to incorporate a change of strength envelope in the FLAC model, such that
residual cohesion is developed as the material liquefies. For example, a FISH function can be used
to adjust the strength parameters as a function of change in the effective confining stress. A more
rigorous approach is to modify a bilinear strength model (such as the strain-softening bilinear model,
MODEL subi) to include the liquefaction behavior (e.g., the Byrne model). The existing MODEL nn
in FLAC incorporates the pore pressure generation effect into the Mohr-Coulomb model. This can
also be done with other models using either the FISH constitutive model facility (see Section 2.8
in the FISH volume) or C++ DLL model facility (see Section 4 in Optional Features) to create a
user-defined model.
* where ue is the excess pore pressure and c is the initial effective confining stress. Note that a
liquefaction state is reached when ue / c = 1.
FLAC Version 5.0
18 - 33
JOB TITLE : .
(*10^2)
FLAC (Version 5.00)

8.000
LEGEND
1-Feb-05 10:32
step 142728
Flow Time
1.5168E+09
2.205E+02 <x< 9.219E+02
1.851E+02 <y< 8.865E+02
7.000
6.000
-1.50E+00
-1.00E+00
-5.00E-01
0.00E+00
5.00E-01
1.00E+00
5.000

Boundary plot
0
4.000
2E 2
3.000

2.000
2.500
3.500
4.500
5.500
(*10^2)
6.500
7.500
8.500

Byrne (liquefaction) material and hysteretic damping
JOB TITLE : .
(*10^3)
FLAC (Version 5.00)
1.400
LEGEND
1-Feb-05 10:32
step 142728
Flow Time
1.5168E+09
-1.001E+02 <x< 1.900E+03
-4.598E+02 <y< 1.540E+03
1.000

0.00E+00
1.00E-02
2.00E-02
3.00E-02
4.00E-02
5.00E-02
6.00E-02
0.600
0.200
-0.200

0.200
0.600
1.000
1.400
1.800
(*10^3)

FLAC Version 5.0
18 - 34
JOB TITLE : .
FLAC (Version 5.00)

LEGEND
1-Feb-05 10:32
step 142728
Flow Time
1.5168E+09
0.000
-0.400
HISTORY PLOT
Y-axis :
5 reldispx
(FISH)
13 reldispy
-0.800
(FISH)
X-axis :
12 Dynamic time
-1.200
-1.600
-2.000
10
15
20
25
30
35


JOB TITLE : .
FLAC (Version 5.00)

(10
04
LEGEND
1-Feb-05 10:32
step 142728
Flow Time
1.5168E+09
0.800
0.600
HISTORY PLOT
Y-axis :
0.400

X-axis :
12 Dynamic time
0.200
0.000
-0.200
10
15
20
25
30
35

FLAC Version 5.0
18 - 35
18.4 Acknowledgment
This example is derived from data provided by Dr. Nason McCullough of CH2MHill. His assistance
and critical review of this document are gratefully acknowledged.
18.5 Reference
Byrne, P. A Cyclic Shear-Volume Coupling and Pore-Pressure Model for Sand, in Proceedings:
Second International Conference on Recent Advances in Geotechnical Earthquake Engineering
and Soil Dynamics (St. Louis, Missouri, March, 1991), Paper No. 1.24, 47-55.
Olson, S. M., T. D. Stark, W. H. Walton and G. Castro. 1907 Static Liquefaction Flow Failure of
the North Dike of Wachusett Dam, Journal of Geotechnical and Geoenvironmental Engineering,
126(12), 1184-1193 (2000).
FLAC Version 5.0
18 - 36
18.6 Data File EARTHDAM.DAT

;Title:Earth Dam
;... STATE: EDAM1 ....
config gw ats dyn extra 10
grid 180,28
gen (0.0,400.0) (0.0,680.0) (1800.0,680.0) (1800.0,400.0) ratio 1.0,1.0 &
i=1,181 j=1,29
model elastic
table 1 delete
table 1 0 550.0 350.0 550.0 720.0 680.0 800.0 680.0 1470 550.0 1800 550.0
set echo off
call tabtop.fis
set cd back
set tid=1
gentabletop
gen line 0.0,475.0 1800.0,475.0
gen line 350.0,550.0 1470.0,550.0
gen line 720.0,680.0 630.0,550.0
gen line 800.0,680.0 900.0,550.0
save edam1.sav
group foundation:soil 2 region 83 4
model mohr group foundation:soil 2
friction=40.0 dilation=0.0 tension=0.0 group foundation:soil 2
group foundation:soil 1 region 94 11
model mohr group foundation:soil 1
friction=40.0 dilation=0.0 tension=0.0 group foundation:soil 1
group embankment:soil 1 region 106 23
model mohr group embankment:soil 1
prop density=3.51 bulk=5.698E6 shear=2.630e6 cohesion=120.0 friction=35.0 &
dilation=0.0 tension=0.0 group embankment:soil 1
save edam2.sav
FLAC Version 5.0
18 - 37

model null group embankment:soil 1
model null group embankment:soil 2
fix x y j 1
fix x i 181
fix x i 1
set gravity=32.2
set flow=off
water density=1.94
set dyn=off
set echo off
call Ininv.fis
set wth=550 k0x=0.5 k0z=0.5
ininv
solve elastic
save edam3.sav
solve
save edam4.sav
apply pp 0.0 var 0.0 16866.36 from 70,29 to 1,1
fix pp i 82 149 j 29
fix pp i 150 181 j 29
prop por=0.3 perm=5.25E-9 region 74 3
history 4 gwtime
set mechanical=off
set flow=on
set step=10000000
water bulk 4.1e6
FLAC Version 5.0
18 - 38
set funsat=on
set fastwb=on
solve
save edam5.sav
apply pressure 0.0 var 0.0 8120.8 from 70,29 to 1,29
set mechanical=on
set flow=off
water bulk=0.0
set step 10000000
solve
save edam6.sav
;*** BRANCH: MOHR COULOMB ****
water bulk=4100000.0
set dyn=on
set =large
hist 100 read acc.his
call filter.fis
set filter in 100
set filter out 101
set fc = 5.0
filter
call int.fis
set int in 101
set int out 102
integrate
set echo off
call baseline.fis
set itab unc=102 itab corr=120 drift=0.3 ttime=40.0
set npnts=2048 itab cvel=104
baseline
history 12 dytime
FLAC Version 5.0
18 - 39

history 104 yvel i=49, j=24
history 107 xdisp i=77 j=29
history 108 ydisp i=77 j=29
history 110 esxy i=63, j=26
save edam7.sav
;*** BRANCH: NO DAMP ****
apply ffield
apply sxy -7631.6 hist table 104 from 1,1 to 181,1
apply xquiet yquiet from 1,1 to 181,1
save edam8.sav
solve dytime 20.0
set echo off
call fft inp.fis
fft inp
set echo off
call Fft.fis
fftransform
save edam9.sav
;*** BRANCH: HYST DAMP ****
restore edam7.sav
;... STATE: EDAM10 ....
initial hyst default -3.156 1.904
set dy damping rayleigh=0.0050 1.25 stiffness
apply ffield
FLAC Version 5.0
18 - 40

history 118 hyst modfac i=61, j=26
set echo off
call reldispx.fis
reldispx
hist reldispx
hist reldispy
set echo off
call inipp.fis
inipp
set echo off
call excpp.fis
excpp
hist excpp
save edam10.sav
;... STATE: EDAM11 ....
solve dytime 10
save edam11.sav
;*** BRANCH: RAYLEIGH DAMP ****
restore edam7.sav
;... STATE: EDAM12 ....
apply ffield
set echo off
call reldispx.fis
reldispx
hist reldispx
hist reldispy
set echo off
call inipp.fis
inipp
set echo off
call excpp.fis
excpp
hist excpp
save edam12.sav
;... STATE: EDAM13 ....
solve dytime 40
FLAC Version 5.0
18 - 41
save edam13.sav
;*** BRANCH: FINN HYST DAMP ****
restore edam6.sav
;... STATE: EDAM14 ....
model finn region 59 24
ff latency=1000000 ff c1=0.2947 ff c2=1.357 ff switch=1 region 59 24
solve
save edam14.sav
;... STATE: EDAM15 ....
water bulk=4100000.0
set dyn=on
set =large
call filter.fis
set filter in 100
set filter out 101
set fc = 5.0
filter
call int.fis
set int in 101
set int out 102
integrate
set echo off
call baseline.fis
set itab unc=102 itab corr=120 drift=0.3 ttime=40.0
set npnts=2048 itab cvel=104
baseline
FLAC Version 5.0
18 - 42

history 12 dytime
history 107 xdisp i=77 j=29
history 108 ydisp i=77 j=29
history 110 esxy i=63, j=26
save edam15.sav
;... STATE: EDAM16 ....
initial hyst default -3.156 1.904
set dy damping rayleigh=0.0050 1.25 stiffness
apply ffield
history 118 hyst modfac i=61, j=26
set echo off
call reldispx.fis
reldispx
hist reldispx
hist reldispy
set echo off
call inipp.fis
inipp
set echo off
call excpp.fis
excpp
hist excpp
save edam16.sav
;... STATE: EDAM17 ....
solve dytime 40
FLAC Version 5.0
18 - 43
save edam17.sav
;plot name: pp
;plot name: ssi
plot hold ssi fill
;plot name: disp
plot hold xdisp fill inv bound
;plot name: pp esyy hist
plot hold history 6 9 vs 12
;plot name: rel. disp
plot hold history 5 13 line vs 12
;plot name: base disp hist
;plot name: excess pp
;plot name: corrected velocity
;plot name: vel hist
plot hold history 102 vs 12
FLAC Version 5.0
18 - 44
18.7 FISH Functions

RELDISPX.FIS
;Name:reldispx
;Diagram:
def reldispx
reldispx = xdisp(62,29) - xdisp(62,1)
reldispy = ydisp(62,29) - ydisp(62,1)
end
INIPP.FIS
;Name:inipp
;Diagram:
def inipp
ppini
= pp(49,23)
end
EXCPP.FIS
;Name:excpp
;Diagram:
def excpp
excpp = pp(49,23) - ppini
end
FLAC Version 5.0
18 - 45
18.8 Data File INPUT.DAT

;Title:Input wave correction
;... STATE: INP1 ....
config gwflow ats dynamic
save inp1.sav
;*** BRANCH: UNFILTERED ****
;... STATE: INP2 ....
set echo off
call fft inp.fis
fft inp
set echo off
call Fft.fis
fftransform
save inp2.sav
;*** BRANCH: FILTERED ****
restore inp1.sav
;... STATE: INP3 ....
set echo off
call Filter.fis
set filter in=100 filter out=101 Fc=5
filter
set echo off
call fft inp.fis
fft inp
set echo off
call Fft.fis
fftransform
save inp3.sav
;*** BRANCH: UNCORRECTED ****
;... STATE: INP4 ....
set echo off
call INT.FIS
FLAC Version 5.0
18 - 46
integrate
set echo off
call INT.FIS
integrate
save inp4.sav
;*** BRANCH: CORRECTED ****
restore inp3.sav
;... STATE: INP5 ....
set echo off
call INT.FIS
integrate
set echo off
call baseline.fis
set itab unc=102 itab corr=120 drift=0.3 ttime=40.0 npnts=2048
set itab cvel=104
baseline
set echo off
call INT.FIS
integrate
save inp5.sav
;plot name: input acc
;plot name: fft
;plot name: input disp
;plot name: input vel
FLAC Version 5.0
18 - 47
18.9 Data File BASELINE.FIS

BASELINE.FIS
;Name:baseline
;Diagram:
;Input:itab unc/int/102/uncorrected velocity table
;Input:itab corr/int/120/low frequency sine wave correction
;Input:drift/float/0.3/residual displ. at end of record
;Input:ttime/float/40.0/total time of record
;Input:npnts/int/2048/number of points in record
;Input:itab cvel/int/105/baseline corrected velocity
def baseline
loop ii (1,npnts)
tt = float(ii-1) * ttime / float(npnts)
vv = pi * tt / ttime
cor d = drift * pi / (2.0 * ttime)
ytable(itab corr,ii) = -(cor d*sin(vv))
xtable(itab corr,ii) = tt
ytable(itab cvel,ii) = ytable(itab corr,ii) + ytable(itab unc,ii)
xtable(itab cvel,ii) = xtable(itab unc,ii)
endloop
end
FLAC Version 5.0
18 - 48
FLAC Version 5.0

FLAC Slope Stability Example

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

FLAC Slope Stability Example

Hochgeladen von

Copyright:

Verfügbare Formate

Example Applications

FLAC Version 5.0

Summary of Example Applications

1 Slope Stability for a Cohesive

3 Cemented Backfill Pillar

and Frictional Soil

and the Effects of Backfill

7 Stresses around a Pressurized

Grouted Cable Anchors

17 Lined Tunnel Construction in

FLAC Version 5.0

Slope Stability for a Cohesive and Frictional Soil

1 Slope Stability for a Cohesive and Frictional Soil

* See Section 10 for a comparison of FLAC results to a limit equilibrium analysis.

FLAC Version 5.0

FLAC Version 5.0

Slope Stability for a Cohesive and Frictional Soil

JOB TITLE : SLOPE UNDER GRAVITATIONAL LOAD

FLAC (Version 5.00)

Itasca Consulting Group, Inc.

Grid plot of initial slope

FLAC Version 5.0

1.2.2 Slope Collapse: Dry Conditions

FLAC Version 5.0

Slope Stability for a Cohesive and Frictional Soil

FLAC (Version 5.00)

Itasca Consulting Group, Inc.

Plot of displacement vectors and x-velocity contours at step 915

FLAC Version 5.0

JOB TITLE : SLOPE UNDER GRAVITATIONAL LOAD

FLAC (Version 5.00)

Itasca Consulting Group, Inc.

Plot of displacement vectors and x-velocity contours at step 1715

Following the procedure used earlier, enter the following commands:

FLAC Version 5.0

Slope Stability for a Cohesive and Frictional Soil

1.2.3 Effective Stress Analysis with WATER table

FLAC (Version 5.00)

Contour interval= 5.00E+01

FLAC Version 5.0

JOB TITLE : SLOPE UNDER GRAVITATIONAL LOAD

FLAC (Version 5.00)

Pore pressure contours

Itasca Consulting Group, Inc.

1.2.4 Effective Stress Analysis with CONFIG gw

FLAC Version 5.0

Slope Stability for a Cohesive and Frictional Soil

JOB TITLE : SLOPE UNDER GRAVITATIONAL LOAD

FLAC (Version 5.00)

Itasca Consulting Group, Inc.

Steady-state flow through slope

FLAC Version 5.0

JOB TITLE : SLOPE UNDER GRAVITATIONAL LOAD

FLAC (Version 5.00)

Pore pressure contours

Itasca Consulting Group, Inc.

FLAC Version 5.0

Slope Stability for a Cohesive and Frictional Soil

1.3 Data File SLOPE.DAT

FLAC Version 5.0

;... STATE: SL3 ....

FLAC Version 5.0

Slope Stability for a Cohesive and Frictional Soil