3 9th International Conference on Geosynthetics, Brazil, 2010

The influence of hoop stresses and earth resistance on the reinforcement

mechanism of single and multiple geocells
Emersleben, A.
Institute of Geotechnical Engineering and Mine Surveying, Clausthal University of Technology, Erzstrasse 18, 38678 Clausthal-
Zellerfeld, Germany, e-mail: [email protected]
Meyer, M.
Institute of Geotechnical Engineering and Mine Surveying, Clausthal University of Technology, Erzstrasse 18, 38678 Clausthal-
Zellerfeld, Germany, e-mail: [email protected]

Keywords: geocells, soil reinforcement, soil stabilization, index testing

ABSTRACT: Geocell structures consist of a series of interconnected single cells, made from different poly-
mer materials. The cell walls completely encase the infill material and provide an all-around confinement to
the soil. The confinement effect is based on three main mechanisms: active earth pressure within loaded cell,
earth resistance in the adjacent cells and hoop stresses in the cell walls. To evaluate the influence of different
parameters e.g. geocell stiffness, number of interconnected cells, geocell height and soil cover height on the
interaction mechanism between hoop stresses and earth resistance static radial load tests were carried out. The
test results have shown that the stiffness of the geocell material and the number of adjacent cells are the most
important parameters. The test concept and the test results are presented in the paper.
2 MATERIALS

1 INTRODUCTION 2.1 Soil

Different authors have carried out triaxial tests on An uniform dry sand with a maximum grain size of
single and multiple geocells to evaluate the influence 2 mm was used as confinement material of single
of the mobilized hoop stresses in the cell walls on geocells and infill material of multiple geocells. The
the shear strength of the infill material (Bathurst & soil is classified as SP according to the Unified Soil
Karpurapu 1993; Madhavi Latha & Murphy 2007; Classification System (USCS). The sand was filled
Rajagopal et al. 1999). They reported that the geo- in the geocells by hand and was compacted after-
cells mobilized an apparent cohesion to the soil wards by a hand tamper to a relative density of at
while the friction angle of the confined and uncon- least 80 percent. The relevant soil parameters of the
fined was the same. The value of the cohesion was sand are shown in table 1.
depending on the geocell material stiffness and the
number of adjacent geocells. The cohesion was in- Table 1. Characteristics of soil
creasing both with increasing stiffness of the geocell
characteristics sand (SP)
material and the number of adjacent cells. Neverthe-
less no attempt was made to quantify the influence coefficient of uniformity U 2.10
of the geocells other than its influence on the peak coefficient of curvature Cc 1.00
strength. To evaluate the strain-dependent influence porosity in loosest state 0.45
of different parameters on the mobilization of hoop porosity in densest state 0.34
stresses within the geocell walls and the earth resis- angle of friction  38.9°
tance in the surrounding soil/adjacent geocells radial
load tests were carried out. A horizontal load is ap- 2.2 Geocells
plied on the cell walls of a single or multiple geocell
structure, while the strains in the cell wall and the Three different types of geocells were used in
horizontal stresses within the surrounding soil were model tests. Geocell “Type 1” was made from high
measured in different distances to the loaded cell. density polyethylene (HDPE) with a density of
Due to the application of a horizontal load influ- 0.95 g/cm³. Single cells are 210 mm long and
ences e.g. of the soil beneath the geocells can be dis- 250 mm width. Single cells with a cell area of 262
regarded and the interaction between hoop stresses cm² were welded together to form a uniform geocell
and earth resistance can be evaluated without the in- mattress. The cell walls are perforated with 10 mm
fluence of other factors. diameter holes. The total open area is 16 % of the

713
cell wall area. The surface of cell walls is textured. tions of the loaded geocell was the same as in field
Geocell “Type 2” was made from the same material construction. For the type 1 and type 2 geocells the
but was manufactured without perforations. Geocell opening angle was chosen to be 90 degree. Because
“Type 3” was made form thermally solidified non- of less stiffness of geocell material the opening an-
woven (PES). The height of all geocells was 15 cm. gle of the type 3 geocell was 130 degree. After
preparation of the geocells the earth pressure cells
Table 2. Characteristics of geocell material were installed in different distances to the air pres-
sure bag (figure 1).
characteristics type 1 type 2 type 3
wall thickness (mm) 1.70 1.70 1.50
secant modulus 5 5
1.59 x 10 3.68 x 10 1.85 x 105
at 2 % strain [N/cm]
secant modulus
1.50 x 105 6.95 x 105 1.42 x 105
at 5 % strain [N/cm]

3 RADIAL LOAD TESTS

3.1 Test Device and Measurement Equipment

The radial load tests were conducted in a test device

in the soil mechanical laboratory of Institute of Geo-
technical Engineering and Mine Surveying, Claust-
hal University of Technology. The test device con-
sists of a test box with inside dimensions of 1.27 m
in width, 1.27 m in length and 0.30 m in depth. Ra-
dial load is applied over an air pressure bag with a
diameter of 210 mm and a height of 300 mm which
is placed in the centre of the test box (figure 1).
Maximum static loads ranging up to 600 kN can be
applied. The air pressure bag is placed directly
within a single geocell and the load is applied radial
symmetrically to the geocell wall.
Figure 1. Radial load test device
The deformations of cell walls were measured by an
electrical potentiometer which is placed outside the After installation of earth pressure cells dry sand
test box. The potentiometer is connected to the geo- was poured into the tests box by hand in different
cells by a wire and the wire is passed through a layers and compacted by a tamper afterwards. The
trench into the test box. The strains within the cell infill density was controlled by the infill weight of
walls are back calculated from the measured defor- the sand and the volume of the test box. After the
mation on basis of mathematical coherences. The test preparation was finished the tests were started.
horizontal pressure in the surrounding soil is meas- Static tests were carried out deformation controlled
ured by four earth pressure cells in distances of 10 until a failure of the tests samples was observed.
cm, 20 cm, 30 cm and 40 cm to the air pressure bag
(Figure 1). The earth pressure cells (EPC) have a di- 3.3 Test Program
ameter of 5 cm and a maximum pressure capacity of
500 kN/m² (figure 1). Deformation of the surface is Five different test series were carried out to evaluate
measured by six inductive settlement transducers the influence of the number of adjacent cells, open-
which are placed in different distances to the air ing of junctions, material stiffness, geocell height
pressure bag (figure 1). and cover thickness on the stress-strain behaviour of
geocells and the horizontal pressure distribution in
3.2 Test Preparation the surrounding soil. Each test series was carried out
with three different numbers of adjacent geocells
The air pressure bag was placed in the centre of the (one, nine and twenty-five). Each single test was
test box. The centre geocell of the geocell structure carried out at least three times to control the repro-
was placed on the air pressure bag and the strain ducibility of the test results. Because of the excellent
measurement system was connected to the centre reproducibility the average values of three test re-
geocell. To simulate in-situ field conditions an initial sults were taken for further analysis.
pressure was applied until the opening of the junc-

714
4 TEST RESULTS 4.2 Influence of the opening angle of the junctions

4.1 Influence of number of adjacent cells Expanding of geocells during installation at the con-
struction site leads to an opening of the junctions.
Because the restraint of infill material and the earth Index tests have show that the opening angle of the
resistance increases also the horizontal pressure in- junctions is dependent on the stiffness of the geocell
creases with increasing number adjacent cells inde- material. With increasing stiffness of the geocell ma-
pendent of the geocell type (figure 2). Large scale terial the opening angle of junction is decreasing.
model tests (Emersleben & Meyer 2008a, 2008b) The opening angle has got a large influence on the
have shown no further increase of load-settlement stress-strain behaviour of soil confined and uncon-
behaviour if the number of geocells was increased to fined geocells (figure 4). With increasing opening
more than twenty-five. angle smaller horizontal pressure can be applied at
the same strain level.
180
240
type 2 geocell, h = 15 cm
type 1 geocell, h = 15 cm 220 type 2 geocell with soil confinement, h = 15 cm
160
horizontal pressure [kN/m²]

type 3 geocell, h = 15 cm type 2 without soil confinement, h = 15 cm

200

horizontal pressure [kN/m²]

HDPE glatt orig., h= 15 cm
at 5 % strain
infill material: sand 180 angle of junction 90°
140 angle of junction 180°
160
140
120
120 angle of junction 150°
100
100
80
opening angle
60 of junctions 30°
80
40
influence of soil
20 confinement or earth resistance
60
0 5 10 15 20 25 0
number of geocells 0 2 4 6 8 10 12 14
Figure 2. Influence of number of geocells on the ho- strain [%]
rizontal pressure at 5 % percent strain Figure 4. Influence of junction opening angle on the
stress-strain behavior
Because of the increasing soil restraint with increas-
ing number of geocells within the cell walls the At the same time the influence of the soil confine-
measured horizontal pressure within the surrounding ment/earth resistance is decreasing with increasing
soil also increase with increasing number of geocells opening angle. If a horizontal load is applied and the
(figure 3). opening angle is small first of all the junctions are
opened. Due to the opening no strains and because
of that no hoop stresses are mobilized within the
horizontal pressure in soil EPC [kN/m²]

14 25x type 2 geocells, h = 15 cm geocell walls. But because of the large deformations
9x type 2 geocells, h = 15 cm
1x type 2 geocells, h = 15 cm large amount of earth resistance is mobilized. With
12 exp. Fit 25x type 2 larger opening angles hoop stresses are mobilized
exp. Fit 9x type 2
exp. Fit 1x type 2 within the geocell walls and the applied pressure is
10
completely taken by the geocell material. Only small
horizontal pressure: 80 kN/m²
infill material: sand earth resistance is mobilized within the surrounding
8
soil.
6
4.3 Influence of geocell stiffness
4
The influence of the material stiffness on the open-
2 ing angle of the junctions is described above. The
material stiffness also has got a large influence on
0 the horizontal pressure. Independent of the number
10 20 30 40 of interconnected geocells the horizontal pressure
distance to air pressure bag [cm] which can be applied to the system is increasing
Figure 3. Influence of number of geocells on the ho- with increasing geocell stiffness (figure 5).
rizontal pressure distribution in the surrounding soil
at 80 kN/m²

715
 300 The horizontal stress distribution in the surrounding
25x geocell with sand infill soil/geocell system is also influenced by the material
270 9x geocells with sand infill
1x geocell with sand confinement
stiffness (figure 6). With increasing material stiff-
horizontal pressure [kN/m²]

240 1x geocell without sand confinement ness the measured horizontal pressure in different
infill material: sand
210 at 5% strains distances to the air pressure bag is decreasing. The
y 0, 0014 ˜ x  62, 28 stresses are transferred over a larger area in fact of
180 lower material stiffness. Because the geocells with
150 y 0, 0013 ˜ x  58,15
higher material stiffness carry more loads by hoop
stresses than geocell with smaller stiffness at the
120 same load or strain level, smaller amount of pressure
extrapolated
90 extrapolated is transferred into the adjacent cells. Because of that
the restraint of the soil within the adjacent cells is
60 y 0, 00089 ˜ x  60,13
higher at a smaller material stiffness.
30 y 0, 0015 ˜ x  22, 43
59.840 N
160 up to 4210 N
4.4 Influence of geocell height and cover thickness
0
0 15.000 30.000 45.000 60.000 75.000
geocell stiffness EA [N] With increasing geocell height and cover thickness
of the adjacent cells the horizontal pressure increase
Figure 5. Influence of geocell stiffness on the hori- independent of the number of adjacent cells and
zontal pressure at 5 percent strain, linear regression stains. At the same time the horizontal pressure dis-
for all tests tribution in the surrounding soil also increases with
increasing geocell height and cover thickness be-
The regression analysis of all tests at 5 percent strain cause of an increase of the restraint of soil.
has shown that the influence of the soil confine-
ment/earth resistance on the horizontal pressure is
decreasing with increasing geocell stiffness. If the 5 REFERENCES
geocell material has got stiffness (at 5% strain) of
more than 59.840 N the soil confinement of a single Bathurst R.J., Karpurapu R. (1993): Large Scale tri-
cell has no influence on the horizontal pressure. At a axial compression testing of geocells reinforced
calculated stiffness of 223.000 N and 498.000 N granular soils, Geotechnical Testing Journal
nine and twenty-five interconnected geocells have 1993, ASTM, Vol. 16, 296 – 303
no influence on the horizontal pressure. At these Emersleben A.; Meyer M. (2008a): Bearing Capaci-
stiffness values the horizontal pressure is taken by ty Improvement of Gravel Base Layers in Road
the large mobilized hoop stresses within the geocell Constructions using Geocells. International
walls. Because of that the soil confinement/adjacent Conference for computer Methods and Ad-
cells leads to no further increase of horizontal pres- vances in Geomechanics (IACMAG), S. 3538-
sure. Same analysis can be done for all strain levels. 3545, Goa, Indien
Emersleben A.; Meyer M. (2008b): The use of geo-
18 cells in road constructions - falling weight def-
horizontal pressure in soil EPC [kN/m²]

16
25x type 1 geocell, h = 15 cm
25x type 2 geocell, h = 15 cm
lecetor and vertical stress measurements. Euro-
25x type 3 geocell, h = 15 cm Geo 4, Proceedings of the 4th European
14 log. Fit 25x type 1 geocell
Geosynthetics Conference, Edinburgh, Schott-
log. Fit 25x type 2 geocell
12
log. Fit 25x type 3 geocell land
horizontal pressure: 80 kN/m²
Emersleben A.; Meyer M. (2008c): Bearing capaci-
10 infill material: sand ty improvement of asphalt paved road construc-
tion due to the use of geocells - falling weight
8 EA = 21.400 N und 22.500 N
deflecetor and vertical stress measurements.
6 Geosynthetics Asia 2008 "Geosynthetics in civil
and environmental engineering, Proceedings of
4 the 4th Asian Regional Conference in Geosyn-
EA = 55.200 N thetics, S. 474-753, Shanghai, China 200
2
Madhavi Latha G., Vidya S. Murphy (2007): Effects
0 of reinforcement form on the behavior of geo-
10 20 30 40 synthetic reinforced sand, Geotextiles and
distance to air pressure bag [cm] Geomembranes, Volume 25, Issue 1, 23-32
Rajagopal K., Krishnaswamy N. R., Madhavi Latha
Figure 6. Influence of geocell stiffness on the hori-
(1999): Behaviour of Sand reinforced with sin-
zontal pressure distribution in the surrounding soil at
gle and multiple geocells, Geotextilies and
horizontal pressure of 80 kN/m²
Geomembranes, Vol. 17, 171 – 184

716