Geotextiles and Geomembranes 28 (2010) 503–513

Contents lists available at ScienceDirect

Geotextiles and Geomembranes

journal homepage: www.elsevier.com/locate/geotexmem

Ten year creep puncture study of HDPE geomembranes protected by

needle-punched nonwoven geotextiles
Robert M. Koerner a, *, Y.G. Hsuan a,1, George R. Koerner b, 2, David Gryger c, 3
a
Department of Civil, Architectural, and Environmental Engineering, Drexel University, Philadelphia, PA, USA
b
Geosynthetic Institute, Folsom, PA, USA
c
Gannett Fleming, Camp Hill, PA, USA

a r t i c l e i n f o a b s t r a c t

Article history: The need for a geotextile to be used for protection against geomembrane puncture by stones and gravel
Received 4 December 2008 has been recognized for many years. There are presently several methods available for selecting such
Received in revised form geotextiles. This paper, however, focuses on the ‘‘GRI-Method’’, which was originally based on short-term
9 October 2009
tests and was extended empirically for long-term performance. The reduction factor for creep behavior
Accepted 22 December 2009
Available online 18 February 2010
(RFCR) is of particular interest since its impact on the resulting geotextile design is the greatest.
The paper presents results of a 10-year long creep puncture study which is configured exactly the
same as was the original short-term testing program. The results indicate that the six z38 mm high
Keywords:
Geomembranes puncturing cones result in yield of the geomembrane at pressures of 34 and 52 kPa and one even had
Creep Puncture a small break. The six 12 mm high cones at pressures of 430 and 580 kPa also resulted in geomembrane
Long-Term Puncture yield but only by a nominal amount and there were no breaks.
Laboratory Tests As a consequence of these creep test results, the original table for creep reduction factors (RFCR) has
Cone Puncture been revised into more conservative values. In this regard, the originally published RFCR table should be
Simulated subgrade replaced accordingly.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction geomembrane are the most troublesome. Nosko and Touze-Folz

(2000) report that 71.2% of geomembrane damage found during
The disposal of solid waste materials in landfills generally electrical leak location audits are caused by stone puncture.
requires a liner system so as to prevent leachate from escaping into Exacerbating the situation is the current tendency to build
the adjacent surface or groundwater. This is almost always the case larger and higher landfills as well as to purposely add liquids
for hazardous and municipal solid waste, and is sometimes thereby creating bioreactor landfills, (Reinhart and Townsend,
required for inert solid waste as well. A recent worldwide survey of 1998). With higher and wetter landfills, along with increased field
regulations in 52 countries indicates that the majority of landfills operations activity involving greater compaction, a unit weight for
require a geomembrane as indicated in Table 1. municipal solid waste of 15 kN/m3 to even 20 kN/m2 is not unrea-
While there are many facets and considerations involved in sonable. With landfills as high as 100 m (so-called ‘‘megafills’’) the
a geomembrane’s design, they obviously should not be punctured. stress imposed on the liner system can readily mobilize stone
This would compromise their continuity and reduce their efficiency puncture of unprotected or poorly protected geomembranes.
as a barrier material. In this regard, stones either in the subgrade This situation brings about the need to be sure that stones do
beneath the geomembrane (even in a CCL, but certainly in a soil not impinge on the bottom of the geomembrane (requiring good
foundation) or from the granular leachate collection layer above the inspection practice) and the use of a cushioning material above the
geomembrane (generally being a geotextile) to protect it from the
coarser fraction of a granular leachate collection system. Regarding
* Corresponding author. Tel.: þ1 610 522 8440; fax: þ1 610 522 8441. the latter, needle-punched nonwoven geotextiles have been widely
E-mail addresses: [email protected] (R.M. Koerner), ghsuan@coe. used. Such geotextiles can be made from different polymers
drexel.edu (Y.G. Hsuan), [email protected] (G.R. Koerner), [email protected] (D. (previously polyester, but now commonly being polypropylene),
Gryger).
1 different mass per unit area, and different strengths.
Tel.: þ1 215 895 2785; fax: þ1 215 895 1363.
2
Tel.: þ1 610 522 8440; fax: þ1 610 522 8441. This paper reviews several design methods for geotextiles used
3
Tel.: þ1 717 763 7211; fax: þ1 717 763 1808. as protection materials.

0266-1144/$ – see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.geotexmem.2009.12.014
504 R.M. Koerner et al. / Geotextiles and Geomembranes 28 (2010) 503–513

Table 1 Shortly thereafter a European Working Group was formed to

Worldwide landfill liner requirements of 52 countries (Koerner and Koerner, 2008). develop a performance test for geotextile protection layers in order
Type of solid CCL alone GM alone GM/CCL Not designated to make comparative evaluations (Witte, 1990). As described by
waste composite Müller (2007), a chamber of at least 300 mm diameter is configured
Hazardous waste 1 (2%) 0 (0%) 34 (65%) 17 (33%) (from top to bottom) as follows:
(HSW)
Municipal waste 9 (17%) 3 (6%) 32 (62%) 8 (15%)
 steel plate
(ISW)
Inert Waste 2 (4%) 1 (2%) 21 (40%) 28 (54%)  sand layer
(ISW)  separation geotextile
Notes: CCL, compacted clay liner; GM, geomembrane.
 puncturing layer, e.g., gravel
 candidate protection geotextile
Focus is on one of these methods, the ‘‘GRI-Method’’, where  candidate geomembrane
a reduction factor for long-term creep (‘‘RFCR’’) of the geo-  0.5–1.0 mm soft metal plate
membrane against the protruding object is required. The original  support layer
values of RFCR suggested in 1997 were based on relatively short  steel plate
creep tests (maximum of 1000 h.) and then extrapolated into long
term values. Since the method was first proposed we are now in The desired load (generally, the maximum service load times
a position to report on a ten-year long series of laboratory experi- a 1.5) is applied with a pressurized piston on the upper steel plate
ments (87,000 h) to examine the relevancy of the originally sug- and maintained for 1000 h. Fig. 1 shows the apparatus just
gested creep reduction factors. described. The soft metal plate (usually a tin-lead alloy) retains the
It should also be mentioned at the outset that only one type and maximum deformation that the geomembrane has experienced
thickness of geomembrane is being evaluated. It is 1.5 mm thick during its pressurization. The test is formalized as European Stan-
smooth high density polyethylene (HDPE). dard BS-EN 13719:2002. Using one of several types of profil-
The reason being that HDPE is by far the most widely used ometers, strains in the identations formed by the impinging objects
polymer type for landfill liners and is actually prescribed by many are then determined. At this point, however, an acceptance crite-
countries. Regarding the thickness, minimum technology guidance rion must be legislated or decided upon. This same group suggested
for HDPE in the U.S. is 1.5 mm. In other countries, notably Europe, it a 3% strain maximum, but assumed that construction represents
varies from 1.5 to 2.5 mm. That said, the minimum (hence, most 2.75% of this value. The remaining value is then the maximum
sensitive) thickness was chosen for all of the testing in the original strain allowed during the test and if any of the measured strains
testing program in and the creep puncture experiments to follow. exceed the 0.25% value, the protection layer is not acceptable. By
In addition only one type and mass per unit area of nonwoven trial and error an acceptable layer, e.g., a thicker geotextile, is
needle punched geotextile was used as a control, or reference, eventually identified. It is important to note that the working
material. The variables were cone height and applied hydrostatic group’s concern regarding these relatively low allowable strain
pressure. values (HDPE yield is about 12% strain) is deformation leading to
long-term stress cracking issues and not the specific concern over
the immediate puncture of the geomembrane. Brummermann et al.
2. Geotextile protection design methods
(1994) and Seeger and Müller (1996) give additional detail.
The above discussion on 3% maximum strain so as to prevent
The first paper to bring attention to the need for geomembrane
long-term stress cracking can be challenged (in the author’s
protection was by Giroud (1982) who also suggested the use of
opinion) from two perspectives. The first is that the stress crack
needle-punched nonwoven geotextiles. The paper, however, does
sensitive resins used in the past have been virtually eliminated
not give a design method per se, rather it discusses the various
from the geomembrane market by use of rigorous generic
functions that such a geotextile can provide.

Fig. 1. Test vessel used in Europe to evaluate geomembrane protection using geotextiles, Müller (2007).
 R.M. Koerner et al. / Geotextiles and Geomembranes 28 (2010) 503–513 505

Fig. 2. Test vessels used to evaluate geomembrane puncture in the GRI-Method of geotextile protection design. (a) Sketches of truncated cones, their arrangement, and test vessel,
(Müller, 2007). (b) Single pressure vessel for long-term testing. (c) Two of our identical pressure vessels with readout boxes.

specifications, e.g., GRI-GM13, which utilizes the notched constant are used in the testing vessel. A lead plate is placed beneath the
load test (ASTM D5397) to challenge the resin’s stress crack geomembrane from which strain is measured. Again, a maximum
susceptibility in an accelerated manner. The cited specification then 0.25% strain in the geomembrane is suggested for geotextile
sets the minimum allowable value. Stress crack failures have simply acceptance. In performing such tests, Shercliff and his colleagues
not been observed in the field for HDPE geomembranes produced have found that both mass per unit area and tensile strength
since the year 2000 which follow the GRI-GM13 specification. The (represented by the CBR value per ISO 12236) of the protection
second is that the HDPE geomembranes currently being produced geotextile are important. As will be seen in the next described
are actually made from medium density polyethylene (MDPE) method, results of the GRI Method indicate that mass per unit area
resin. The resin density is between 0.935 and0.938 g/cc and only by is the controlling parameter. This is an important issue and specific
addition of carbon black and additives does the formulation reach papers have been written on this very subject, e.g., Jones et al.
0.941 g/cc, which is the lower limit for HDPE according to ASTM (2000).
D883. That said, as the resin density of polyethylene decreases, the Variations of a geotextile-only protection material are also
stress crack resistance greatly increases. Thus, if stress cracking is available in the open literature. A layer of sand is obviously available
a greatly reduced issue for present-day HDPE geomembranes, a 3% to provide puncture protection, as is a sand layer contained within
maximum strain criteria for puncture seems to be excessively geotextiles, i.e., sand filled cushions. Such tests have been reported
conservative. Incidentally, Peggs et al. (2005) use an allowable by Tongon et al. (2000) and more recently by Dickinson and
strain of 6–8%. Brackman (2008). The latter also has proposed and evaluated
A somewhat simplified version of the previously described layered geotextiles consisting of textile components needle-
performance test has been developed by Shercliff (1996). The punched together having different physical and mechanical prop-
proposed stone, geotextile, and geomembrane under consideration erties. In addition, geomembrane protection using geosynthetic
506 R.M. Koerner et al. / Geotextiles and Geomembranes 28 (2010) 503–513

Fig. 3. Selected puncture results using the GRI testing vessel with truncated cones; see Fig. 2 for details (Narejo et al., 1997). (a) Behavior of different types of geomembranes.
(b) Behavior of various stone types compared to cones. (c) Behavior of different geotextile mass per unit area. (d) Behavior of various types of nonwoven geotextiles.

clay liners (GCLs) has also been evaluated (Heerten, 1994; Narejo and reclaimed resins each of which were either staple or contin-
et al., 2002, 2007). While options to thick needle-punched uous filament) resulted in an empirical equation which correlated
nonwoven geotextiles are available for geomembrane protection to the data as shown in Fig. 3d. This important graph clearly indi-
materials, such nonwoven geotextiles are, by far, the most widely cates that the geotextile’s mass per unit area is the controlling
used. parameter. The regression analysis results in R2 ¼ 0.973. This is not
Recognizing the variation of geomembrane puncturing objects, to say that geotextile strength is not important, it only suggests that
e.g., various stone sizes and shapes, experiments with what was strength, per se, is of secondary importance to mass per unit area.
considered a ‘‘worst-case scenario’’ puncturing object were per- While many theses and technical papers have been written
formed and reported accordingly; see Hullings and Koerner (1991). using this experimental setup, a series of three papers captures the
The concept was not based on a stress crack scenario of 3% strain entire program (Wilson-Fahmy et al., 1996; Narejo et al., 1996;
but on the object actually puncturing through the geomembrane. Koerner et al., 1996).
What eventually resulted was a 250 mm spaced arrangement of
three truncated cones made from rigid acrylic plastic in a triangular
array and placed in a test vessel, see Fig. 2. Sand coverage of Table 2
Modification factors and reduction factors for geotextile protection material design
portions of the base of the cones allows for a specific protrusion using Eq. (2), i.e., the ‘‘GRI-Method’’.
height against which the protection geotextile is placed and then
the designated geomembrane. Hydrostatic pressure is applied to (a) Modification factors (all  1.0)

the geomembrane until a puncture failure of the geomembrane is MFs MFPD MFA
indicated. Water passing through this puncture hole is electroni- Angular 1.0 Isolated 1.0 Hydrostatic 1.0
cally sensed by the plastic cones and registered on readout boxes. Subrounded 0.5 Dense, 38 mm 0.83 Geostatic, shallow 0.75
A number of parametric evaluations have been developed using Rounded 0.25 Dense, 25 mm 0.67 Geostatic, mod. 0.50
Dense, 12 mm 0.50 Geostatic, deep 0.25
this setup. Fig. 3a shows the response of four geomembrane types
in which HDPE is seen to be the most sensitive to out-of-plane (b) Reduction factors (all  1.0)

puncture of this type. Fig. 3b shows results of 1.5 mm smooth HDPE RFCBD RFCR
geomembranes being punctured by three types of stones as Mass per unit area (gm/m2) Protrusion height
compared to the truncated cones used for the entirety of this (mm)
project. The puncture correlation of angular stones to the truncated 38 25 12
stones is readily seen. Fig. 3c shows a 1.5 mm smooth HDPE geo-
Mild leachate 1.1 Geomembrane alone N/R N/R N/R
membrane being protected by various mass per unit area geo- Moderate leachate 1.3 270 N/R N/R >1.5
textiles indicating that higher mass per unit area of the same Harsh leachate 1.5 550 N/R 1.5 1.3
geotextile type does indeed provide greater protection. Using 1100 1.3 1.2 1.1
1.5 mm thick smooth HDPE geomembranes with a large variation of >1100 y1.2 y1.1 y1.0

needle-punched nonwoven geotextiles (polyester, polypropylene Abbreviation: N/R, not recommended.

 R.M. Koerner et al. / Geotextiles and Geomembranes 28 (2010) 503–513 507

Table 3   
Properties of HDPE geomembranes used in these test M 1
pallow ¼ 50 0:00045 2
Property Test method Units Value H MFS  MFPD  MFA
 
Thickness, (min. ave.) D5199 mm 1.58 1
 lowest individual of 10 values mm 1.53
 (2)
RFCBD  RFCR
Density (min.) D1505/D792 g/cc 0.949
 Tensile properties (min. ave.) Type IV
where pallow is the allowable pressure so as to prevent puncture
 yield stress D6693 kN/m 32.7
 break stress D6693 kN/m 53.2 (kPa); M, geotextile mass per unit area (g/m2); H, protrusion height
 yield elongation D6693 % 17.0 (m); MFS, modification factor for protrusion shape; MFPD, modifi-
 break elongation D6693 % 800 cation factor for packing density; MFA, modification factor for
Tear resistance (min. ave.) D1004 N 250 arching in solid materials; RFCBD, reduction factor for long-term
Puncture resistance (min. ave.) D4833 N 668
chemical/biological degradation; and RFCR, reduction factor for
Stress crack resistance D5397 (app.) h 209
Carbon black content (range) D1603 % 2.3 long-term creep.
Carbon black dispersion D5596 Cat. 1 The design situation can be approached from a given mass per
Oxidative Induction Time (OIT) (min. ave.)
unit area geotextile to determine the unknown FS-value, or from
(a) Standard OIT D3895 min. 94 a given FS-value to determine the unknown geotextile mass per
or unit area value. In the latter case with ‘‘M’’ being unknown,
(b) High pressure OIT D5885 min. 224 a reasonable value of FSCR is assumed for the calculation and then

Oven aging at 85 C D5721 this assumed value must be checked for validity in Table 2b.
(a) Standard OIT (min. ave.) – D3895 % 50 Koerner (2005) gives numeric examples, and Valero and Austin
% retained after 90 days (1999) present design charts for the many variables contained in
or D5885 % 95
(b) High pressure OIT (min. ave) –
the design equation. It might be noted that this method is the only
% retained after 90 days design method that allows for direct selection of a geotextile
protection material without the need for large scale trial-and-error
UV resistance D7238
(a) Standard OIT (min. ave.) D3895 N.R experimental testing.
or In Eq. (2b) the two terms ‘‘RFCBD’’ and ‘‘RFCR’’ are intended to
(b) High pressure OIT (min. ave.) – D5885 % 59 extend the short term test results into a simulated long term
% retained after 1600 h performance behavior. Under typical landfill conditions of
moderate temperature and typical leachate the chemical and bio-
logical degradation term, RFCBD is comparatively small; see Koerner
and Koerner (2005) and Tisinger (1996), respectively. The term
The resulting design formula uses a conventional factor of safety RFCR, however, is not small and in many cases ‘‘not recommended’’
as follows: is suggested. Due to its importance in the overall design, a series of
long term creep tests using this same methodology, i.e,. truncated
FS ¼ pallow =pact (1)
cones, has been undertaken for the past ten years. This paper
where FS is the factor of safety (against geomembrane puncture); presents these new results which will attempt to explain many of
pact, actual pressure due to the applied normal stress, e.g., contents the issues raised in this review, particularly the validity of the RFCR-
of a landfill or surface impoundment; and pallow, allowable values of Table 2b.
pressure using different types of geotextiles and site-specific
conditions.
An empirical relationship for ‘‘pallow’’ has been generated. It is 3. Results after 10-years of creep puncture
given as Eq. (2) and its correlation to the experimental data was
shown in Fig. 3d. Its use, however, so as to accomodate site-specific This section of the paper presents the resulting behavior and
conditions requires the use of modification factors and reduction deformations of the test specimen geomembranes after 10-years of
factors as given in Table 2. Note that in Table 2, all MF values 1.0 creep puncture testing. It should be noted that the vessels were
and all RF values 1.0. never opened until after the ten-year period elapsed.

Fig. 4. The hard acrylic truncated cone protrusions producing the puncturing action. (a) Array of three cones rising above sand level. (b) Detail of an individual cone.
508 R.M. Koerner et al. / Geotextiles and Geomembranes 28 (2010) 503–513

Table 4
Results of truncated cone 10-year creep puncture tests (600 g/m2 NP-NW-PET geotextile protecting a 1.5 mm smooth HDPE geomembrane).

Vessel no. Vertical cone heightsa Applied puncture Final description of geomembranee
stressc

Initial (mm) Final (mm)b (kpa) (%)d Visual Thickness (mm)f

1 12 12.1 430 23 3-Subtile yields – no breaks 1.16 (70% of original)
2 12 11.5 580 28 3-Subtile yields – no breaks 8.80 (48% of original)
3 38 29.7 52 15 3-Pronounced yields; one break 0.32 (20% of original)
4 38 31.1 34 10 3-Pronounced yields; no breaks 0.34 (22% of original)
a
Each test apparatus had three identical truncated cones beneath the geomembrane; see Fig. 4.
b
The cone heights changed during, or after, pressurization due to movement of the initially placed sand layers, i.e., the sand was essentially pushed up around the stationary
cones somewhat reducing their initial heights (except for Vessel #1).
c
Hydrostatic stress applied to geomembrane; beneath which is the geotextile and then the three puncturing cones.
d
Percentage of short-term failure stress using Eq. (2) with all MF and RF-values of ‘‘one’’ for the calculations.
e
Refers to geomembrane response after 10-years of hydrostatic stress.
f
Results are average minimum geomembrane thickness above the cone tips.

3.1. Details of the test setups 200 g/m2 needle-punched nonwoven (continuous filament)
polyester geotextiles. They will be collectively referred to as
There are four identical test vessels used in this study, each 600 g/m2 protection materials.
containing three identical truncated cones shaped and config- The differences in the four test vessels are the heights of the
ured as shown in Fig. 2. In all cases the geomembranes being truncated cones causing the puncture to occur and the applied
evaluated are 1.5 mm nominal thickness smooth HDPE which hydrostatic pressures.
conforms to the original GRI-GM13 specification insofar as Regarding the cone heights, sand is placed and compacted in the
physical, mechanical, and endurance properties are concerned; vessels leaving a protrusion height rising above the sand level; see
see Table 3. To be noted in the geomembrane’s endurance Fig. 4. As placed, two vessels had initial cone heights of 12 mm and
properties of Table 3 is that the stress crack resistance of 209 h. the other two had initial cone heights of 38 mm. It was recognized
met the original version of GRI-GM13 (which was 200 h.) but that 38 mm was unacceptable, e.g., in Table 2b, ‘‘not recommended’’
does not meet the current version (which has been increased to is listed, but this limit was in need of being verified. Regarding the
300 h). Thus the HDPE geomembrane being currently used is 12 mm cone heights, Table 2b indicates that it should be acceptable
better in its stress crack resistance than the one used in this providing a RFCR ¼ 1.3 is used in the design procedure of Eq. (2).
study’s testing. In this same regard, the very high density of This, of course, had to be verified as well.
0.949 g/cc is significantly greater than the current specification Regarding the applied hydrostatic pressure, there was consid-
value which is 0.941 g/cc. Also common to all four setups is the erable uncertainty. The design procedure using Eq. (1) does not
geotextile cushioning materials. They consisted of three layers of address a maximum pressure. As a result high hydrostatic pressures

Fig. 5. Photographs of geomembrane (upper) and geotextile (lower) deformations in Vessel #1 at x12 mm cone heights and 23% short term failure stress.
 R.M. Koerner et al. / Geotextiles and Geomembranes 28 (2010) 503–513 509

Fig. 6. Photographs of geomembrane (upper) and geotextile (lower) deformations in Vessel #2 at x12 mm cone heights and 28% short term failure stress.

were used for the two sets of 12 mm cone heights (430 and that occurs within the solid waste as deformation occurs. The
580 kPa) and low hydrostatic pressures were used for the two sets ‘‘MFA’’ term in Table 2a attempts to take such geostatic stresses into
of 38 mm cone heights (52 and 34 kPa). It is worth mentioning that account. It should also be mentioned that these applied hydrostatic
hydrostatic pressure represents surface impoundment (liquid) pressures represent 10–28% of the short-term failure stresses as
stresses but overestimates solid waste stresses due to the arching indicated in Table 4.

Fig. 7. Photographs of geomembrane (upper) and geotextile (lower) deformation in Vessel #3 at x38 mm cone heights and 15% short term failure stress. (Note crack in geo-
membrane and wet spot on underlying geotextile).
510 R.M. Koerner et al. / Geotextiles and Geomembranes 28 (2010) 503–513

Fig. 8. Photographs of geomembrane (upper) and geotextile (lower) deformation in Vessel #4 and x38 mm cone heights and 10% short term failure stress.

3.2. Maintenance of the test setup break that occurred during the entire test program. Also seen is that
the protection geotextile was highly deformed in a radial pattern
The four test vessels were located in an isolated room along with around the 38 mm cone heights. This was not observed in the
other ongoing experiments where the temperature was maintained 12 mm cone height tests. The summary description of these visual
at 22  C  2  C. The pressure to which the geomembranes within findings is given in Table 4.
the vessels were subjected was held constant throughout the ten-
year test period. The truncated cones shown in Fig. 4 are embedded 3.4. Measured cone heights
with two copper electrodes emerging on their flat faced surfaces.
They are filled level with the cone surface so as not to have any The cone heights were measured after removal of the geo-
protrusion themselves, i.e., only the acrylic cone tips cause the membranes and geotextiles and found to be the same or somewhat
actual puncturing action. These copper wires are connected to less than originally placed; see Table 4 for initial versus final cone
a junction box (see Fig. 2c) and when a voltage is registered a light
indicator turns on. When a puncture occurs a circuit is completed
by the water above the geomembrane leaking through a hole
caused by the truncated cone, i.e., an electrical circuit is completed.
This only occurred in one of the twelve cones used during the 10-
year test time. It occurred near the end of the time frame indicated.
Otherwise, the test vessels were essentially free of maintenance
and were disturbed only once. That was when they were decom-
missioned after 10-years of service.

3.3. Visual test results

Figs. 5–8 show photographs of the geomembranes and geo-

textiles associated with the four setups listed as Vessels 1–4 in
Table 4 after they were dismantled. In all cases both the geo-
membrane and the geotextile were visibly deformed. The outline of
the protruding cone tip is very apparent even in Figs. 5 and 6 for the
12 mm cone heights. An extremely small yield area is seen in the
upper right photographs of each figure. In contrast, Figs. 7 and 8
show the deformations caused by the 38 mm cone heights. Here
the deformations are seen to be much greater and a more
pronounced yielding has occurred in the geomembrane at the cone
tips. Additionally, the upper right photograph of Fig. 7 shows that
a small break occurred within the yielded area. This is the only Fig. 9. Resulting crescent moon shape of yield region in geomembrane test specimens.
 R.M. Koerner et al. / Geotextiles and Geomembranes 28 (2010) 503–513 511

Table 5
Measured values (in mm) associated with the designations of Fig. 9.

Vessel no.

1 2 3 4

Cone no. 1.1 1.2 1.3 2.1 2.2 2.3 3.1 3.2 3.3 4.1 4.2 4.3
A (Z-axis) 9.70 9.53 10.29 12.12 11.79 11.66 15.85 16.31 16.81 17.58 18.19 16.92
B (Z-axis) 9.40 9.37 10.13 11.20 11.48 11.00 13.69 13.44 13.92 15.80 16.10 16.43
C (Z-axis) 8.15 8.05 8.86 10.36 10.11 9.98 11.63 11.25 11.63 14.10 14.66 14.25
D (Z-axis) 8.51 8.31 8.86 10.19 10.52 9.78 11.71 11.68 12.40 14.25 14.35 15.01
E (X-axis) 13.15 14.60 14.67 16.03 14.97 16.74 20.06 20.19 20.62 19.11 18.65 19.06
F (X–Z axis) 2.03 2.22 2.01 2.41 2.20 2.48 4.70 4.72 5.34 3.65 3.73 3.00
G 1.07 1.13 1.05 0.65 0.81 0.72 0.47 0.47 0.44 0.36 0.46 0.94
H 0.60 0.44 0.35 0.60 0.33 0.42
I 1.58 3.63 4.52 4.52 2.56 2.66
J 1.07 1.16 1.19 0.69 0.85 0.76 0.57 0.44 0.47 0.47 0.47 0.94
K 0.59 0.41 0.36 0.54 0.35 0.40
L 1.83 3.41 4.62 4.55 2.57 2.33

heights. This reduction in cone height is believed to be caused by comparing them to the calculated values, a percentage of short-
a shifting of sand from the central region of the test setup up term pressures can be obtained. They are 23, 28, 15 and 10%,
against the stationary cones during the initial vessel pressurization. respectively and are listed in Table 4.
The sand was a cohesionless well graded concrete sand. The final
cone heights will be used in the subsequent calculations. 4.2. Calculated strains based on the length of the yield region

All geomembranes adjacent to the twelve cones were dimen-

4. Data analysis
sionally measured in their yield regions and then were cross
sectioned and measured for thickness using a digital caliper. Fig. 9
This section presents calculations using the experimental data
identifies the data given in Table 5. As shown in Fig. 9 and the
which when added to the visual identifications will result in the
previous photographs, the yield region was always in the shape of
summary and conclusion section to follow.
a crescent moon resulting from the truncated cone configuration.
The dimension ‘‘E’’ was considered as being a possible calculation
4.1. Calculation of percent short term stresses target and while these final deformation values are properly
ordered, i.e., the 12 mm cone height values are less than the 38 mm
The allowable pressure is determined using Eq. (2) and cone height values, the initial value is zero which makes the
substituting all MF-values with ‘‘ones’’ and all RF-values also with calculation of strain not possible.
‘‘ones’’, since these experiments adopt the exact configurations
from which the equation was originally generated. The equation 4.3. Calculated strains based on the width of the yield region
which follows is that obtained directly from the short term test data
which had the correlation shown in Fig. 3d. The dimension ‘‘F’’ was also considered as being a possible
calculation target and again the values are properly ordered but an
M initial zero value makes the strain calculation impossible.
pallow ¼ 50 þ 0:00045 2 (3)
H
It is of interest to note that in Eq. (3), the first term of 50 kPa 4.4. Calculated strains based on thickness reduction
represents the puncture resistance of the 1.5 mm nominal thick-
ness HDPE geomembrane by itself. This very low value indicates The left and right cross sections shown in Fig. 9 were used to
that the geomembrane does indeed require a geotextile (or other) obtain a thickness strain. This is readily accomplished since the
for protection and it is the second term in the equation that original thickness is easily measured and these are the initial values
becomes the major contributor to this protection. Using M ¼ 600 g/ from which to make thickness strain calculations. Note that this is
m2 for all cases and the final cone heights as given in Table 4, the not a tensile strain in the conventional sense of describing geo-
predicted allowable short-term puncture pressures (at failure) are membrane strains. In the last column of Table 4 average values are
1894, 2092, 356, and 329 kPa, respectively. By taking the actual used in the calculations. The resulting strains indicate that the
hydrostatic applied pressures in the four test vessels and 38 mm cones resulted in the highest strains with the 80% value for
Vessel No. 3 being the one in which one of the cones had an actual
break, recall Fig. 7. The numeric values themselves, however, are
Table 6 not comparable to the usual tensile strains and, as such, are
Revised values for ‘‘RFCR’’ to be used in Eq. (2) for geotextile protection materials
inconclusive.
Mass per unit area (g/m2) ‘‘RFCR’’-values

Protrusion height (mm) 4.5. Discussion

38 25 12
The fact that yield occurred in the deformed shape of the HDPE
Geomembrane alone N/R N/R N/R
270 N/R N/R N/R geomembranes under evaluation was fortunate from a measure-
550 N/R N/R >1.5 ment perspective. Once yield occurs, the deformation retains its
1100 N/R 1.5 1.3 shape and stress relaxation becomes a minor issue. This is not the
>1100 1.3 1.2 1.1 case for the preyield deformation in which lead plates are required
Abbreviation: N/R, not recommended. to capture the deformations after pressure is relaxed. Equations for
512 R.M. Koerner et al. / Geotextiles and Geomembranes 28 (2010) 503–513

tensile strain stemming from uniform geometric shapes as pre- RFCR ¼ 1.3. Since all six of these cases resulted in geomembrane
sented in Seeger and Müller (2000), however, are not possible to yield (albeit small yields in comparison to the higher cone heights),
use for this crescent moon shape after yield occurs. Thus the only the Table 2b values must be changed and made more conservative
strain value that could be calculated came from a decrease in in their design guidance.
thickness from the original value. While thickness strain is quite To be noted in Table 2 for RFCR, a not recommended (N/R)
unconventional, the data appears consistent between the low and comment exists for the various protrusion heights in descending
high cone height behavior. Here the higher cone heights resulted in order as the cone heights decrease and the protection geotextile
the greater thickness strains. The highest value of 80% was in the mass increases. By virtue of these test results, the N/R comments
particular set of cones in which a break actually occurred in one of must be extended. Thus, our conclusion as a result of this creep testing
them. program is to replace the existing RFCR-values with the more restrictive
values in Table 6 following:
5. Summary and conclusions Lastly, the entry of ‘‘>1.5’’ for a 12 mm cone height associated
with a 550 g/m2 geotextile is felt to be appropriate considering the
The need for geomembrane protection against puncture by following items.
objects such as stones and gravel has been apparent for many years.
Commonly used for this purpose are relatively thick needle-  The geotextiles used at present are made from polypropylene
punched nonwoven geotextiles (although there are other possible fibers versus the tested geotextiles which were made from
strategies). In essence, such geotextiles provide a cushion in polyester fibers. Since the specific gravity of PP is 0.91and that
blunting the inherent aggressiveness of the puncturing object of PET is between 1.22 and 1.38, one has from 25% to 34% more
against the geomembrane. Even further, geomembranes used as filaments in an equivalent mass per unit area geotextile with
liner materials beneath solid waste landfills are commonplace and polypropylene fibers. It is felt that this provides for consider-
for large landfills the normal stresses on such geomembranes are ably greater protection capability although it has not yet been
very high, i.e., the puncture situation is greatly exacerbated. quantified. Work is ongoing in this regard.
The type of geomembrane is also an issue. By virtue of its good  The area of yield for the six w12 mm cone heights was
chemical resistance and long anticipated lifetime, HDPE geo- extremely small and the thicknesses of the remaining geo-
membranes are routinely used to line solid waste landfills. Many membrane was such that considerable deformation could still
countries even specifically specify this type of geomembrane. That be sustained before break is even close to occurring.
said, HDPE is (other than scrim reinforced geomembranes) the  The greater than symbol in the ‘‘>1.5’’ recommendation is
most sensitive geomembrane type to out-of-plane deformation, as precisely for additional conservatism and safety and if
is the situation arising from a puncturing object located above or a designer wishes to be more conservative than the new rec-
below the geomembrane. ommended table suggests he/she is free to do so.
As a result of the above issues, several approaches toward
selecting a proper geotextile protection material are in the litera-
Acknowledgement
ture and have been reviewed herein. This paper, however, has
focused on the ‘‘GRI-Method’’ which was developed based on the
This project was funded by the Geosynthetic Institute’s research
result of a large short-term testing project that has been published
and development unit, the Geosynthetic Research Institute.
and widely used for at least twelve-years. Needed, however, is the
Members and Associate Members can be found on the institute’s
projection of short-term testing results into long-term behavior.
website at www.geosynthetic-institute.org.
This has been done in the past using empirical tables for both
degradation (RFCBD) and creep (RFCR) reduction factors; recall Eq.
(2). The degradation by chemical and biological agents is the lesser References
of the two reduction factors and, as a result, this present study is
focused entirely on the validity of the RFCR-values. These values are ASTM D883, Standard Terminology Relating to Plastics.
ASTM D5397, Standard Test Method for Evaluation of Stress Crack Resistance of
given in Table 2b. To verify or refute the given values, a 10-year
Polyolefin Geomembranes Using Notched Constant Tensile Load Testing.
creep study of HDPE geomembranes and their associated geotextile Brummermann, K., Blumel, W., Stoewahse, C., 1994. Protection layers for geo-
protection materials against puncture has been concluded and this membranes: effectiveness and testing procedures. In: Proceedings of the Fifth
paper presents the results. IGS Conference, Singapore, pp. 1003–1006.
Bs EN 13719:2002, Geotextile and geotextile-related products. Determination of the
The same type of pressure vessel and truncated puncturing long term protection efficiency of geotextiles in contact with geosynthetic
cones as used in the short-term tests were used for these long-term barriers, ISBN: 0 580 40575-3.
tests. In all cases, 1.5 mm thick smooth HDPE geomembranes were Dickinson, S., Brackman, R.W.I., 2008. Assessment of alternative protection layers
for a geomembrane-geosynthetic clay liner (GM-GCL) composite linter. Cana-
used and protected by 600 g/m2 needle-punched nonwoven PET dian Geotechnical Journal 45, 1594–1610.
geotextiles. Three truncated cones were used in each of four pres- Giroud, J.P., 1982. Design of geotextile associated with geomembranes. In:
sure vessels, the differences being the protruding cone heights (six Proceedings of the Second IGS Conference, Las Vegas, USA, pp. 27–30.
GRI-GM13, Standard Specification for Test Methods, Test Properties, and Testing
at z12 mm and six at z38 mm) and applied hydrostatic pressures Frequency for High Density Polyethylene (HDPE) Smooth and Textured
(varying from 34 to 580 kPa). Geomembranes.
After 10-years of pressurization the vessels were dismantled Heerten, G., 1994. Geotextile and/or GCL protection systems for geomembranes.
GRI-7 Conference on Geosynthetic Liner Systems. IFAI Publ., pp 155–167.
and it was found that all six of the high cone heights (w38 mm) had
Hullings, D., Koerner, R.M., 1991. Puncture resistance of geomembranes using
pronounced yield zones in the geomembranes and one of the six a truncated cone test. Geosynthetics 1991 Conference Proceedings. IFAI, Rose-
had a small break within its yield zone. Clearly, such high cone ville, MN, pp. 273–285.
Jones, D.R.V., Shercliff, D.A., Dixon, N., 2000. Difficulties associated with the speci-
heights with this type of geotextile are unacceptable. The entry in
fication of protection geotextiles using only unit weight. In: Proceedings of the
Table 2b in this regard mentions ‘‘not recommended’’ and this Second European Geosynthetics Conference (EuroGeo 2), Italy, pp. 551–556.
comment is hereby substantiated. Koerner, G.R., Koerner, R.M., 2005. In-situ temperature monitoring of geo-
However, the creep test results at low cone heights (z12 mm) membranes. Proceedings of GRI-18 Conference in GeoFrontiers. ASCE, Austin,
TX (on CD).
provide a different conclusion. Table 2 indicates that a 12 mm cone Koerner, R.M., 2005. Designing with Geosynthetics. Pearson-Prentice Hall Publ. Co.,
height with a 550 g/m2 protection geotextile is acceptable with an Upper Saddle River, NJ, 796 pp.
 R.M. Koerner et al. / Geotextiles and Geomembranes 28 (2010) 503–513 513

Koerner, R.M., Koerner, J.R., 2008. GRI’s second worldwide survey of solid waste Reinhart, D.R., Townsend, T.G., 1998. Landfill Bioreactor Design and Operation. Lewis
landfill liner and cover systems. GRI Report No. 34. Geosynthetic Institute, Publishers, New York, 189 pp.
Folsom, PA, USA, 137 pp. Seeger, S., Müller, W., 1996. Requirements and testing of protection layer systems
Koerner, R.M., Wilson-Fahmy, R.R., Narejo, D., 1996. Puncture protection of geo- for geomembranes. In: Proceedings of GRI-9, Geosynthetics in Infrastructure
membranes; part III examples. Geosynthetics International 3 (5), 655–675. IFAI, Enhancement and Remediation. Geosynthetic Information Institute, Phila-
St. Paul, MN. delphia, PA, pp. 185–194 (Jour, Geotextiles and Geomembranes, 14 (7/8) July/
Müller, W., 2007. HDPE Geomembranes in Geotechnics. Springer-Verlag, Berlin, 485 pp. August 1996, pp. 365–376).
Narejo, D., Corcoran, G., Zunker, R., 2002. An evaluation of geosynthetic clay liners Shercliff, D.A., 1996. Optimization and testing of liner protection geotextiles used in
to minimize geomembrane leakage caused by protrusions in subgrades and landfills. In: DeGrott, Den Hoedt, Termaat (Eds.), Proceedings of Geosynthetics:
compacted clay liners. In: International Symposium on Clay Geosynthetic Applications, Design and Construction. Balkema, Rotterdam, pp. 823–828.
Barriers (IS Nuremberg 2002), Nuremberg, Germany, pp. 61–72. Tisinger, L., 1996. Chemical compatibility testing: the state-of-the-practice.
Narejo, D., Kavazanjian, E., Erickson, R., January 16–19 2007. Maximum protrusion Geotechnical Fabrics Report 14 (8), 28–29. IFAI, Roseville, MN.
size under geomembrane/GCL composite liners. In: Geosynthetics 2007 Tongon, A.R., Rowe, R.K., Moore, I.D., December 2000. Geomembrane strain
Conference Proceedings, Washington, DC (on CD). observed in large-scale testing of protection layers. Journal of Geotechnical and
Narejo, D., Koerner, R.M., Wilson-Fahmy, R.F., 1996. Puncture protection of geo- Geoenvironmental Engineering 126 (12), 1194–1208.
membranes; part II experimental. Geosynthetics International 3 (5), 629–653. Valero, S.N., Austin, D.N., 1999. Simplified design charts for geomembrane cushions.
IFAI, St. Paul, MN. In: Proceedings of Geosynthetics 1999. IFAI, Roseville, MN, pp. 357–372.
Nosko, V., Touze-Folz, N., 2000. Geomembrane liner failure: modeling of its influ- Wilson-Fahmy, R.F., Narejo, D., Koerner, R.M., 1996. Puncture protection of geo-
ence on contaminant transfer. In: Proceedings of Second European Geo- membranes; part I theory. Geosynthetics International 3 (5), 605–628. IFAI, St.
synthetics Conference, Bologna, Italy, pp. 557–560. Paul, MN.
Peggs, I.D., Schmucker, B., Carey, P., 2005. Assessment of maximum allowable Witte, R., 1990. Kurzebericht: weiterentwicklung des schutzwirksamkeits-nach-
strains in polyethylene and polypropylene geomembranes. In: Proceedings of weises für geotextile schutzsysteme in der deponiebasisabdichtung. Müll und
GeoFrontiers Conference. ASCE GeoInstitute (on CD). Abfall 22, 788–789.