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

net/publication/280085579

Buckling Resistance of Geotextile-encased Columns

Conference Paper · September 2014

CITATIONS READS

0 147

3 authors, including:

Stefan Vogt
Technische Universität München
19 PUBLICATIONS   39 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Micro-Mechanics of Sand View project

All content following this page was uploaded by Stefan Vogt on 16 July 2015.

The user has requested enhancement of the downloaded file.

Buckling Resistance of Geotextile-encased Columns
S. Vogt & G. Bräu
Zentrum Geotechnik, Technische Universität München, Munich, Germany
T. Hampel
Bauer Spezialtiefbau GmbH, Schrobenhausen, Germany

ABSTRACT: Buckling of Geotextile-Encased Columns (GEC) is usually resisted by the surrounding soil.
Therefore, in sufficiently stiff soils there is no risk of buckling GECs having conventional geometry (L/D
ratio). For many conditions GECs are therefore a suitable and cost effective system to transfer loads from
a deformation tolerant structure into firm subsoil through layers several meters thick consisting of soft
soils. Cohesive clayey or organic soils having a consistency near the liquid limit may provide undrained
shear strengths well below 10 kN/m2 and may be found throughout large deposits at maritime delta re-
gions around the world. For these ultra-soft soils the internal resistance of buckling may limit the bearing
capacity of a GEC system. Despite the soil stiffness the main parameter influencing the buckling re-
sistance is the bending stiffness (flexural rigidity) of a GEC. The paper gives results from an experimental
work to determine the bending stiffness of GECs. It is shown how to use the worked out stiffness parame-
ters for calculating the buckling load of a single GEC. The method can cover imperfections of the GEC
axis from installation processes and subsoil conditions and can describe lateral soil support for undrained
and drained loading conditions.

Keywords: Geotextile-Encased Columns (GEC), Soft Soils, Buckling, Stability, Bending Stiffness

1 INTRODUCTION

A Geotextile-Encased Column (GEC) is generally a foundation element that is carrying loads as compres-
sion forces into a deep bearing soil layer. GECs are used as a very efficient method for foundations of
constructions such storage areas or dams that are tolerant to settlements (Kempfert und Wallis 1997, Bräu
1999, Raithel 1999, Alexiew et al. 2012 and Sobolewski et al. 2012). Under subsoil conditions where a
soft layer may show a thickness of only several meters as given in most parts of central Europe there is no
danger of buckling in general (Kempfert et al. 2002, Küster et al. 2011, Raithel et al. 2011 and 2012).
However, at estuaries formed by rives carrying large sediment freights consisting mainly of fine grained
or organic soils large soft soil deposits exist consolidating under their self-weight. If the rate of sedimen-
tation is high and therefore the increase of the total vertical stress z and in addition the permeability k of
the soil is low or the drainage length is long, soft soils may show layers where underconsolidated condi-
tions are governing the soil behavior. The excess pore pressure in such deposits lead to extremely low ef-
fective stresses 'z and hence undrained shear strengths cu well below 10 kN/m2. The water content may
show values higher than the water content at liquid limit. A project where GEC foundation system is used
successfully within a very soft and thick soil layer is discussed by Alexiew & Moormann 2010.
Different failures may occur and must be considered:
 Failure of the bearing soil in deeper layers around the GEC due to exceeding the maximum possible
skin friction and tip resistance; to be proofed according to EC 1997-1.
 Yielding of the GEC cross section because of normal forces that lead to a tensile failure in within the
geotextile; shown in detail by Raithel 1999.
 Loss of stability from exceeding the buckling load; as discussed in this paper.
The failure of the subsoil is to be excluded by proofing as per EN 1997-1. The following considera-
tions assume that proofing using the standard EN 1997-1 is satisfied. Since then checking the yielding of
the cross section and the loss of stability is remaining. With an increase of the normal force within the
GEC whether maximum stress that depend on the strength of the GEC cross section exceed or the GEC
will buckle with increasing lateral deformations that is controlled by the lateral soil resistance.
For calculating the buckling load of beam-like foundation elements supported by soft soils analytical
models exist. According to the construction method certain imperfections have to be considered. The soil
response is described by an elastic-plastic approach. Unlike piles or pile-like structures made up of steel
or concrete, that behave roughly elastic within a certain range of strains, the stiffness of the granular soil
within a GEC and the resulting bending stiffness of the GEC cross section where the geotextile strongly
interacts with the granular soil within the geotextile and the supporting soil is hard to estimate. Prelimi-
nary considerations show, that for a proper calculation of the flexural stiffness of a GEC many decisive
boundary conditions must be taken into account. Unless further experiences are available the assumption
of the bending stiffness for proofing the buckling resistance of GECs without experimental evidence may
lead to uncertainties.
Therefore, a conducted laboratory test series aims for a better understanding about the most important
parameters that govern the bending stiffness of GECs. The loading tests were back-calculated using both a
simple idealization of the cross section and the finite-element method using advanced soil models for
simulating the interaction between the fill material and the geotextile encasement tensile stresses while
acting as an membrane. The obtained values were used in an analysis on the buckling resistance of GECs
that cover extreme conditions made up of a large layer of soil with undrained shear strengths as low as
cu = 2 kN/m2. The discussion of the finite-element model and the parametric study for theoretical buckling
loads of GECs under different boundary conditions are not part of this paper.

2 BUCKLING LOAD OF AN EMBEDDED GEC

2.1 Analytical Solution

Since the effective buckling length Lcr can in most cases develop free of the boundary conditions at the
top or bottom end of the soft soil layer, an infinitely long GEC can be assumed for most practical purpos-
es. For the calculation an equivalent structural system with a simply supported beam, as shown in Fig-
ure 1, is used. The length of the beam is Lcr and the waviness of the sinusoidal deformation is fixed at
n = 1. Lcr which is in general much smaller than the thickness of the soft soil layer is not an input parame-
ter of the calculation but needs to be determined by an iterative procedure as illustrated below.

N
T=P

Lateral soil
z
support
Imperfect
GEC axis
P

M
Lcr
p(z,y) T = 0

e0
Figure 1. Static System
The buckling load of the system consisting of the GEC and the supporting soil can be computed according
to the analytical solutions provided by Wennerstrand, J. & Fredriksson, A. 1988 or Vogt, N., Vogt, S. &
Kellner, C. 2005 and 2009 respectively. The equation (1) covers an imperfection e0 in the middle of the
buckling figure Lcr / 2 and a lateral elasto-plastic soil support of the GEC. For most cases Lcr is unknown
and the decisive buckling load Ncr must be found in an iterative procedure.
2
  
2
L 
y f  EI     p f  D   cr 
 Lcr    
N cr  (1)
y f  e0
The bending stiffness EI of the GEC having a diameter D must be determined in general by taking into
account the stress level and the interaction between the fill material and the geotextile acting as an encas-
ing membrane. The influencing parameters are discussed in detail in chapter 2.4. The maximum soil sup-
port pf mobilized at a lateral displacement of yf that is acting on the GEC diameter D must be available for
the expected time the GEC foundation will be used. For simple cases pf and yf may be chosen according
to the information given in chapter 2.2. Lateral thrust on the GEC foundation because of a close surface
loading or situation where the GECs are placed within inclined slowly creeping slopes, soil liquefaction
under dynamic loadings or other influencing loads where one may expect no lateral soil support at all lead
to very small buckling loads. These cases must be taken with special care and Lcr must be considered ac-
cording to the “fixities” governed by bearing soil layers above and below the soft soil layer.
According to installation processes and the geology of the subsoil imperfections regarding the axis of
the GEC must be considered. The parameter e0 reduces the buckling load Ncr and should be taken with
care. For micro-piles e.g. it is common that a radius of curvature 1/κ is given. By using the parameter 1/κ
for describing imperfection of a GEC the parameter e0 must be calculated according to equation (2).
Lcr 1  L 
e0  tan arcsin  cr   (2)
2 2  2 1   
The parameter e0 influences both the decisive buckling load Ncr and the length Lcr. This demands an itera-
tive procedure.

2.2 Lateral Soil Support

The soft soil surrounding the GEC may only provide limited lateral support in case the mean stress p ex-
ceeds the maximum soil shearing resistance. Figure 2 gives a curve close to reality of the soil support p
against the lateral deformation y (dashed line). If the GEC is fully embedded by the soil failure stress p f
will mark the point where soil starts to flow fully around the cross section of the GEC. Beside this failure
mechanism pf may be limited because of a wedge failure at shallow depths. If the section of the GEC ex-
amined is close to the surface failure will happen in form of fracturing upwards to the surface where there
are no effective stresses at the boundary surface.

Lateral soil support p

pf
Characteristic support of the
soft soil close to reality

p
k pf Idealized characteristic
k y of the soft soil support
1

yf Deformation y
Figure 2. Elasto-plastic soil support

Different models for determining the buckling load Ncr of piles use an elasto-plastic approach to cover
the soil support p for varying lateral deformation y (Wennerstrand & Fredriksson 1988, Ofner & Wimmer
2007, Vogt et. al 2005 and 2009). Using this model the soil support is described elastic up to a critical lat-
eral deformation yf. This branch of the p-y curve can be also described by a bedding modulus k = pf / yf.
For deformations y > yf the soil behavior is fully plastic and pf is limiting the lateral support of the GEC.
Support stresses p > pf are not allowed.
Vogt & Vogt 2011 provide brief information about a model scaled lateral loading tests on piles and a
series of triaxial compression and extension tests. An average value of yf / D ratios under undrained and
drained loading conditions can be described by the two equations (3) covering the experimental results by
trend lines. The empirical equations in (3) for calculating the dimensionless ratio yf / D [-] are valid for in-
serting the undrained shear strength cu of the soil in units [kN/m2].
yf 0,2 yf 0,4
 0, 4
for undrained loading and  0, 4
for drained loading (3)
D cu D cu
For undrained loading conditions the ultimate lateral soil support pf can be described by using the plas-
ticity theory (Randolph, M. F. & Houlsby 1984). In the case that the soil fails according to a full flow
around the GEC diameter pf may be calculated by the first expression in formula (4), where 'z is the ef-
fective vertical stress. At shallow depths the soil resistance from the first expression in (4) is larger than
from the second equation in formula (4) which describes a failure mechanism upwards to the surface
(wedge failure).
 
 2 z 

p f  min  4 2  2  cu ;  ' z  2     cu  (4)
  3 D 
Moreover, there can be found smaller values of pf in the literature that aim for covering effects raising
form the sensitivity (structural strength) and viscosity of normally consolidated soft soils.
If the rate of loading is slow or the permeability of the soil is high the soil resistance pf can be calculat-
ed under assumption of drained conditions. Especially in the case of a GEC foundation the presumption of
drained conditions in the soil surrounding the GEC may lead to a realistic estimation of the bearing be-
havior. The well permeable geotextile and the coarse grained fill lead to quick pore pressure dissipation in
the soft soil layer. Water outflow from the soft soil under the static lateral loading of the GEC prohibiting
flexural buckling decreases the void ratio and therefor increases the shear strength and the maximum lat-
eral support pf. A simplifying approach to calculate pf under drained loading conditions in respect to the
drained effective shear parameters ' and c' is given by Hansen 1961 and is described by the following
equation (5).
p f  K q   ' z  K c  c' (5)

20 100
′k = 35
′k = 35
10 ′k = 30 50
′k = 30
Parameter Kc [-]
Parameter Kq [-]

′k = 25
5 ′k = 25
′k = 20
′k = 20

′k = 15 ′k = 15
2 10

1 5
0 5 10 15 20 0 5 10 15 20
Relative depth z / D [-] Relative depth z / D [-]
Figure 3. Coefficient Kq und Kc for calculating the ultimate drained soil resistance according to Hansen 1961
 The parameters Kq and Kc can be determined from the curves given in the diagrams in Figure 3. The
effective vertical stress 'z can be taken from the mean value given at the depth according to Lcr / 2.
If the net distance between the GEC is close areas within the soft soil needed for supporting the lateral-
ly bending GEC overlap. This means that the ultimate stress pf calculated by (4) or (5) overestimates the
decisive soil resistance. A rough estimation for the net distance may be given by 3·D. In such cases the
soil resistances according to the passive earth pressure coefficients valid for plain conditions have to be
considered (see DIN 4085 e.g.).

2.3 States of equilibrium (second order theory)

Solving relationship (1) for yf = y, the axial load N on the pile required to keep the deformed member
(beam, pile or GEC) in equilibrium is obtained as a function of the displacement y. Such states of equilib-
rium are shown in Figure 4 for various imperfections in the structural member (i.e. beam, column or pile).
The curves are obtained by assuming a constant flexural stiffness EI and an elastic or elastic-plastic soil
support. For comparison the solutions of the unsupported beam is also given by Figure 4.

Normal force N

Perfect plastically supported beam

e0[A] = 0 Perfect elastic-plastic supported beam

Buckling load according to the
elastically supported beam
e0[B] > e0[A]
Imperfect elastically supported beam
e0[C] > e0[B] Buckling load according to
Imperfect elastic-plastic supported beam the elastic-plastic supported
imperfect beam
e0[D] > e0[C]

Buckling load according to the

Perfect unsupported beam unsupported beam

Imperfect unsupported beam

yf Deformation y
Figure 4. States of equilibrium (second order theory)

In the case of a perfectly straight, elastically supported beam or GEC respectively subjected to an axial
load N there are no lateral displacements y and bending moments M until the bifurcation load Ncr is
reached. Should the beam be displaced laterally, it will subsequently move back to its initial position. On
reaching the buckling load, any lateral displacement leads to neutral equilibrium. It is thereby necessary
that increasing lateral displacements leads to additional supporting forces from the lateral bedding. How-
ever, as soon as the lateral soil reaction p does not increase linearly under the cumulating deformations y,
unstable conditions are reached. Thereby the deformation yf until the maximum soil support pf is mobi-
lized is decisive, particularly when a bi-linear function p(y) is assumed as a first approximation (see chap-
ter 2.2). The same applies to pre-deformed supported GECs where e0 > 0. As soon as the lateral deflection
of the GEC reaches the magnitude at which the ultimate soil reaction pf is attained, an unstable situation
arises. The bifurcation load is calculated for y = yf by means of the above equation (1). In parallel to this
result it is shown in a study on the buckling behavior of slender piles and drilling rods based on finite-
element calculations carried out by Meier 2009 that according to the parameter yf, that describes in partic-
ular the stiffness of the supporting soil, the influence of both the magnitude of the imperfection e0 and the
initial stiffness of the supporting soil on the buckling load is large.
 2.4 Bending Stiffness
The bending stiffness of the GEC has a very important influence on the buckling load calculated by
equation (1). The different effects arising from the interaction between the fill material (i.e. sand or grav-
el) and the geotextile on the input parameter EI for a GEC foundation are rather complex compared to the
calculation of the bending stiffness of concrete piles or even steel piles. Figure 5 discusses the develop-
ment of stresses within the fill material and the geotextile for different deformations y.

Compression
Normal force N Compressive stresses within the fill material
A
Tensile stresses in the geotextile

Ncr N Tension

C Compression
Compressive stresses within the fill material
y B
Tensile stresses in the geotextile
B D
Tension

Compression Partly compressive stresses within the fill

N material
C
Tensile stresses in the geotextile, no
compressive stresses possible
Tension
A
Compression
Rupture within the geotextile?
D Shear strength of the fill material reached
Tension
yf Deformation y
Figure 5. Development of stresses within the GEC cross section

The normal force N acting on a GEC will cause bending moments from the very beginning of a loading
because of the initial imperfection e0. In the beginning (point A) stresses arising from the moment are ra-
ther small compared to the stresses resulting from the normal force N. Furthermore compressive stresses
within the fill material and tensile stresses in the geotextile are dependent form the installation process,
which causes pre-stressing within the GEC cross section. By rising N also the moment M increases (point
B). This leads to a rearrangement of stresses within the fill material and the geotextile. The maximum ten-
sile stress in the geotextile increases. At this point there are two competing effects determining the EI of
the GEC:
 The stiffness of the fill material increases non-linear. By assuming oedometric conditions within the
fill material an exponential law that describes the stress dependency on the stiffness of the fill material
can be used (see e.g. Ohde 1939). At higher stress levels or normal force N respectively the fill materi-
al pushes towards the geotextile membrane and causes radial strain within the geotextile and hence lat-
eral strains within the fill material. In this state the oedometric stiffness must be reduced because of in-
creasing shear strains within the fill material.
 Opposite to the stress-strain behavior of the fill material the geotextile tends to loose stiffness as tensile
stresses increase from the very beginning of the loading.
Because of the increasing bending moment from point B to point C the geotextile itself would need to
carry compressible stresses from a certain deformation y on. For further deformation to point D the in-
creasing tensile stresses within the geotextile may cause failure mostly because of the limited seam
strength of the tube-like geotextile. It is very important to note that buckling may happen within a defor-
mation y where the GEC reacts more or less like an elastic beam. This means that buckling failure is not
necessarily linked to the rupture of the geotextile as given by the information in Figure 5. This was also
proofed by the loading tests illustrated in chapter 3.3.
To get more detailed information about the complex mechanism between the fill material and the geo-
textile encasement that determines the bending stiffness EI the main purpose of the research work was to
examine two different experimental setups that give a basis for interpreting the bending stiffness. Fur-
thermore, the bearing behavior was modelled using 3-d finite-element calculations on the basis of the per-
formed laboratory tests. These results are not part of this publication and will be discussed in the future.

3 EXPERIMENTAL WORK

3.1 General
By using two different loading mechanisms the bending stiffness of different GECs was examined in
laboratory tests. First bending tests were performed to give a rough insight to the interaction between the
fill material and the geotextile encasement. In these bending tests the GEC was placed horizontally on 3
supports. The maximum deformation was measured and a bending stiffness was calculated on the basis of
the beam equation. To examine the very decisive effect of axial stresses within the GEC fill material
buckling tests on small scaled GECs were performed. Measurement data from these tests where used to
back calculate the bending stiffness using Euler’s equation.
Three different GECs where examined. The main properties are given by Table 1. Type 1 and Type 2a
are model scaled GECs. For this 2 GECs both bending tests and buckling tests were conducted. Type 2b is
a commercially available GEC. Note that beside the diameter D, the seam strength Smax,r and tensile
strength at Smax,r Type 2a does not differ from GEC Type 2b.

Table 1. Geotextiles used in the experimental tests

Woven geotextile Woven geotextile Woven geotextile
Characteristic
Type 1 Type 2a Type 2b
Material Polypropylen Polyethylenterephthalat Polyethylenterephthalat
Diameter D 0,1 m 0,1 m 0,4 m
Mass per area ρA 100 g/m2 575 g/m2 575 g/m2
Tensile strength, axial Tmax,a 24,9 kN/m 71,3 kN/m 71,3 kN/m
Tensile strength, radial Tmax,r 15,7 kN/m 159,1 kN/m 159,1 kN/m
Seam strength, radial Smax,r 4,9 kN/m 4,6 kN/m 32,4 kN/m
Tensile strain at Tmax,a 22,3 % 10,3 % 10,3 %
Tensile strain at Tmax,r 14,5 % 7,8 % 7,8 %
Tensile strain at Smax,r 25,1 % 14,5 % 25,7 %

Two fill materials where used in the test series to get information about the influence of the stiffness
and shear strength of the fill material on the bearing behavior of the GEC. The main geotechnical proper-
ties of the used well-graded sand and uniformly-graded gravel are given in Table 2.

Table 2. Fill materials used in the experimental tests

Sand Gravel
Characteristic Remarks
well-graded uniformly-graded
Particle shape - rounded rounded
Loosest density d,min according to DIN 18126 1,551 g/cm3 1,627 g/cm3
Highest density d,min according to DIN 18126 2,106 g/cm3 1,827 g/cm3
Angle of friction ' slope response, dry soil 35,0 ° 32,5 °
Oedometer stiffness Es at 'z = 100 kN/m2 15 MN/m2 27 MN/m2
Ohde exponent m as introduced by Ohde 1939 0,61 0,43

3.2 Bending tests

The test procedure of the conducted bending tests is illustrated by the pictures in Figure 6. The GEC
(here: Type 2b) was first mounted to a filling frame above a deep shaft (see picture 1 and 2). The column
was supported right at the bottom to prevent the loosening of the fill material due to the expansion of the
geotextile when moving the GEC out of the shaft. According to the desired target value of the initial rela-
tive density ID the fill material was placed using different compaction efforts. Before the bending test
started the horizontally aligned GEC was supported uniformly along its length (see picture 3). The sup-
ports at the edges of the GEC where made up of stable concrete blocks. For the middle support a crane
was used. To solve the beam equation of the statically indeterminate system in order to calculate the bend-
ing stiffness EI the measurement of the force carried by the middle support is necessary. This was
achieved by using a load cell above the hook. The measurement started after the uniform support was re-
moved (see picture 4). The bending deformation and the reaction force in the crane were monitored con-
tinuously during the test while middle support was lowered by the crane. Since the loading is only due to
the self-weight of the GEC the total weight was determined after the bending test.

1 3

2 4

Figure 6. Bending tests on GEC

Results of 8 bending tests on 4 m long GEC type 2b are given in the charts of Figure 7. The data points
show the influence of the fill material and the initial relative density ID on the back calculated bending
stiffness EI. First the two diagrams give the measured deformation y perpendicular to the GEC axis. It
can be interpreted that both the fill material itself and the density of the fill material strongly influences
the stiffness EI. For low values of ID the bending stiffness remains small. A higher stiffness can be
achieved for the well graded stand and a high density ID.

Distance from midpoint of the GEC axis z [m]

2,50
-2,0 -1,5 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 2,5 Fill material: Sand,
0 well-graded
Deformation y [cm]

2,25
ID high
5
Bending stiffness EI [kNm2]

2,00
10
1,75
15 Fill material: Gravel, ID low
uniformly-graded 1,50
20
Fill material: Gravel,
1,25 uniformly-graded
0
Deformation y [cm]

ID high 1,00
5

10 0,75

15 Fill material: Sand, ID low 0,50

well-graded 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00
20
Initial relative density ID [-]
Figure 7. Results from large scale bending tests on GECs made from woven geotextile type 2b
 Since the bending tests do not cover the effect of compressive stresses from the normal force N within
fill material additionally buckling tests were performed. On the basis of this test results the influence of
the stiffening compressive stresses can be evaluated.

3.3 Buckling tests

Because of the massive complexity for loading the 4 m long GEC Type 2b within buckling tests only
tests on the model scaled GECs Type 1 and Type 2a providing a diameter of D = 0,1 m filled with the
well-graded sand were conducted. The main aim for the buckling tests was to cover the effect of the com-
pressive stresses within the fill material and evaluate their influence on the bending stiffness EI. The load-
ing was performed using an electro-mechanical jack providing a large traverse path. A constant rate of ax-
ial deformation was applied to the GEC. Pictures of the bucking tests on the 80 cm long model scaled
GECs are given on the left side of Figure 8.
During the loading the axial force N and the lateral deformation y was monitored. The results of the 4
buckling test are plotted in the diagrams of Figure 8. It can be seen first that the buckling load of the GEC
Type 1 was significantly smaller that for the Type 2a. This can be explained by the lower geotextile stiff-
ness of Type 1 GEC. The influence of the relative density ID of the fill material can be also interpreted by
the graphs in Figure 8. Tests with a higher density of the fill material provided higher initial stiffness re-
garding the deformation y and therefore the settlement of the GEC. Also the buckling load Ncr increased
for the tests with higher values for ID. It is noteworthy that during all loading tests at the maximum normal
force N = Ncr there was no failure of the seam notable. Only at larger deformations y seam rupture hap-
pened. At this deformations y the test was stopped as given by the curves in the diagrams of Figure 8.

Geotextile: Type 1

2,0
Normal force N [kN]

1,8 Test 1: Buckling loads:

1,5 ID ≈ 0,90
Ncr,1 = 1,6 kN
1,3
Ncr,2 = 1,2 kN
1,0
0,8 Test 2:
0,5 ID ≈ 0,75
0,3
0,0
0 1 2 3 4 5 6 7 8 9 10 11 12
Maximum deformation y [cm]
Geotextile: Type 2a
8,0
Test 3: Buckling loads:
Normal force N [kN]

7,0
ID ≈ 0,90
6,0 Nki,3 = 6,3 kN
5,0 Test 4:
ID ≈ 0,75 Nki,4 = 5,6 kN
4,0
3,0
2,0
1,0
0,0
0 1 2 3 4 5 6 7 8 9 10 11 12
Maximum deformation y [cm]
Figure 8. Buckling tests on model scaled GECs made from woven geotextile type 1 und 2a

The buckling loads and the back calculated EI using Euler’s formula are summed up by Table 3. To de-
termine EI also the measurement of the buckling length Lcr is necessary. This was done by judging the
buckling figure during the test and measuring the distance between the two points of inflection along the
GEC axis when the maximum normal force was reached. The comparison to the back calculated values of
EI from bending tests verify the major influence of the compressible stresses rising from the increasing
normal fore N during the tests. It can be seen that for all tests bending stiffness EI back calculated from
buckling tests is more than ten-times the EI from bending tests.

Table 3. Buckling load Ncr and comparison between back calculated bending stiffness EI from buckling and bending tests
 Geotextile Woven geotextile - Type 1 Woven geotextile - Type 2a
Initial relative
0,90 0,75 0,90 0,75
density ID
Buckling load Ncr 1,6 kN 1,2 kN 6,3 kN 5,6 kN
Back calculated EI
0,099 kNm2 0,084 kNm2 0,238 kNm2 0,191 kNm2
form buckling test
Back calculated EI
0,007 kNm2 0,006 kNm2 0,014 kNm2 0,012 kNm2
form bending test

4 CONCLUSIONS

For checking the buckling load for Geotextile-encased Columns (GECs) bedded by ultra-soft soils the
bending stiffness is a decisive parameter that is influenced by the complex interaction between the fill ma-
terial and the geotextile. In test series made up of bending and buckling tests the influence of the geotex-
tile stiffness, the fill material and the density of the fill material on the bending stiffness was found to be
dominant. Furthermore, it was verified that the bending tests give clearly smaller values of the back calcu-
lated bending stiffness than the buckling tests. This is mainly because of the great influence of compres-
sive stresses within the fill material that increase its stiffness strongly. Rupture of the geotextile was not
observed while reaching the buckling load.
The paper gives a frame to calculate buckling loads of a GEC. Because of the highly permeable GEC
the soil support can be determined considering drained loading conditions. The initial imperfections
should be covered regarding the installation method and the subsoil conditions. Until there is no further
information it is recommended to back calculate the bending stiffness from loading tests. A bending test
can be performed easily and fast without any further loading equipment. Nevertheless the bending test
does not cover the effect of the normal force within the GEC resulting in much higher values of the bend-
ing stiffness. From the test experiences a 10-fold of the measured bending stiffness during bending tests
can be used for estimating the bending stiffness to be used for proofing GEC-buckling.

REFERENCES

Alexiew, D., Moormann, C. 2010. Foundation of a coal/coke stockyard on soft soil with geotextile encased columns and hori-
zontal reinforcement. 9th International Conference on Geosynthetics, Guarujá, Brazil.
Alexiew, D., Raithel, M., Küster, V.,Detert, O. 2012. 15 years of experience with geotextile encased granular columns as foun-
dation system.; ISSMGE – TC 211 International Symposium on Ground Improvements IS-GI, Brussels
Bräu, G. 1999. Geotextile Bauweisen beim Ausbau der BAB A8 München-Salzburg imBereich des Chiemsees Heft 8,
Schriftreihe der Arbeitsgruppe „Erd- und Grundbau“ FGSV – Erd- und Grundbautagung, Erfurt.
EA-Pfähle. 2012. Empfehlungen des Arbeitskreises „Pfähle“, Deutsche Gesellschaft für Geotechnik e.V., Issue 2, Publisher
Ernst & Sohn, Berlin.
Hansen, B. J. 1961. The Ultimate Resistance of Rigid Piles against Transversal Forces. Geoteknish Institute, Copenhagen, Bul-
letin No. 12
Kempfert, G., Wallis, P. 1997. Geokunststoffummantelte Sandsäulen – ein neues Gründungsverfahren im Verkehrswegebau;
geotechnik Sonderheft zur 5. Informationsveranstaltung über „Kunststoffe in der Geotechnik“, Munich.
Kempfert, G., Raithel, M., Möbius, W., Wallis, P. 2002. Polderumschließung durch einen auf geokunststoffummantelten
Sandsäulen gegründeten Deich; 27. Baugrundtagung, Mainz
Küster, V., Raithel, M., Alexiew, D. 2011. Zum Langzeitverhalten von Gründungen mit geokunststoffummantelten Säulen;
Österreichische Geotechniktagung
Meier, T., (2009): Application of Hypoplastic and Viscohypoplastic Constitutive Models for Geotechnical Problems; Veröf-
fentlichungen des Institutes für Bodenmechanik und Felsmechanik der Universität Fridericiana in Karlsruhe, Issue 171.
Ofner, R., Wimmer, H. 2007. Knickbemessung von Mikropfählen in weichen Böden. Bauingenieur, Volume 82, Issue 5, p.
206-213.
Ohde, J. 1939. Zur Theorie der Druckverteilung im Baugrund; Der Bauingenieur, Volume 20, p. 451-459
Raithel, M. 1999. Zum Trag- und Verformungsverhalten von geokunststoffummantelten Sandsäulen. Dissertation. Heft 6,
Schriftreihe Geotechnik, Universität Gh Kassel
Raithel, M., Werner, S., Küster, V., Alexiew, D. 2011. Analyse des Trag- und Verformungsverhaltens einer Gruppe geokunst-
stoffummantelter Säulen im Großversuch. Bautechnik, Volume 88, Issue 9, p. 593-601.
Raithel, M., Alexiew, D., Küster, V. 2012. Loading Test on a Group of Geotextile Encased Columns and Analysis of the Bear-
ing and Deformation Behaviour and Global Stability. International Conference on Ground Improvement and Ground Con-
trol (ICGI 2012)
Sobolewski, J., Raithel, M., Küster, V., Friedl, G. 2012. A2 Highway Embankment in Poland founded on Geotextile Encased
Columns (GEC) – Case History Report with Monitoring Data. Proceedings of the 5th European Geosynthetics Congress.
Valencia
Vogt, N., Vogt, S. & Kellner, C. 2005. Knicken von schlanken Pfählen in weichen Böden. Bautechnik 82 (2005), Heft 12, p.
889-901.
Vogt, N., Vogt, S. & Kellner, C. 2009. Buckling of slender piles in soft soils. Bautechnik Supplement: Geotechnical Engineer-
ing, Volume 86, Issue 1, p. 98-112.
Vogt, S. & Vogt, N. 2011. Zeitabhängigkeit von Stützspannungen sehr weicher Böden und ihre Relevanz bei der Bemessung
stabilitätsgefährdeter Pfahlgründungen. Bauingenieur, Volume 88, Issue 2, p. 84-94.
Wennerstrand, J. & Fredriksson, A. 1988. Capacity of slender Steel Piles. 1st Geotechnical Seminar on Deep Foundations and
Auger Piles, Gent.
Randolph, M. F. & Houlsby, G. T. 1984. The limiting pressure on a circular pile loaded laterally in cohesive soil. Geotech-
nique, Volume 34, Issue 4, p. 613-623.

View publication stats