You are on page 1of 18

ARTICLE IN PRESS

Geotextiles and Geomembranes 22 (2004) 205222

Lateral and axial deformation of PP, HDPE and


PET geogrids under tensile load
Masahiro Shinodaa, Richard J. Bathurstb,*
a

Railway Technical Research Institute, Foundation and Geotechnical Engineering Structures, Technology
Division, 2-8-38, Hikari-cho, Kokubunji-shi, Tokyo 185-8540, Japan
b
GeoEngineering Centre at Queens-RMC, Civil Engineering Department, Sawyer Building, Room 2085,
Royal Military College of Canada, 13 General Crerar, Kingston, Ont., Canada K7K 7B4
Received 20 March 2003; received in revised form 10 March 2004; accepted 15 March 2004

Abstract
The paper describes a series of short-term in-isolation tensile tests that were carried out to
investigate the loadstraintime performance of typical knitted polyester (PET), biaxial
polypropylene (PP) and uniaxial high-density polyethylene (HDPE) geogrid products. The
specimens were subjected to constant rate of strain (CRS) loading and variable rate of strain
(VRS) loading. Variables between test specimens included specimen length and specimen
aspect ratio (e.g. number of ribs). A novel feature of these tests was that internal strains in the
specimens were recorded using non-contact measurements taken with a video-extensometer
apparatus. The data shows that at large strains the PP and HDPE geogrid specimens exhibited
large lateral strains while lateral strains were negligible for the PET product. Results of CRS
tests carried out with different specimen aspect ratios showed that there was no change in
tensile strength, axial strain and lateral strain at rupture for HDPE and PET materials. The PP
material in this investigation did show an increase in axial and lateral strain at rupture for a
specimen tested at an aspect ratio approaching one. The polyolen geogrids (PP and HDPE)
were shown to exhibit tensile stiffness and ultimate tensile load capacities that increased with
rate of strain. Implications of the contractive behaviour of the polyolen geogrid products to
the selection of in-soil reinforcement stiffness values and rheological modelling are identied.
r 2004 Elsevier Ltd. All rights reserved.
Keywords: Geosynthetics; Geogrids; Tensile strength; Axial strains; Lateral strains; Stiffness; Videoextensometer

*Corresponding author. Tel.: +1-613-541-6000 ext 6479/6347/6391; fax: +1-613-545-8336.


E-mail address: bathurst-r@rmc.ca (R.J. Bathurst).
0266-1144/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.geotexmem.2004.03.003

ARTICLE IN PRESS
206

M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

1. Introduction
In-isolation tensile loading of geosynthetic reinforcement products according to
ASTM test methods D 4595, D 6637 and D 5262 is routine practice to characterise
the loadstrain properties of these materials.
The importance of measurement of tensile load-strain properties of geotextiles and
geogrids under soil connement has been noted in the literature (McGown et al.,
1982, 1994, El-Fermaoui and Nowatzki, 1982, Siel et al., 1987, Christopher et al.,
1986, Leshchinsky and Field, 1987). One important nding from in-soil tests was
that there can be a signicant increase in the stiffness and strength of some
geotextiles conned in soil compared to the unconned condition (Ling et al., 1992,
Boyle et al., 1996). Yuan et al. (1998) performed in-soil constant rate of strain tests
on woven and nonwoven polypropylene (PP) geotextiles, an extruded high-density
polyethylene (HDPE) geogrid and a woven polyester (PET) geogrid. Walters et al.
(2002) reinterpreted their test results and concluded that soil connement has a large
effect on the secant stiffness values (measured at 5% axial strain) for nonwoven
geotextiles (an increase of up to 500% depending on the conning pressure and soil
type), a small but measurable inuence on the stiffness values of woven geotextiles
and geogrids (an increase of 530% over the in-isolation value), and no inuence on
the extruded HDPE geogrid stiffness response.
In-isolation tensile tests on nonwoven geotextiles show that necking or lateral
contraction occurs during axial extension (Rowe and Ho, 1986). Surprisingly, the
onset and magnitude of lateral contractions for both geotextiles and geogrid
reinforcement products during testing has not been the subject of quantitative
investigation. Geogrid reinforcement products are routinely used in reinforced soil
retaining wall construction. The selection of an appropriate reinforcement stiffness
value has been demonstrated to be a precursor to the correct estimation of
reinforcement loads from in situ strain readings with consideration of the duration of
loading (Allen and Bathurst, 2002). Lateral contraction of geogrids under tensile
load is, therefore, of interest if the argument is accepted that constrained lateral
deformation of these materials in-soil can lead to a stiffer axial response of these
materials compared to the same response in air. Evidence of lateral contraction of
typical geogrid reinforcement products tested in-isolation and the axial strain levels
at which lateral contraction is initiated could lead to the need to re-examine current
assumptions regarding the applicability of in-isolation tensile test results to the insoil environment.

2. Objective
The principle objective of the investigation described here was to examine the inisolation lateral and axial deformation of three typical geogrid reinforcement
products up to rupture using different specimen dimensions and different axial
loading paths and loading rates. A novel feature of the testing program was the

ARTICLE IN PRESS
M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

207

measurement of local axial and lateral displacements using a non-contact videoextensometer technique (Shinoda and Bathurst, 2004).

3. Test program
3.1. Geogrid materials
Specimens of biaxial PP geogrid, uniaxial HDPE geogrid and a knitted PET
geogrid were used in the current investigation (Table 1). These materials are typical
geogrid reinforcement products used in soil reinforcement applications. All three
materials have been tested previously at RMC as part of an on-going program of
full-scale testing of retaining walls and slopes (Bathurst et al., 2002). The visco-elastic
behaviour of the PP geogrid material used in this study has also been investigated by
Kaliakin et al. (2000) and Thornton (2001).
3.2. Loading apparatus
A MTS servo-controlled hydraulic actuator with a stroke of 160 mm and rated to
100 kN was used to carry out the tensile load tests. The controller for this actuator
provided strain rate and load control functions and the ability to switch between
modes without interruption. Roller clamps described by Bathurst and Cai (1994)
were used to grip the geogrid specimens (Fig. 1a). A load cell mounted between the
actuator piston and top loading clamp was used to record the tensile load during
each test. Axial displacements at the top clamp were recorded by an LVDT mounted

Table 1
Geogrid properties
Property

Materials

Polymer type
Structure
Coating
Mass/unit area (g/m2)
Aperture size (mm)
Machine direction
Cross-machine direction
Thickness (mm)
At longitudinal member
At junction
Wide-width tensile strength (kN/m)
At 5% strain
Ultimate

PP
Punched sheet and drawn
Uncoated
215

PET
Knitted
PVC
114

HDPE
Punched sheet and drawn
Uncoated
NA

25
33

27
22

140
15 (maximum)

1.0
2.9

1.1
1.2

1.0
2.7

8.3
12.5

4.4
17.5

35.7
68.9

NA = Not available; PP = polypropylene; PET = polyester; HDPE = high-density polyethylene; PVC


= polyvinyl chloride. Sourcemanufacturers literature.

ARTICLE IN PRESS
208

M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

Fig. 1. General arrangement of wide-width strip tensile loading apparatus, test specimens and strain
monitoring points on specimens. Note: Targets on HDPE specimens placed vertically from midpoint of
transverse member to midpoint of longitudinal members. (a) Cross-section view of test apparatus, (b) PP
specimen, (c) PET specimen, (d) HDPE specimen.

in the MTS actuator piston. A 0.2-kN-preload was applied to each specimen prior to
testing to remove any initial slack in the specimen, particularly at the roller clamps.
All tests were carried out at 2072 C consistent with related ASTM specications.
3.3. Strain monitoring
A high-resolution digital Charged Coupled Device (CCD) camera (videoextensometer) and ancillary data acquisition system were used to record specimen

ARTICLE IN PRESS
M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

209

axial load and displacements of targets mounted directly on longitudinal members of


the test specimens. An advantage of this approach is that local displacements of
geogrid points can be tracked in X and Y directions at locations well removed from
the specimen clamps. Another advantage of this equipment is that specimen response
is not inuenced by the measurement technique as may be the case for clip-on
extensometers and strain gauges glued directly to the surface of geosynthetic test
specimens. In this investigation the targets comprised 2-mm-diameter painted white
circles. The video-extensometer camera and software locks on to points dened by
the contrast between the white targets and the black geogrid surface. The X and Y
displacements of each monitoring point are recorded and specimen strains calculated
from selected pairs of targets. The system used in this investigation is capable of
tracking multiple points at a frequency of up to 25 Hz and plotting the displacements
and computed strains against load in real time. The resolution of the measurements
is a function of the size of the eld of view. In this investigation the measurements
were taken at a resolution of 0.64 mm. The ease of use, exibility and high resolution
of strain measurements using a video-extensometer system has been demonstrated by
Jones (2000) who used similar equipment to monitor load-extension tests on
geotextiles. Additional details of the equipment and monitoring technique used in
the current study have been reported by Shinoda and Bathurst (2004).
Preliminary tests showed that axial and lateral strains calculated from videoextensometer targets placed on adjacent junctions or at points midway between
junctions of PET and PP specimens gave the same values indicating that strain
distributions were uniform for these materials. Local axial tensile strains reported
hereafter were calculated using displacement measurements for targets a and b
and lateral strains using targets a and d located close to the centre of PET and
PP specimens (Fig. 1b and c). Non-uniform strain distributions were observed for
the HDPE specimens due to local variation in the cross-sectional area and modulus
of the drawn polymer material. Non-uniform axial strain distributions in thick
uniaxial drawn HDPE geogrids have also been noted by McGown et al. (1994). In
the current investigation, six targets were placed on the surface of the HDPE
specimens in the pattern illustrated in Fig. 1d. Targets a and d were located at
approximately the midpoint of the longitudinal members between transverse
members. Average axial tensile strains were calculated using deformations recorded
between points a and c, and d and f. Lateral strains were calculated by
taking the average of horizontal deformations for pairs of points at the same
elevation on the specimen. This array of six targets allowed average strains
corresponding to one aperture spacing in longitudinal and transverse directions to be
calculated as well as to investigate the distribution of lateral strains between
longitudinal members.
3.4. Loadextensiontime paths for tensile tests
3.4.1. Constant rate of strain (CRS) tests
As part of the development of the experimental methodology, reference
conventional CRS tests at 10% strain/minute (based on cross head speed) were

ARTICLE IN PRESS

Axial strain

M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

Axial strain

210

Time

Axial load

Axial load

Time

(a)

Time

(b)

Time

Fig. 2. Loadextensiontime paths for test program. (a) Constant rate of strain (CRS) test, (b) variable
rate of strain (VRS) test.

carried out on each material type (Fig. 2a). These tests were repeated three times to
conrm that test results were repeatable (Shinoda and Bathurst, 2004). In the current
investigation, two series of CRS tests were performed on specimens trimmed to
different lengths and widths. The rst series comprised 200-mm-wide PP and PET
geogrid specimens trimmed to lengths of 224, 314, 432 and 540 mm and 226, 286, 395
and 496 mm, respectively. A second series of tests was carried out at 10% strain/
minute on specimens trimmed to a constant length of approximately 200 mm but
with different widths (representing different number of longitudinal members). The
PP specimens were prepared with 2, 4, 5 and 6 longitudinal members. Four PET
geogrid specimens were prepared with 2, 4, 6 and 8 longitudinal members and six
HDPE geogrid specimens with 2, 4, 6, 8, 9 and 10 longitudinal members. The
inuence of strain rate on PP, PET and HDPE geogrid specimen response was
investigated by applying strain rates of 0.1, 1.0, 10 and 100% strain/minute up to
rupture. In some cases the actuator stroke was exceeded before rupture could be
achieved.

3.4.2. Variable rate of strain (VRS) tests


The results of variable rate of strain load testing have been used by Hirakawa et al.
(2003) to verify an advanced three-component rheological model for different
polymer geogrid reinforcement materials. The inuence of variable strain rate during
load-extension of PP, PET and HDPE geogrid specimens was investigated in the
current study by changing the axial strain rate (based on cross head speed) during the
test (Fig. 2b). The strain rate was varied in the sequence of 10, 1.0, 0.1, and 1.0%
strain/min to rupture with each strain rate increment held for about 20, 450 and
4500 s, respectively for the PP and PET specimens, and 20, 250 and 2600 s for the
stiffer HDPE specimens.

ARTICLE IN PRESS
M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

211

4. Test results
4.1. Influence of specimen dimensions on loadstrain response
Fig. 3 shows the inuence of specimen length on load-extension results for
specimens with constant width (200 mm). The tensile strength at rupture (or large
strain) was reasonably constant between specimens of each type and did not show a
dependence on aspect ratio (Fig. 3a). With the exception of the shortest test
specimen, the PP specimens were not taken to rupture due to the stroke capacity of
the actuator. The average tensile strength of the PP and PET specimens was 12.9 and
17.3 kN/m, respectively, which are close to the manufacturers reported values of
12.5 and 17.5 kN/m based on wide-width strip tensile tests (ASTM D 4595) (Table
1). Small deviations from loadstrain response using mechanical extensometers and

20

PET geogrid

Axial load (kN/m)

0.40 (6)
0.70 (6)
0.88 (6)

0.51 (6)

15

X = Specimen rupture

0.37 (8)

PP geogrid

0.90 (8)

10

0.46 (8)

0.64 (8)

Number of longitudinal members


Width to length ratio
0
0

10

(a)

20
Axial strain (%)

30

40

-15
0.90 (8)
PP geogrid

Lateral strain (%)

-10

0.64 (8)

0.46 (8)
0.51 (6)
0.70 (6)

-5

0.37 (8)
PET geogrid

0.88 (6)
0.40 (6)
X = Specimen rupture
5
0

(b)

10

20

30

40

Axial strain (%)

Fig. 3. Inuence of specimen length on CRS load-extension results for specimens with constant width
(200 mm). (a) Axial load versus axial strain, (b) lateral strain versus axial strain.

ARTICLE IN PRESS
212

M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

the non-contact measurement technique employed here may be one source of the
discrepancy as demonstrated by Shinoda and Bathurst (2004). Another source of the
difference may be variations in material properties between product rolls. The loadextension response of the specimens of each material over the monitored axial strain
range were sensibly the same and appear to be independent of the aspect ratio of the
specimens for these CRS tests. Fig. 3b shows plots of lateral strain versus axial strain
for the same tests. The PET specimens exhibited very little lateral contraction
(maximum of 1.5%) while lateral strains for the integral PP specimens were
pronounced after about 10% axial strain. At axial strains >10%, the lateral strains
varied linearly with increasing axial tensile strain. The axiallateral strain responses
of the PP specimens over the range of strains investigated (i.e. local slope of the
curves) were independent of specimen dimensions.
Fig. 4 shows the inuence of specimen width on load-extension results for
specimens with constant length (200 mm). The rupture strengths of the specimens of
20
PP and PET geogrid

Axial load (kN/m)

0.79 (5)
0.60 (4)
0.21 (2)

15

X = Specimen rupture
PET geogrid

0.90 (6)
PP geogrid
0.36 (4)

10

0.13 (2)
0.53 (6)
5

0.88 (8)
Number of longitudinal members
Width to length ratio

0
0

10

(a)

20

30

40

Axial strain (%)

-15

PP and PET geogrid

X = Specimen rupture

0.90 (6)
0.21 (2)

Lateral strain (%)

-10

0.60 (4)

0.79 (5)
PP geogrid

-5
0.88 (8)
PET geogrid

0
0.36 (4) 0.53 (6)

0.13 (2)

5
0

(b)

10

20

30

40

Axial strain (%)

Fig. 4. Inuence of specimen width on CRS load-extension results for specimens with constant length
(200 mm). (a) Axial load versus axial strain, (b) lateral strain versus axial strain.

ARTICLE IN PRESS
M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

213

both types were the same (i.e. independent of specimen width) (Fig. 4a). However,
the tensile strain at rupture for the widest PP specimen (180 mm0.9 aspect ratio)
was notably larger (25% axial strain) than the narrower specimens (E 12% axial
strain). Fig. 4b shows that the lateral contraction of the PP specimens was greater for
the specimens with a width less than 180 mm (0.9 aspect ratio) at the same axial
strain. The greater lateral contraction (i.e. greater aperture distortion) may be the
reason that the narrower specimens failed at a lower axial strain. The load-extension
curves for the PET specimens were independent of specimen dimension (Fig. 4a) and
lateral contractions were small (p0.05%) (Fig. 4b). The largest contraction at
rupture was recorded by the widest specimen with 8 longitudinal members.
Nevertheless, this value is about one-half of the values reported for the longer
specimens in the previous data set. The small amount of contraction recorded for the
PET specimens in both data sets is thought to be due to straightening of the multiple
longitudinal laments during axial extension and appears to increase with specimen
length and to a lesser degree with specimen width.
The results of CRS tests on HDPE specimens with constant length but variable
width are plotted in Fig. 5. No signicant systematic variation in the load-extension
response and load or strain values at rupture can be seen in Fig. 5a. The average
lateral strains recorded for the same tests are plotted in Fig. 5b. Lateral strains at
rupture were reasonably constant for all specimens and varied from 40% to 80% of
the axial strains at rupture. Otherwise there was no consistent trend in the average
lateral strain at rupture with specimen width. The lateralaxial strain curve for the
narrowest specimen shows apparent dilation at small axial strains. However, this is
thought to be an erroneous result related to the difculty that was encountering
securing the very narrow specimen in the large roller clamps used in this
investigation.
Fig. 6a shows a summary of the rupture strengths for all specimens versus aspect
ratio. Taken together, the data shows that tensile strength at rupture was
independent of specimen dimensions for each of the three material types tested.
Fig. 6b summarises axial strain at rupture versus aspect ratio for all tests. For PET
and HDPE specimens the strain at rupture was reasonably constant. The widest PP
specimen showed a very much higher strain at rupture compared to narrower
specimens. Fig. 6c shows that the trend in data noted above for axial strains at
rupture is preserved for the lateral strains at rupture with only the PP material
showing a large increase in lateral strain at rupture for the widest specimen.
4.2. Influence of strain rate on loadstrain response
The load-extension results discussed in this section correspond to CRS and VRS
tests carried out on specimens of all three material types prepared to the same
dimensions (i.e. 220 mm long and 200 mm wideaspect ratio=0.9). The continuous
curves in the gures correspond to CRS tests carried out at a constant strain rate
(Fig. 2a). The stepped curves correspond to VRS tests carried out at variable target
strain rates (measured at the cross-head) of 10, 1.0, 0.1 and 1.0% strain/min to
rupture (Fig. 2b). The small differences in magnitude between target strain rate

ARTICLE IN PRESS
214

M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222


80

X = Specimen rupture

HDPE geogrid
0.12 (2)
0.58 (9)

Axial load (kN/m)

60

0.40 (6)

0.66 (10)
0.52 (8)
0.25 (4)

40

Number of
longitudinal members
Width to length ratio

20

0
0

10

15

Axial strain (%)

(a)
-10
Average lateral strain (%)

HDPE geogrid

X = Specimen rupture

-8

0.40 (6)
0.58 (9)

-6
0.52 (8)
0.66 (10)

-4
-2
0.25 (4)
0
0.12 (2)

2
0

(b)

10

15

Axial strain (%)

Fig. 5. Results of CRS tests on HDPE specimens with constant length but variable width. (a) Axial load
versus axial strain, (b) average lateral strain versus axial strain.

values and those noted on the following gures is that the latter were recomputed
after the specimens had been subjected to the initial preload while the former were
based on the initial distance between the cross-heads prior to preloading. For the
purpose of this investigation, these differences are considered inconsequential. The
constant VRS loadstrain responses (Fig. 7a and b) show that the polyolen
materials were very sensitive to rate of loading and axial strains at rupture decreased
with increasing rate of strain. The variable strain rate VRS tests showed that with
decreasing strain rate increments the loadstrain segments rapidly approached and
fell on top of the corresponding constant rate of strain curves. The corresponding
data for the PET specimens shows that the loadstrain response of the PET material
was sensibly independent of the strain rate-time loading path. The observations for
axial load axial-strain plots for CRS and VRS data described here also hold true for
the corresponding axial load versus lateral strain curves but are not reproduced here
for brevity.

ARTICLE IN PRESS
M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

215

80

Tensile strength (kN/m)

HDPE geogrid
60

*60.9 kN/m

40

PET geogrid

20 *17.2 kN/m

PP geogrid
* Average value

*12.8 kN/m
0
0.0

0.2

(a)

0.4

0.6

0.8

1.0

Aspect ratio

30

Axial strain at rupture (%)

PP geogrid

20
PET geogrid

*15.9 %

10
*9.1 %

0
0.0

HDPE geogrid
* Average value
0.2

(b)

0.4

0.6

0.8

1.0

Aspect ratio

Lateral strain at rupture (%)

-20

-15
PP geogrid
-10
*6.6 %
-5
*0.5 %
0

5
0.0

(c)

HDPE geogrid

PET geogrid
* Average value
0.2

0.4

0.6

0.8

1.0

Aspect ratio

Fig. 6. Inuence of specimen aspect ratio on tensile strength and strains at rupture. (a) Tensile strength
versus aspect ratio, (b) axial strain at rupture versus aspect ratio, (c) lateral strain at rupture versus aspect
ratio.

ARTICLE IN PRESS
216

M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222


20

PP geogrid
10.6% strain/minute

15
Axial load (kN/m)

X = Specimen rupture
99.3% strain/minute
10.0% strain/minute
1.0% strain/minute

10

1.1% strain/minute
0.1% strain/minute
5

1.1% strain/minute
0.1% strain/minute

0
0

10

20

(a)

30

40

Axial strain (%)


80
HDPE geogrid 98.0% strain/minute
10.1% strain/minute
1.0% strain/minute

Axial load (kN/m)

60

0.1% strain/minute
10.3% strain/minute

40

1.0% strain/minute
0.1% strain/minute

20

1.0% strain/minute
10.3% strain/minute

X = Specimen rupture

0
0

10

(b)

15

20

25

Axial strain (%)


20
PET geogrid

11.1% strain/minute
103.4% strain/minute
0.1% strain/minute

Axial load (kN/m)

15

0.9% strain/minute
9.6% strain/minute

10

1.0% strain/minute
5

0.1% strain/minute
1.0% strain/minute

0
0

(c)

X = Specimen rupture

9.5% strain/minute
10

20

30

40

Axial strain (%)

Fig. 7. Inuence of strain rate and strainrate path on load-extension response (a) PP geogrid specimens,
(b) HDPE geogrid specimens, (c) PET geogrid specimens.

ARTICLE IN PRESS
M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

217

The general trend in the lateral strainaxial strain CRS curves for the biaxial PP
specimens (Fig. 8a) was increasing magnitude of lateral strain with increasing axial
strain rate taken at the same axial strain value greater than 10% (i.e. the material
lateral stiffness decreased with increasing strain rate, or equivalently, necking of the
specimens increased). This trend was reversed for the HDPE material (Fig. 8b) which
may be due to the very different structure of the uniaxially drawn material and the
effect of averaging the lateral strains over three different rows of monitoring points.
The VRS curve segments for the same tests that are superimposed on the gures do
not consistently fall on the corresponding CRS test curves. For example, the VRS
curves continue to track over the 10% strain/min CRS curve regardless of axial
strain rate. Lateral strain response of the PET specimens was small (p0.08%) and
strain-rate path effects were not detectable at this level of deformation (Fig. 8c).
The variation in normalised tensile strength at rupture versus axial strain rate from
CRS tests is summarised in Fig. 9. The data shows a general trend of increasing
strength with increasing strain rate for polyolen materials and less strain-rate
sensitivity for the PET material.

5. Conclusions
The inuence of specimen dimensions and strain rate on the loadextension
response of geotextiles in wide-width tensile testing performed in-isolation has been
reported by others (Rowe and Ho, 1986, Shrestha and Bell, 1982). Similar
investigations of the loadstrain response of geogrids with different specimen
dimensions and the inuence of specimen geometry on strength and strains in both
lateral and axial directions have not been reported in the literature to the best of the
writers knowledge. It appears that lateral contractions of typical geogrid
reinforcement products have been assumed to be negligible in magnitude and,
perhaps for this reason, of little practical interest in the area of soil-reinforcement
interaction. This paper describes the results of a series of careful tests carried out to
examine the local load-extension response of three typical geogrid reinforcement
(integral PP and HDPE geogrids and one knitted PET geogrid) in both axial and
lateral directions. The specimen strain measurements were made using a noncontacting video-extensometer technique. The following observations were made:
(1) For the range of strain rates investigated, the strength at rupture of the PP,
HDPE and PET specimens was not sensitive to specimen aspect ratio.
(2) The axial strain and lateral strains at rupture for the HDPE and PET specimens
were also independent of specimen dimensions. However, the axial and lateral
strains at rupture were greater for the PP geogrid specimen with the largest width
(180 mm or 6 longitudinal members) compared to narrower specimens.
(3) Lateral strains at rupture were about 6080% of the axial strains at rupture for
the HDPE material and about 4% of the axial strain values at rupture for the
PET materials. The lateral strains at rupture for the PP specimens varied from
about 3050% of the axial strain at rupture.

ARTICLE IN PRESS
218

M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222


-15
PP geogrid

1.0% strain/minute

10.0% strain/minute
Lateral strain (%)

-10
Test result with
variable strain rate
-5
0.1% strain/minute
99.3% strain/minute
0

X = Specimen rupture
5
0

10

(a)

20

30

40

Axial strain (%)

-15

X = Specimen rupture

HDPE geogrid

0.1% strain/minute
Lateral strain (%)

-10
1.0% strain/minute
10.4% strain/minute

-5

Test result with variable strain rate


98.0% strain/minute

5
0

10

(b)

20

30

40

Axial strain (%)


-4
X = Specimen rupture

Lateral strain (%)

PET geogrid

-2
11.1% strain/minute
0.9% strain/minute
0
Test result with variable strain rate

2
0

(c)

10
Axial strain (%)

15

20

Fig. 8. Lateral strainaxial strain response from CRS load-extension tests. (a) PP geogrid specimens, (b)
HDPE geogrid specimens, (c) PET geogrid specimens.

ARTICLE IN PRESS
M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

219

Tensile strength /
Tensile strength at 10% strain/minute

1.50

1.25
HDPE geogrids

1.00

0.75

PET geogrids
PP geogrids

0.50
0.1

10
Strain rate (%)

100

Fig. 9. Normalised tensile strength at rupture versus axial strain rate from CRS load-extension tests.

(4) Loadstrain response in both orthogonal directions for the polyolen materials
was observed to be rate-of-strain dependent while the PET material was not.
(5) Lateral contraction of the widest PP specimens (180 and 200 mm wide) was
observed to be negligible at strains less than about 810% axial strain but to
increase almost linearly with axial strains thereafter.
(6) Lateral contraction (or necking) of the HDPE specimens was initiated at the
onset of the load-extension tests.
(7) Variable rate-of-strain loadstrain paths for the polyolen materials were
observed to fall on the corresponding curves for constant rate-of-strain
tests provided each VRS load segment was maintained for a sufcient length
of time.

6. Implications
The results of this investigation have implications to the understanding of soilstructure interaction for the types of reinforcement used in this study when used insitu and consequently for design. Fortunately, for ultimate limit state design of
reinforced structures in which the reinforcement selection is based on a reference inisolation rupture strength of the material (e.g. ASTM D 6637), the rate of strain and
the aspect ratio of the specimens do not appear to be a factor. However, for
serviceability design the conversion of strain to load requires selection of a suitable
stiffness value. The polyolen materials in the current study have demonstrated rateof-strain dependence under both CRS and VRS load paths which makes the
selection of a suitable time-dependent stiffness value problematic. Walters et al.
(2002) gave examples that showed the equivalent rate-of-strain loading of
reinforcement layers in actual walls was 4 or 5 orders of magnitude slower than

ARTICLE IN PRESS
220

M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

that of the standard CRS test carried out at 10% strain/min (ASTM D 4595, D
6637). Walters et al. proposed that for design purposes, CRS tests could be carried
out at different strain rates and this data used to calculate a 2% secant stiffness
value (J2%) as a function of time. This value can then be used to calculate an
end-of-construction reinforcement load (or post-construction loads at a prescribed
design life).
The rate of loading of a reinforcement layer is also an issue for seismic design
(Bathurst and Alfaro, 1996). The stiffness, rupture strength and strain at rupture are
functions of rate of strain for polyolen materials as demonstrated in the tests
reported here. Of particular importance is the trend toward lower rupture strains
with increasing rates of axial strain.
The loadlateral strain response records reported here give indirect evidence of the
possible inuence of soil connement on reinforcement stiffness. Reinforcement
contractions observed in-isolation may be expected to be restrained in-soil due to soil
connement, particularly in a stiff dense soil. Using this line of argument suggests
that for PP geogrids of the type investigated here, the connement effect may not be
pronounced until axial strains have achieved strains in excess of (say) 10% (e.g.
Fig. 8a). Fortunately, this is well beyond any serviceability strain limit for design.
Allen and Bathurst (2002) have estimated that a maximum reinforcement strain of
about 3% is required to generate contiguous failure zones in reinforced soil retaining
walls constructed with high-quality and well-compacted granular soil backlls and
proposed that this strain level be used for design against a soil failure limit state. The
3% reinforcement strain associated with this limit state is much less than the strain
levels observed in this investigation to generate contractive behaviour in the PP
geogrid specimens. However, for HDPE geogrids, in situ stiffness increases due to
constrained contraction may develop at the onset of axial load (e.g. Fig. 5b). Hence,
in-soil stiffness values may be higher at all load levels for this material under
operational conditions. For the knitted PET geogrid tested in this study there were
no signicant lateral contractions of the specimens and hence restrained contraction
is not a likely contributor to increased in-soil soil stiffness for this material.
Nevertheless, other mechanisms such as the interaction between geogrid transverse
members and the conning soil (i.e. passive bearing) and interface friction between
the surfaces of the geogrid members have been identied as contributors to pullout
capacity of integral drawn polyolen reinforcement products (Palmeira and
Milligan, 1990). The structure of these materials and their interaction with the soil
may also play a part to increase in-soil stiffness of these materials that is independent
of the restrained contraction effect described above.
Finally, the results of the tests reported here have implications to the development
of rheological models for planar polyolen geogrid reinforcement materials such as
the products used in this investigation. Advanced loadstraintime models for
planar geosynthetic reinforcement products are exclusively one-dimensional. The
load-extension data for the PP and HDPE geogrid products reported in this study
have demonstrated a pronounced two-dimensional response under axial tensile
loading that should be considered if rheological models are to be used to predict
reinforcement behaviour, particularly at large strains.

ARTICLE IN PRESS
M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

221

Acknowledgements
The rst author is grateful for the nancial support from Integrated Geotechnical
Institute Limited of Tokyo, Japan to carry out the work described in this paper while
a visiting Research Fellow with the GeoEngineering Centre at Queens-RMC at the
Royal Military College of Canada, Kingston, Ontario, Canada. Financial support
was also provided by the Natural Sciences and Engineering Research Council of
Canada and the Department of National Defence (Canada) in the form of
equipment and operating grants awarded to the second author.

References
Allen, T.M., Bathurst, R.J., 2002. Soil reinforcement loads in geosynthetic walls at working stress
conditions. Geosynthetics International 9 (56), 525566.
ASTM D 4595. Standard Test Method for Tensile Properties of Geotextiles by the Wide-width Strip
Method. American Society for Testing and Materials, West Conshohocken, Pennsylvania, USA.
ASTM D 5262. Standard Test Method for Evaluating the Unconned Tension Creep Behavior of
Geosynthetics. American Society for Testing and Materials, West Conshohocken, Pennsylvania, USA.
ASTM D 6637. Standard Test Method for Determining Tensile Properties of Geogrids by the Single or
Multi-rib Tensile Method. American Society for Testing and Materials, West Conshohocken,
Pennsylvania, USA.
Bathurst, R.J., Alfaro, M.C., 1996. Review of seismic design, analysis and performance of geosynthetic
reinforced walls, slopes and embankments. Invited keynote paper, IS-Kyushu 96, Third International
Symposium on Earth Reinforcement, Fukuoka, Kyushu, Japan, 1214 November 1996, Balkema, Vol.
2, pp. 887918.
Bathurst, R.J., Cai, Z., 1994. In-isolation cyclic load-extension behavior of two geogrids. Geosynthetics
International 1 (1), 119.
Bathurst, R.J., Walters, D.L., Esfehani, M., El-Emam, M., Blatz, J.A., 2002. Physical modelling of
geosynthetic walls and embankments, Keynote paper, International Conference on Physical Modeling
in Geotechnics, (ICPMG), St. Johns, Newfoundland, pp. 2130.
Boyle, S.R., Gallagher, M., Holtz, R.D., 1996. Inuence of strain rate, specimen length and connement
on measured geotextile properties. Geosynthetics International 3 (2), 205225.
Christopher, B.R., Holtz, R.D., Bell, W.D., 1986. New tests for determining the in-soil stressstrain
properties of geotextiles. Proceedings of the Third International Conference on Geotextiles, Vienna,
Austria, Vol. 3, pp. 683688.
El-Fermaoui, A., Nowatzki, E., 1982. Effect of conning pressure on performance of geotextiles in soils.
Proceedings of the Second International Conference on Geotextiles, Las Vegas, Nevada, USA, Vol. 3,
pp. 799804.
Hirakawa, D., Kongkitkul, W., Tatsuoka, F., Uchimura, T., 2003. Time dependent stressstrain
behaviour due to viscous properties of geogrid reinforcement. Geosynthetics International 10 (6),
176199.
Kaliakin, V.N., Dechasakulsom, M., Leshchinsky, D., 2000. Investigation of the isochrone concept for
predicting relaxation of geogrids. Geosynthetics International 2 (2), 7999.
Jones, D., 2000. Wide-width geotextile testing with video extensometer, In: Stevenson, P.E. (Ed.), Grips,
Clamps, Clamping Techniques, and Strain Measurement for Testing of Geosynthetics. ASTM STP
American Society for Testing and Materials, West Conshohocken, Pennsylvania, USA, Vol. 1379,
pp. 8388.
Leshchinsky, D., Field, D.A., 1987. In-soil load elongation, tensile strength and interface friction
of nonwoven geotextiles. Proceedings of Geosynthetics 87, New Orleans, Louisiana, USA,
pp. 238249.

ARTICLE IN PRESS
222

M. Shinoda, R.J. Bathurst / Geotextiles and Geomembranes 22 (2004) 205222

Ling, H.I., Wu, J.T.H., Tatsuoka, F., 1992. Short-term strength and deformation characteristics of
geotextiles under typical operational conditions. Geotextiles and Geomembranes 11 (2), 185219.
McGown, A., Andrawes, K.Z., Kakir, M.H., 1982. Load extension testing of geotextiles conned in soil,
Proceedings of the Second International Conference on Geotextiles, Las Vegas, Nevada, USA, Vol. 3,
pp. 793798.
McGown, A., Yogarajah, I., Andrawes, K.Z., Saad, M.A., 1994. Strain behaviour of polymeric geogrids
subjected to sustained and repeated loading in air and in soil. Geosynthetics International 1 (3),
341355.
Palmeira, E.M., Milligan, G.W.E., 1990. Scale and other factors affecting the results of pull-out tests of
grids buried in sand. G!eotechnique 39 (3), 511524.
Rowe, R.K., Ho, S.K., 1986. Determination of geotextile stressstrain characteristics using a wide
strip test. Proceedings of the Third International Conference on Geotextiles, Vienna, Austria, Vol. 3,
pp. 885890.
Shinoda, M., Bathurst, R.J., 2004. Strain measurement of geogrids using a video-extensometer technique.
ASTM Geotechnical Testing Journal, in press.
Shrestha, S.C., Bell, J.R., 1982. A wide strip tensile test of geotextiles. Proceedings of the Second
International Conference on Geotextiles, Las Vegas, Nevada, USA,Vol. 3, pp. 739744.
Siel, B.D., Tzong, W.H., Chou, N.N.S., 1987. In-soil stressstrain behavior of geotextile, Proceedings of
Geosynthetics 87, New Orleans, Louisiana, USA, Vol. 1, pp. 260265.
Thornton, J.S., 2001. Characterization of short and long term creep and relaxation properties of a
polypropylene geogrid. Proceedings of Geosynthetics Conference 2001, Portland, Oregon, USA,
pp. 835845.
Walters, D., Allen, T.M., Bathurst, R.J., 2002. Conversion of geosynthetic strain to load using
reinforcement stiffness. Geosynthetics International 9 (56), 483523.
Yuan, Z., Swan, R.H., Bachus, R.C., 1998. Soil connement effect on stressstrain properties of
geosynthetics. Proceedings of the Sixth International Conference on Geosynthetics, Atlanta, Georgia,
Vol. 2, pp. 523528.

You might also like