Full-Scale Test Wall in Sand

6-1

6 Full-Scale Test Wall in Sand

6.1 Problem Statement
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
tiebacks are anchored over a 24 ft length by pressure grouting, and are pretensioned at a lockoff
load of 186.0 kips. 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;
(3) pretensioning the tiebacks to the specified lockoff load; and
(4) 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.

FLAC Version 8.0

6-2

Example Applications

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.

Table 6.1
Row
1
2
3
4
5
6
7

Reinforcement spacing
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

Area (ft )
Bond Stiffness (lbf /ft/ft)

8.5 10

-3

6.3 10

8.5 10

-3

6.3 10

8.5 10

-3

6.3 10

.0103

.0103

6.3 10

8.5 10

-3

6.3 10

Bond Strength (lbf /ft)

5000

5000

6000

9000

6000

Yield Strength (lbf )

73620

73620

73620

222750

222750

73620

The pressure dependency of the grout bond strength for the soil nails and tiebacks is neglected
for this analysis. See Section 9 for an example that includes the pressure dependency of the bond
strength.
The tiebacks are assumed to be installed and pretensioned to a lockoff load of 168.0 kips before
further excavation is performed.

FLAC Version 8.0

Full-Scale Test Wall in Sand

6-3

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.
The soil is assumed both 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

6.2 Modeling Procedure

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 project file for the analysis is contained in testwall.prj.
Note that the spacing property is used with the STRUCTURE 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.

FLAC Version 8.0

6-4

Example Applications

Initial Equilibrium Stage

The FLAC grid for the problem region (shown in Figure 6.1) is created using the Build / Generate /
Geometry Builder
tool. The boundary geometry is an imported file, testwall.geo. This geometry can
be obtained from a dxf or drawn in the sketch tool. The final grid is shown in Figure 6.2. The grid
contains 1564 zones; elements representing soil behind the wall are one foot square, and elements
farther from the wall are slightly larger.
JOB TITLE : FULL-SCALE TEST WALL IN SAND

(*10^2)
1.150

FLAC (Version 8.00)

LEGEND

1.050

1-Jun-15 14:25
step
0
-5.333E+01 <x< 1.333E+01
5.100E+01 <y< 1.170E+02
0.950

Grid plot
0

2E 1
0.850

0.750

0.650

0.550

Itasca Consulting Group, INC.

Minneapolis, MN, USA
-4.500

-3.500

-2.500

-1.500

-0.500

0.500

(*10^1)

Figure 6.2

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.
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.

FLAC Version 8.0

Full-Scale Test Wall in Sand

6-5

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 ten-segment cable
with a grouted unstressed length of 24 feet connected to a three-segment cable with a pretensioned
un-grouted length of 21 feet. A pretension force of 186.0 kips is applied to the un-grouted cable to
simulate the effect of the pressure grouting. (Note that the pretensioning 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 un-grouted 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 un-grouted cables.
The pretensioning is performed using the FISH function PRETENSION.FIS, which is described
in Section 1.10.5 in Structural Elements. The middle cable segment of the pretensioned section
is deleted, and a pretension axial load is applied to the two remaining segments and increased
gradually (in six step increments) up to the lockoff load. Once the lockoff load is reached, the
remaining pretensioned segments are deleted, and a one-segment cable with tension set to 186.0
kips is installed and connected to the grouted cable section and the wall. Figure 6.3 displays the
axial force distribution in the tieback at the lockoff load.
JOB TITLE : FULL-SCALE TEST WALL IN SAND

(*10^2)
1.150

FLAC (Version 8.00)

LEGEND

1.050

18-Jun-15 12:35
step 69030
-5.333E+01 <x< 1.333E+01
5.100E+01 <y< 1.170E+02
0.950

Cable Plot
Axial Force on
Structure
Max. Value
# 5 (Cable) -1.435E+05
# 6 (Cable) -1.860E+05
Boundary plot
0

0.850

2E 1
0.750

0.650

0.550

Itasca Consulting Group, INC.

Minneapolis, MN, USA
-4.500

-3.500

-2.500

-1.500

-0.500

0.500

(*10^1)

Figure 6.3

Axial forces in tiebacks after pretensioning

FLAC Version 8.0

6-6

Example Applications

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.
JOB TITLE : FULL-SCALE TEST WALL IN SAND

(*10^2)
1.150

FLAC (Version 8.00)

LEGEND

1.050

1-Jun-15 14:25
step 142021
-5.333E+01 <x< 1.333E+01
5.100E+01 <y< 1.170E+02
0.950

Grid plot
0

2E 1
0.850

0.750

0.650

0.550

Itasca Consulting Group, INC.

Minneapolis, MN, USA
-4.500

-3.500

-2.500

-1.500

-0.500

0.500

(*10^1)

Figure 6.4

FLAC Version 8.0

Plot of grid following removal of elements representing soil in

front of the wall

Full-Scale Test Wall in Sand

6-7

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 STRUCTURE 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 contained in WALL.FIS.
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 maximum axial force in cable ID 6 (184.6 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 available in
Lorig (1991).
JOB TITLE : FULL-SCALE TEST WALL IN SAND

(*10^2)
1.150

FLAC (Version 8.00)

LEGEND

1.050

4-Jun-15 15:33
step 142021
-5.333E+01 <x< 1.333E+01
5.100E+01 <y< 1.170E+02
0.950

Boundary plot
0

2E 1

Cable Plot

0.850

Axial Force on
Structure
Max. Value
# 2 (Cable) -6.277E+03
# 3 (Cable) -3.656E+03
# 4 (Cable) -1.133E+03
# 5 (Cable) -1.475E+05
# 6 (Cable) -1.846E+05
# 7 (Cable) -1.191E+04
# 8 (Cable) -1.395E+04
# 9 (Cable) -1.356E+04
#10 (Cable) -1.080E+04
Cable Plot
Shear Spring Bond Yields

0.750

0.650

0.550

Itasca Consulting Group, INC.

Minneapolis, MN, USA
-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

FLAC Version 8.0

6-8

Example Applications

JOB TITLE : FULL-SCALE TEST WALL IN SAND

FLAC (Version 8.00)

(10

02

LEGEND
1.800

1-Jun-15 14:25
step 142021

1.600
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

Itasca Consulting Group, INC.

Minneapolis, MN, USA

Figure 6.6

Distribution of axial tensile forces in the tieback distance from

the face (ft) vs actual axial force (kips)

JOB TITLE : FULL-SCALE TEST WALL IN SAND

FLAC (Version 8.00)

(10

01

LEGEND
1-Jun-15 14:25
step 142021

1.200

Table Plot
Stage I nail

1.000

Stage II nail
Stage III nail

0.800

Stage IV nail
Stage V nail
0.600

Stage VI nail
Stage VII nail

0.400

0.200

-10

-8

-6

-4

-2

Itasca Consulting Group, INC.

Minneapolis, MN, USA

Figure 6.7

FLAC Version 8.0

Distribution of axial tensile forces in the soil nails distance from

the face (ft) vs actual axial force (kips)

Full-Scale Test Wall in Sand

6-9

JOB TITLE : FULL-SCALE TEST WALL IN SAND

(*10^2)
1.150

FLAC (Version 8.00)

LEGEND

1.050

1-Jun-15 14:25
step 142021
-5.333E+01 <x< 1.333E+01
5.100E+01 <y< 1.170E+02
0.950

Displacement vectors
scaled to max = 5.000E-02
max vector = 2.263E-02
0

0.850

1E -1

Boundary plot
0

2E 1
0.750

0.650

0.550

Itasca Consulting Group, INC.

Minneapolis, MN, USA
-4.500

-3.500

-2.500

-1.500

-0.500

0.500

(*10^1)

Figure 6.8

Displacement vectors for test wall analysis

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 8.0