See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/277936824

Modeling laterally loaded piles

Article  in  Pollack Periodica · August 2013

DOI: 10.1556/Pollack.8.2013.2.13

CITATIONS READS

3 205

1 author:

Janos Szep
Széchenyi István University, Gyor
3 PUBLICATIONS   4 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Janos Szep on 21 March 2021.

The user has requested enhancement of the downloaded file.

 POLLACK PERIODICA
An International Journal for Engineering and Information Sciences
DOI: 10.1556/Pollack.8.2013.2.13
Vol. 8, No. 2, pp. 117–129 (2013)
www.akademiai.com

MODELING LATERALLY LOADED PILES

János SZÉP

Department of Structures, Széchenyi István University, 1 Egyetem tér, Győr, H-9026, Hungary
e-mail: [email protected]

Received 9 January 2012; accepted 4 November 2013

Abstract: Single row pile groups are gaining popularity as foundation system for both
buildings and bridges. They are a reasonable and economical solutions for structures with low
vertical loads e.g. bridge abutments. The lateral resistance of a pile group is the result of contact
pressure between the individual pile shafts and the soil. When designing this kind of structure, the
estimation of displacements and bending moments is the main focus of the calculation. In most
cases single piles are modeled by assuming a Winkler material. For their calculations, designers
typically employ one of the commonly-used structural FEM codes (e.g. AxisVM, Fem-Design).
On the other hand, advanced, truly 3D FEM packages (like MIDAS GTS, Plaxis 3D) are
available, allowing for realistic modeling of the soil environment and soil-structure interaction.
As an intermediate solution, packages based on traditional approaches are also available (GEO5,
CGU). In this paper, results obtained using three different calculation methods and packages are
compared.

Keywords: FEM, Spring constant, Soil-structure interaction

1. Single row pile group

Piles are is the most extensively-used foundation method when it is necessary to
carry, both vertical and horizontal loads. Well-established methods exist for calculating
the vertical capacity of piles. Numerous methods are also available to calculate pile
deformations, bending moments and shear forces due to lateral loading. Unfortunately,
the results generated by different models show significant variation. The most common
model for piles subjected to lateral loads is the elastically supported (Winkler [1]) beam.
Pile dimensions, forces and moments, as well as reinforcement schedules are generally

HU ISSN 1788–1994 © 2013 Akadémiai Kiadó, Budapest

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
118 J. SZÉP

calculated with FEM packages. As a consequence, the linearly elastic material models
employed in these - so-called ‘structural’ - packages are mostly suitable for
superstructure calculations. As the soil behavior is neither elastic, nor linear, the validity
of results based on linear elastic models for foundation design is questionable. To model
such nonlinear 3D soil-structure interaction problems, full-3D geotechnical packages
are the most suitable tools [2].

1.1. Importance of modeling single row pile group behavior

In Hungary, the design concept of a single row of piles has grown as an economical
solution to foundations for precast girder bridges [3]. In most cases a single row of piles
have sufficient vertical load capacity. For horizontal loads, however, reliable bending
moment capacity is more difficult to attain. As a result of the hinged connection
between the beams and the pylons for this type of bridge, the foundation stiffness has
only a minor effect on the bending moments, shear and normal forces in the
superstructure. For other bridge types with rigid deck-abutment joints like frame bridges
and integral bridges, the substructure-superstructure and the soil structure interaction
have much more significant impact on the behavior. Lateral capacity becomes even
more critical for these designs.
It is more and more common to build high embankments on weak soils. In this case,
correct modeling of the behavior of the lateral support for the piles is essential, since not
only must they resist lateral loads applied on the structure but also embankment forces
transferred through the soil. The failure mode induced by the embankment is a very
different design problem and could be as critical as structural loads.
With the new, advanced geotechnical packages more realistic modeling of soil-
structure interaction becomes possible. For some critical problems, calculations show
more favorable mechanical behavior than it was assumed based on routine bridge design
calculations. Calculations show that the piles in the abutment have significantly lower
loads on them than suggested by the Winkler-style models [1], while the piles of the
intermediate supports suffer more significant horizontal displacements and are subject
to greater loads than previously assumed [4].

2. Comparative calculations
As a first step, a single laterally loaded pile (Fig. 1) was analyzed using three
different numerical methods. Results of bending moment distribution and displacements
were then compared. The three numerical methods are:
• AXIS 10VM, the fundamental structural design tool in Hungary;
• GEO4, an increasingly popular geotechnical design code;
• MIDAS GTS, a true 3D geotechnical FEM package that provides more realistic
modeling for soil-structure interaction.
In order to be able to evaluate the effect of the differences between the calculation
methods, a very simple model has been used, as shown in Fig. 1.

Pollack Periodica 8, 2013, 2

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
 MODELING LATERALLY LOADED PILES 119

F v=1800–3600 kN
p=(0–100) kPa

Fh =180÷360 kN

sand
φ=30° qs0 =25 kPa
H=12.0 m
c=(1÷60) kPa qsH=75 kPa
γ=20 kN/m3 qb=4000 kPa
Es0 =15–30 MPa Ch0 =18.75÷37.50 MN/m3
EsH =45–90 MPa ChH =56.25÷112.5 MN/m3

reinforced concrete pile

D=80 cm E=20,000 MPa ν=0,16

Fig. 1. Basic geometry and properties for comparative study

The basic geometry and properties reflect typical values seen in Hungary. Several
other considerations are given below:
• dimensions and the material parameters of the pile being modeled are the most
commonly used Continuous Flight Auger (CFA) pile type;
• the pile tops are free to rotate, which is obviously not always the case, but this
question is beyond the scope of this paper, therefore it is disregarded;
• two load levels were applied, one, a typical working load (Fv=1800 kN,
Fh=180 kN), the other, nearly ultimate load (Fv=3600 kN, Fh=360 kN);
• vertical loading was used only for the MIDAS analysis since vertical stresses
affect the behavior of soil elements. GEO4 and AXIS have no coupled effect, so
vertical loads have no additional influence;
• no surface load has been considered for GEO4 and MIDAS models, while in
AXIS a surface load equivalent to 5.0m embankment height has been applied.
This generated an ultimate passive resistance in AXIS approximately equal to
the passive resistances in GEO4 and MIDAS;
• the shear strength parameters (φ and c), and the shaft and tip resistances (qs and
qb) are compliant with each other;
• the compression moduli (Es) and the horizontal subgrade reaction moduli (Ch)
are compliant with the shear strength parameters. Ch was calculated with the
formula Ch=α×Es/D, with α=1.0 and α=2.0 to investigate the effect of varying
the subgrade reaction modulus as well;
• cohesion of 1 kPa has been used in MIDAS models for numerical stability and
60 kPa for some GEO4 tests in order to simulate the effect the overburden of an
5.0 m high embankment.

Pollack Periodica 8, 2013, 2

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
120 J. SZÉP

2.1. AXIS VM model

The AXIS model is basically a beam element with continuous support [5]. The
support model is shown in Fig. 2.

k(z) = Ch (z) × D = Es (z) / D × D = Es (z) =

z
=Es0 + z × (E sH – Es0 ) / H [MN/m/m]
k(z)
qx(z;ex) = k(z) × ex [kN/m]
∆H
qxmax(z) qxmax(z)= (Kp – K a) × (z × γ × D + p) [kN/m]

qx(z) = qx(z;ex ) if qx(z;ex) < qxmax(z)

qx(z)
ex(z)
qx(z)= qxmax(z) if qx(z;ex) > qxmax(z)

Fig. 2. Horizontal supports for the AXIS model

The pile is divided into N sections with the length of ∆H (in the presented example
N=12). The spring parameters were constant within each section and were determined
with the formula shown in the figure. The spring constant - basically equal to the
compression modulus - increases with the depth as shown in Fig. 1. The slip limit for
the springs, for a D×∆H surface, are calculated from the difference between the active
and the passive earth pressures. (Note that D might be increased due to stress spreading.
Reese and Wang [6] and Sabatini [7] give methods for adjusting the effective diameter.
This issue was not considered in this paper.)
Performing a linear elastic calculation with AXIS, the slip limit forces are ignored,
so the reaction forces are unlimited. With a linear elastic/plastic calculation model, the
slip forces are taken into account, setting a limit to the maximum spring reaction forces.
According to current practice, only the linear elastic calculations are carried out,
allowing for unlimited pressures to develop along the pile shaft. In this case, the
calculated pressures should be at least compared with the passive earth pressure, but this
very rarely happens. With the linear elastic/plastic calculation, this check is
automatically carried out.
Analyses were performed using parameter groups shown in the left side of Table I.
The effects of horizontal force magnitude and limiting earth pressure were analyzed.
Test A was considered the reference test while Tests B to F had one or two parameters
changed.
Fig. 3 and Fig. 4 provide a graphical interpretation of the results, while the right side
of Table I includes a brief summary of the most important results.
By analyzing the results, one can conclude the following:
• Linear elastic/plastic assumptions yield significantly higher displacements than
the linear elastic assumption. In case of large loads, the mobilized passive earth

Pollack Periodica 8, 2013, 2

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
 MODELING LATERALLY LOADED PILES 121

pressure in the near-surface zone is not sufficient for equilibrium. The loads
must be transferred downwards, while excessive pile cap displacement occurs.
In Test B, the ‘mobilized’ part of the pile is 1 m deeper than in Test A. This
difference is about 3 m in Tests D and C. In Tests A and C, the spring forces are
the highest at the top of the pile, while in Tests C and D, due to limits in passive
earth pressure, an opposite trend is valid (no figures of these tests are
presented);

Table I
Results of AXISVM calculations

AXIS calculations D=80 cm, L=12.0 m laterally loaded reinforced concrete pile
Input data Results
Horizontal Maximum
Spring Surface
Load displacement bending
Calculation Figure constant load*
Mode of of pile cap moment
calculation
k p FA emax Mmax
MN/m/m kPa kN mm kNm
A 3. elastic 15 – 45 0 180 6.6 191
B 4. elastic/plastic 15 – 45 0 180 13.5 338
C elastic 15 – 45 0 360 13.2 382
D elastic/plastic 15 – 45 0 360 52.8 965
E elastic/plastic 30 – 90 0 180 10.9 335
F elastic/plastic 15 – 45 100 360 13.2 382
* the surface load increases the ultimate horizontal capacity

• The doubled modulus of subgrade reaction in Test E compared to Test B had

little effect on displacements, as in both cases the plastic limit values control
behavior;
• The surface loads have beneficial effect on displacements by increasing the
yield forces of the nonlinear springs, allowing almost equal displacements for
twice the load in Test F as in Test B. This is because in Test F 13.2 mm pile cap
displacement mobilized significant resistances, since the plastic slip limits
(passive earth pressure) were higher than in case B;
• The bending moments in the pile are also heavily influenced by the applied
calculation method (linear elastic or linear elastic/plastic);
• The higher subgrade reaction modulus applied in Test E had marginal effect on
the pile bending moments compared to Test B. The reaction forces important for
maximum bending moments are not governed by the spring constant, but by the
plastic slip forces (passive earth pressure) of the combined support elements;

Pollack Periodica 8, 2013, 2

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
122 J. SZÉP

-180.00 -6.6 -180.00 -180.00 -180.00

-175
-107
-100
-93 132
-76 132
-81
-30
186
-53 -26
-54 186
191
1 186
5 186

39 155
40 155
112
112
44
43 71
71
0.3 38
0.3 38 38
11 38
28
27 15
10 15
10
17
17 3
3
7
7 -1
-1
-1
-1
Z Z Z Z
-1
-3 -0.1
X X X X

Reaction force Displacement Shear force Bending moment

Fig. 3. AXIS Test A - Linear (for 180 kN/pile load, value for spring constant and slip limit force)

-180.00 -13.5 -180.00 -180.00 -180.00

-179
-22 -179
-177
-155 169
-22 -159 169
-65
-97 295
-65 -89 295
-108
-108
-108
-1 338
5 338
55 304
59 304
77 235
77 235
74 158
73 158
59 91
0.6 57 91
0.6
20
37 43
19 43
20 35
19 14
16
16 18 14

6 2
5 2
-1
-1
Z Z Z Z
-1
-4 -0.1
X X X X

Reaction force Displacement Shear force Bending moment

Fig. 4. AXIS Test B - Nonlinear (for 180 kN/pile load, base value for spring constant and slip
limit force)

• However, as the surface load in Test F increased the plastic slip forces, the
bending moments reduced significantly. Therefore a twofold increase in load
from 180 to 360 kN resulted in almost the same bending moment. In case F,
large reaction forces were generated since the slip limit (passive earth pressure)
was not reached;
• The lower parts of the pile in most cases (except for Test C) remain inactive
with no bending moment present.
In general, the calculation method and the slip limit values have the most significant
effect on the results. Linear calculations underestimate the displacements and the
bending moments, adequate safety can only be obtained if spring constants that

Pollack Periodica 8, 2013, 2

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
 MODELING LATERALLY LOADED PILES 123

gradually increase with depth have been considered, starting from 0 at the level of the
pile. Higher slip limits near the surface will also generate lower moments in the pile.
Similarly, high surcharges (from the surface load) will generate high slip limit values,
causing low bending moments.

2.2. GEO4 model

GEO4 and its successor, GEO5 contain a pile dimensioning block, which similar to
the linear AXIS model, calculates the pile behavior for horizontal loads using a linear
Winkler beam model. It is possible to apply user-defined spring constants or accept
automatically-generated values. With the program it is not possible to carry out analysis
similar to the nonlinear AXIS calculation, as it cannot limit the reaction earth pressures.
Even more problematic is that the limit earth pressures are not even displayed; therefore
additional calculations would be necessary for the check of resulting earth pressures.
Additional issues raise further doubts about the validity of the results. Due to all these
concerns the sheet pile check module was used instead. The module basically applies
the same spring and plastic slip model (see Fig. 2) as AXIS in linear elastic/plastic
mode. A great advantage of GEO4 - as opposed to AXIS - is that the active and the
passive limit values are calculated automatically from input soil parameters significantly
simplifying the model building procedure. An additional advantage is that the output
provides information about the ratio of the reaction forces to the active and passive limit
values and allows comparison with the calculated earth pressure at-rest as well. In case
of similarly difficult problems such interpretation of the results are extremely useful,
especially because the most common argument against FEM is the lack of possibility to
compare results with previous earth-pressure approach based experience.
The sheet pile check module of GEO4 with a few simple modifications is easily
applicable for the analysis of laterally loaded single piles. The following modifications
should be applied:
• To have the same surface level on both sides of the wall, the excavation depth
should be 0.01 m;
• The wall should be defined as pile wall with 1.0m pile spacing and with the
actual pile diameter;
• The horizontal load acting on the piles should be applied as an anchor force and
the anchor spacing should be 1.0 m;
• The subgrade reaction modulus should be referenced to 1.0 m;
• However, the results are referred per meter of sheet pile wall. The pile section
properties should be modified to reflect this assumption;
• The effect of the surface load, if it exists, on the active and passive earth
pressures should be taken into account by a fictitious cohesion, calculated with
the following formula:

p K p − Ka
c= . (1)
2 K p + Ka

Pollack Periodica 8, 2013, 2

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
124 J. SZÉP

Using the above method the results for the members of the sheet pile wall behavior
can be considered as result for a single pile.
With GEO4 the same pile and soil environment was analyzed as with AXIS. Several
models with different boundary conditions have been tested. Their input data is included
in the left side of Table II. The effect of the horizontal force magnitude and the limiting
earth pressure (by changing cohesion) behavior were analyzed on the pile. Test G is
considered a reference while tests H and I had one parameter changed for each.

Table II
Results of GEO4 calculations

GEO4
D=80 cm, L=12.0 m laterally loaded reinforced concrete pile
calculation
Input data Results
Subgrade Horizontal Maximum
reaction Cohesion* Load displacement bending
Calculation modulus of pile cap moment
Ch c FA emax Mmax
3
MN/m kPa kN mm kNm
G 18.75–56.25 0 180 8.8 287
H 18.75–56.25 0 360 34.0 837
I 18.75–56.25 60 360 11.0 380
* The aim of the applied cohesion is to simulate the surface load by increasing the passive
earth pressure and by reducing the active earth pressure.

Fig. 5 and Fig. 6 provide a graphical interpretation of the results, while the right side
of Table II includes a brief summary of most important results.
From the results the following can be concluded:
• The observed trends are very much like those of the AXIS calculations. They
are not repeated here;
• For 180 kN load in Test G, the displacement and the bending moment values
seem to be realistic. Fig. 7 suggests that at the top 2.0 m the earth pressure
limitation was activated. The bottom part of the pile is basically inactive for the
horizontal load. The reaction forces increase downwards together with the earth
pressure limit values;
• For 360 kN horizontal load, extremely large bending moments and
displacements develop within the pile, and for equilibrium, about 3.0 m long
passive zone is required;
• The cohesion has been applied to simulate the effect of the surface load on the
limiting earth pressures, but in other cases can represent true cohesion, is helpful
with respect to the pile behavior;

Pollack Periodica 8, 2013, 2

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
 MODELING LATERALLY LOADED PILES 125

Fig. 5. GEO4 test G (180 kN/pile load, base Fig. 6. Earth pressures plotted relative to
value for subgrade reaction modulus, zero limit values for GEO4 Test G
cohesion)

horizontal horizontal reaction shear force bending moment

displacement
Fig. 7. Results of GEO4 Test G (180 kN/pile load, base value for subgrade reaction modulus, zero
cohesion)

• The displacements and moments are generally smaller than those calculated
with AXIS, for which no reason could be found. While the numerical models
are the same, displacements and moments calculated by GEO4 are about 60-

Pollack Periodica 8, 2013, 2

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
126 J. SZÉP

85% of those calculated by AXIS. In the last calculation set, the results are
basically identical.
The output data of this package is very easily interpreted, and provides a clear
insight into pile behavior. The automatically calculated limiting earth pressures are one
of the advantages of the module, which greatly simplifies the data input. Another great
benefit of GEO4 is the almost negligible calculation time, allowing extensive sensitivity
studies to be carried out for the different parameters within a reasonable time frame.

2.3. MIDAS GTS model

The MIDAS GTS model consists of a reinforced concrete pile embedded in
homogeneous, isotropic, linearly elastic - perfectly plastic continuum with Mohr-
Coulomb failure criteria [8]. In this model no stress dependency is allowed with respect
to the elastic properties, as it is shown in Fig. 1. For the whole box model a uniform
15 MPa compression modulus and ν=0.3 have been considered as it is shown in Fig. 8.
These properties describe a weaker soil than the one described with the subgrade
reaction modulus for AXIS and GEO4.
Between the pile and the soil, special contact elements with relatively complicated
parameters must be placed. For the contact elements of each section a shear resistance
calculated from the unit skin friction presented in Fig. 1 can be defined. Two additional
modulus-like parameters should be defined for the contact elements. One of the
parameters can be referred to as shear stiffness and the other as normal stiffness of the
contact element. The interface elements allow for sliding between the pile and soil.
Tensile limits on the interface allow for gapping to take place on the tension side of the
pile. In two more analyses, both horizontal and vertical loads were applied. As
anticipated some interaction of the vertical and the horizontal stresses occurred within
the soil elements. From the assumed stress interaction, vertical loads should have had
some influence on the horizontal displacements and bending moments.
The test conditions and the most important results are presented in Table III. The
most important graphical output of the tests is shown in Fig. 8 and Fig. 9.
From the results the followings can be concluded.
• The displacement patterns and their spatial distributions are generally
reasonable and coherent with our general kinematic preconceptions. Realistic
displacement levels have been observed. For Test J, where no vertical load was
applied, only a minor vertical displacement was observed. The vertical load in
case K had marginal effect on the horizontal displacements most probably due
to the limited accuracy of the calculations;
• The bending moments are significantly lower than they are in the case of the
other two models;
• The low quality of the bending moment, shear and normal force diagrams are
due to the coarse grid. An unusual feature is that the shear force diagram has a
constant part near the top, but the bending moment diagram is relatively flat
around the peak.

Pollack Periodica 8, 2013, 2

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
 MODELING LATERALLY LOADED PILES 127

Table III
Results of MIDAS GTS calculations

MIDAS
D=80 cm, L=12.0 m laterally loaded reinforced concrete pile
calculation
Input data Results
Horizontal Vertical maximum
Horizontal Vertical
displacement displacement bending
Calculation load load
of pile cap of pile cap moment
Fh FV ex;max ex,max Mmax
kN kN mm mm kNm
J 180 0 5.3 0.8 132
K 180 1800 4.6 16.7 108
L 360 3600 10.8 33.1 230

pile and soil soil

Fig. 8. MIDAS Test K - horizontal displacements for 180 kN horizontal load

3. Summary
The application of single row pile groups are gaining popularity in the Hungarian
civil engineering practice, applied mostly for bridge abutments. The single row pile
group is an economical foundation system by means of which - if applied together with
certain superstructure types - it is possible to reduce the forces and moments of the
superstructure. For the appropriate estimation of the bending moment distribution in the

Pollack Periodica 8, 2013, 2

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
128 J. SZÉP

piles the soil-structure interaction must be modeled reliably. This paper compares the
results obtained from testing three different calculation methods. An important finding
is that the results obtained with the most popular structural FEM code (AXIS) neatly
matches with those from the widely used geotechnical package GEO4. To achieve good
agreement between the results of the two models, the spring constant was determined
using careful geotechnical analysis. Additionally it was essential to apply a linearly
elastic - perfectly plastic spring characteristic with adequately selected slip limit forces
to model the soil behavior appropriately. With these methods it is possible to analyze
the behavior of the superstructure and foundation in one complex model.

bending moment normal force shear force

Fig. 9. MIDAS Test K - Bending moment, shear and normal forces for 180 kN horizontal load

Results of MIDAS calculations will be used to improve the accuracy of the simple
AXIS model by applying more representative spring constants and slip conditions.
For representative modeling of soil-structure interaction the most suitable package
was MIDAS. The results from the MIDAS model are encouraging; however, the
validity of the results is still to be proved. Influence of the interface behavior will be
studied in greater detail as well as the degree of mesh refinement. Research in-progress
has produced more numerically stable results and hopefully more accurate predictions.
In order to improve and validate computational methods, small-scale physical models
and measurement data from full-scale structures will be used.

Acknowledgement
The Author would like to thank to his colleagues, Edina Koch, dr. Róbert
Szepesházi and dr. Richard Ray for their valuable contribution towards writing this
paper.

Pollack Periodica 8, 2013, 2

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
 MODELING LATERALLY LOADED PILES 129

Reference
[1] Hetényi M. Beams on elastic foundation, Theory with applications in the fields of civil and
mechanical engineering, University of Michigan Press, Ann Arbor, Michigan, 1964.
[2] Bak E., Koch E., Ray R., Sharle P., Szepesházi R. Parametric study of combined pile-raft
foundation. Proc. of the XIVth Danube-European Conference on Geotechnical
Engineering, Bratislava, Slovakia, 2-4 June 2010, p. 13.
[3] Szép J., Murinkó G., Szepesházi R. Modelling bridge substructures, (in Hungarian)
Geotechnika 2009 conference, Ráckeve, Hungary, 27-29 October 2009, p. 22.
[4] Szep J. FEM - modeling of abutment, 10th Slovak Geotechnical Conference - Geotechnical
Problems of Engineering Constructions, Bratislava, Slovakia, 30-31 May 2011,
pp. 207-218.
[5] AxisVm, Finite Element Analysis & Design Program, users manual, InterCad Ltd,
Budapest 2010.
[6] Reese L. C., Wang, S. T. LPILE Plus 3.0 Technical manual of documentation of computer
program, Ensoft, Inc., Austin, Texas, 1997.
[7] Strom R. W., Ebeling R. M. State of the practice in the design of tall, stiff, and flexible
tieback retaining walls, US Army Corps of Engineers, Engineering Research and
Development Center, Technical Report ERDC/ITL TR-01-1, 2001, p. 225.
[8] MIDAS GTS users manual, MIDAS Information Technology Co., Ltd. Delft, 2009.

Pollack Periodica 8, 2013, 2

Brought to you by Széchenyi István University | Unauthenticated | Downloaded 03/21/21 03:39 PM UTC
View publication stats