STRESS CRACKING BEHAVIOR OF
HDPE GEOMEMBRANES AND ITS PREVENTION
by
Robert M. Koerner, Yick G. Hsuan, and Arthur E. Lord, Jr.
Geosynthetic Research Institute
Drexel University
Philadelphia, PA 19104
Cooperative Agreement No. CR-815692
Project Officer
David A. Carson
Municipal Solid Waste and Residuals Management Branch
Waste Minimization, Destruction and Disposal Research Division
Cincinnati, OH 45268
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OH 45268
-------
DISCLAIMER
The information in this document has been funded wholly by the United States
Environmental Protection Agency under Cooperative Agreement No. CR-815692 to the
Geosynthetic Research Institute of Drexel University in Philadelphia, Pennsylvania. It has been
subjected to the Agency's peer and administrative review, and it has been approved for publication
as an EPA document. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
11
-------
FOREWORD
Today's rapidly developing and changing technologies and industrial products and practices
frequently carry with them the increased generation of materials that if, improperly dealt with, can
threaten both public health and the environment. The U.S. Environmental Protection Agency is
charged by Congress with protecting the Nation's land, air, and water resources. Under a mandate
of national environmental laws, the agency strives to formulate and implement actions leading to a
compatible balance between human activities and the ability of natural systems to support and
nurture life. These laws direct the EPA to perform research to define our environmental problems,
measure the impacts, and search for solutions.
The Risk Reduction Engineering Laboratory is responsible for planning, implementing, and
managing research, development, and demonstration programs to provide an authoritative,
defensible engineering basis in support of the policies, programs, and regulations of the EPA with
respect to drinking water, wastewater, pesticides, toxic substances, solid and hazardous wastes,
and Superfund-related activities.
This publication is one of the products of that research and provides a vital communication
link between the researcher and the user community. This document focuses on the development
of new test methods to evaluate the stress crack resistance of HDPE geomembranes.
Recommendations are forwarded which should be carefully considered and hopefully adopted by
both resin suppliers and sheet manufacturers providing liner materials to the waste containment
industry.
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory
111
-------
ABSTRACT
Geomembranes made from high density polyethylene (HOPE) have a high percent crystallinity
and are therefore of concern with regard to stress cracking. A review of the literature plus our field
exhuming of various sites-of-opportunity gave rise to twenty-five (25) situations where stress
cracking of HDPE geomembranes has occurred since the introduction of this liner material in 1980.
The stress cracks varied from very small lengths of a few centimeters to large shattering types of
failure. All failures have occurred on exposed geomembranes where ultraviolet light exposure, high
temperature oxidation and low temperature contraction are continuously ongoing processes.
ThecurrentiyusedASTMD1693'M3entsMp"testmethodforq^
contains some generally acknowledged limitations. Therefore an alternative (and more challenging)
test method was investigated. The result of this search was the development of a notched constant
tensile load (NCTL) test. In this test dumbbell shaped test specimens are centrally notched and
stressed to a prescribed percentage of their yield stress. They are then incubated in a wetting agent
at 50°C constant temperature and their failure times recorded. Evaluation of a series of test specimens
at different applied stresses produces a ductile-to-brittle response curve where the onset of brittle
behavior is termed the transition point. The test method and procedure was also verified for
reproducibility and was evaluated in a series of interlaboratory tests with satisfactory results.
Utilization of the above described NCTL test on 18 commercially available sheet materials
and 7 field exhumed geomembranes led to the recommendation that for an acceptable stress crack
resistant HDPE geomembrane the test results should indicate a transition time of greater than 100
hours.
The above recommendations have been shown to be clearly more challenging than those
resulting from the bent strip test. It is recommended to discontinue sole reliance on the bent strip test
for qualification of HDPE geomembranes in favor of the proposed NCTL test.
In consideration of the length of testing time required to completely develop the full NCTL
test response, a single point (SP) test has also been developed. In the SP-NCTL test, five notched test
specimens are stressed at 30% of then- yield stress and evaluated in a similar procedure as the NCTL
test The recommendation is that none of the test specimens should fail within 200 hours.
Still further, a seam constant tensile load (SCTL) test has been developed whereby five
notched seam specimens are stressed at 30% of their yield stress and, again, are evaluated in a similar
procedure as the NCTL test The recommendation is that none of these test specimens should fail
within 200 hours.
This report was submitted in fulfillment of Cooperative Agreement CR 815692 by the
Geosynthetic Research Institute of Drexel University under the partial sponsorship of the US
Environmental Protection Agency. This report covers a period from May 11,1989 to May 10,1992
and was completed as of July 31,1992.
iv
-------
CONTENTS
Disclaimer . ii
Foreword „ Hi
Abstract iv
List of Figures vii
List of Tables ix
Acknowledgements x
1. Stress Cracking in Polymeric Materials 1
1.1 General Phenomenon of Stress Cracking 1
1.2 Types of Stress Cracking 2
1.3 Crack Initiation and Crack Growth Rate 7
1.4 Current Evaluation of Stress Cracking in HDPE Geomembranes 9
2. Occurrences of HDPE Geomembrane Stress Cracking in the Field 14
2.1 Literature Review 14
2.2 Current Study 15
3. The Notched Constant Tensile Load (NCTL) Test 24
3.1 Background and Development 24
3.2 Test Details and Protocol 24
3.2.1 Test Specimen Details , 25
3.2.2 Notching Details 25
3.2.3 TestProcedure 25
3.2.4 Data Collection and Presentation 28
3.2.5 Behavior of Response 28
3.2.6 Analysis of Response „ 31
3.3 Laboratory Testing Program 34
3.3.1 Reproducibility Tests 34
3.3.2 Intei-laboratory Test #1 , 36
3.3.3 InterlaboratoryTest*2 38
3.3.4 InterlaboratoryTest#3 40
-------
3.3.5 Laboratory Testing Program 42
3.3.6 Summary of NCTL Tests 45
3.4 Correlation Between Laboratory and Field Tests 46
3.5 NCTL Test Recommendations 48
3.6 Correlation Between NCTL Test and Bent Strip Test 48
4. Single Point Notched Constant Tensile Load (SP-NCTL) Test 51
4.1 TestDetails andProtocol 51
4.2 SP-NCTL Test Recommendations 51
5. Seam Constant Tensile Load (SCTL) Test 54
5.1 TestDetails andProtocol 54
5.2 SCTL Test Recommendations 56
6. Related Considerations 57
6.1 Effect of Different Temperatures * 57
6.2 Effect of Different Incubation Media 58
6.3 Other Tasks in This Research Report 60
6.3.1 Density Variation in the Seamed Area 60
6.3.2 Residual Stress Determination by Heating 62
6.3.3 Residual Stress Determination by Strain Gage Rosettes 62
6.3.4 Failure Surface Morphology 62
6.3.5 Accelerated Laboratory Degradation 63
7. Summary and Conclusions 64
8. References 66
Appendix "A" 69
• GRI Test Method GM5(a) - Notched Constant Tensile Load (NCTL) 70
Test Method
• GRI Test Method GM5(b) - Single Point Notched Constant Tensile Load 80
(SP-NCTL) Test Method
• GRI Test Method GM5(c) - Seam Constant Tensile Load (SCTL) 83
Test Method
Appendix "B" 87
• Notched Contant Tension Load (NCTL) Curves of Eighteen Laboratory Samples 88
and Seven Field samples Generated and Analyzed in This Report
VI
-------
LIST OF FIGURES
Number page
1 Conceptualization of ductile and brittle failure mechanisms in semir
crystalline polymer materials, after Lustiger and Rosenberg® 3
2 Scanning electron micrographs of ductile and brittle failure surfaces
of HDPE geomembrane sheet 4
3 Transmission electron micrograph showing tie molecule linkages between
crystalline regions of HDPE at 20,000 times magnification, after
Keith, et alP ., 5
4 Stress crack failures in HDPE geomembranes in the field 6
5 Micrographs (at magnification of 100 times) and schematic diagram
showing crack and craze formation in HDPE geomembranes 8
6 , Current method of qualifying HDPE geomembranes with respect to
stress cracking via ASTMD1693, the "Bent Strip Test" 10
7 ASTM D1693 "Bent Strip Test" specimens hi a containment channel
before and after removal from the incubation bath 12
8 Dimensions of a ASTM D1822 dumbbell test specimen 26
9 Schematic diagram of a notched ASTM D1822 test specimen 26
10 Photograph and schematic diagram of twenty station notched constant
tensile load (NCTL) test device 27
11 Typical response curves resulting from a complete notched constant
tensHe load (NCTL) test device 29
12 Photographs of ductile and brittle failures of NCTL test specimens 30
13 Scanning electron micrographs of ductile (pullout) and brittle (fibril)
NCTL test specimen failures 32
14 Various surface morphologies of NCTL test specimens in the transition
region 33
15 Reproducibility study on a single HDPE geomembrane material 35
vii
-------
Number
16
17
18
19
20
ASTM interlaboratory test-1, showing the NCTL test results from five
different laboratories
ASTM interlaboratory test-2, showing the NCTL test results from five
different laboratories
ASTM interlaboratory test-3, showing the NCTL test results from four
different laboratories
Transition time and stress results of NCTL testing on various HDPE
geomembranes
Transition time and transition stress results of NCTL tests on various
Bags
37
39
41
44
commercially available HDPE geomembrane and field exhumed
geomembranes 47
21 Illustration of logic behind a single point notched constant tensile load
(SP-NCTL) test 52
22 Seam constant tensile load (SCTL) test specimens 55
23 Results of NCTL tests performed at different temperatures together
with the result of RPM at 25°C 59
24 NCTL test performed in different incubation media 61
vm
-------
LIST OF TABLES
Number pagg
1 Fracture toughness (K) values of different polymers, after
Williams*?) 9
2 Selected differences between HDPE pipes and HDPE geomembranes 13
3 Summary of stress cracking survey, after GeoServices, Inc.^23^ 15
4(a) General description of the failure sites 16
4(b) Mechanical and physical properties of exhumed HDPE
geomembranes 17
4(c) Description of and cause of stress cracking at field sites 19
5 NCTLreproducibility tests 34
6 NCTLinterlaboratory test#l results 36
7 NCTL interlaboratory test #2 results 38
8 NCTL interlaboratory test #3 results 40
9 NCTL test results of various HDPE geomembrane samples 43
10 Test results on field retrieved HDPE geomembranes that failed in a
stress cracking mode 46
11 Comparison of ASTM D1693 bent strip test results with notched
constant tensile load (NCTL) test results 49
12 Effect of different immersion media on NCTL test results 60
IX
-------
ACKNOWLEDGEMENTS
This study was funded by the U.S. Environmental Protection Agency under Cooperative
Agreement No. CR-815692. David A. Carson was the Project Officer. Our sincere appreciation
extended to the Agency, Mr. Carson and also to Mr. Robert E. Landreth for this opportunity.
The cooperation of the facility owners of the different field exhumed sites is considered to
be invaluable to the results of this study. We offer this acknowledgement anonymously at the
respect of their wishes.
Throughout the course of this project active and very positive participatory involvement was
maintained with HDPE geomembrane manufacturers and resin suppliers. This group was involved
in the interlaboratory testing program and also formed the Geosynthetic Research Institute's
internal peer review panel. They are as follows:
Gundle Lining Systems, Inc. - Mark Cadwallader/Frederick Struve
National Seal Co. - Gary Kolbasuk
Poly-America, Inc. - Gerald E. Fisher/Jim Norbert
SLT Environmental, Inc. - William W. Walling
Novacor Chemicals Ltd. - Gary K. F. Yim
Union Carbide Corp. - Anthony Nicholas/William Hoffman/Juris Aspe
Phillips Petroleum Co. - Rex L. Bobsein
Solvay Polymers, Inc. - Michael R. Mahan/John L. Hammond
Mobile Chemical Co - Frank A. Nagy
Southwest Texas State University - Patrick Cassady
External peer reviewers on behalf of the U.S. Environmental Protection Agency were as follows:
• Precision Laboratories - Ronald Belanger
• J & L Engineering - John Boschuk, Jr.
• Umweltbundesamt - Klaus Stief
The authors extend their appreciation to all of the above for their voluntary involvement in the
subject matter of this report.
Robert M. Koerner
Yick (Grace) Hsuan
Arthur E. Lord, Jr.
-------
SECTION 1
STRESS CRACKING IN POLYMERIC MATERIALS
This initial chapter is presented to introduce the general subject of environmental stress
cracking in polymeric materials, which in this report will be referred to as "stress cracking". The
chapter will also serve to give an overview of the various types of stress cracking and will present
the current status of stress crack evaluation for high density polyethylene (HDPE) geomembranes.
It should be understood that throughout this report the HDPE polymer being referred to is a linear
medium density polyethylene resin which, when formulated and manufactured into sheet, has the
following general properties:
• Density = 0.940 to 0.950 g/cc
• Melt Index (conditionE) = 0.08 to 0.40 g/lOmin.
• Carbon Black Content = 1.5 to 3.0%
1.1 GENERAL PHENOMENON OF STRESS CRACKING
Stress cracking in polymeric materials is not a newly discovered phenomenon in the plastic
industry, nor is it associated with a single type of polymer. The specific mechanisms leading to
stress cracking are numerous, e.g., chain scission, bond breaking, cross linking, extraction of
various components, etc. See Reference 1 for an overview of how degradation proceeds from the
original polymeric material to the point of embrittlement (a form of stress cracking) and the
eventual termination of the material's functional life.
The definition of stress cracking according to ASTM D883 is "an external or internal crack
in a plastic caused by tensile stress less than its short-term mechanical strength"/2) This type of
cracking generally refers to brittle cracking whereby there is no, or very little, ductile drawing of
the polymeric material from its adjacent failure surfaces.
"V
The evaluation of stress cracking in plastics has been studied by a large number of
investigators. A sampling of key references are Choi and Broutman®, Chan and Williams^, and
Lu and Brown®, but many more are available. In general, the phenomenon is referred to as
"environmental stress cracking" in the literature since the test is always performed at an elevated
temperature and/or immersed in an accelerated wetting agent (i.e., a surfactant).
-------
Of particular interest with regard to an understanding of the stress cracking phenomenon in
semi-crystalline polymers is the conceptual molecular models of Lustiger and Rosenberg® which
are shown in Figure 1. These sketches present the hypothetical behavior of polyethylene under
small deformation (Figure l(a)), then during a ductile failure whereby the crystalline regions
gradually pull apart in a cold drawing mode (Figure l(b)), and finally during a brittle failure
whereby the crystalline regions separate virtually intact via tie molecule breaking in an abrupt mode
(Figure. l(c)). These failure modes can be further appreciated via the micrographs shown in
Figure 2. Figure 2(a) shows a typical macroscopic ductile failure morphology, while Figure 2(b)
shows a typical brittle failure morphology in HDPE geomembranes. The ductile failure
morphology is one wherein the polymer is uniformly "pulled out" of the adjacent material. In
contrast, the brittle failure shows a fibril morphology wherein the polymer "cracked" with no, or
little, drawing from one surface to another.
Upon considering the above described behavior it should be stated that tie molecules are the
only polymer chains which can transfer load from one crystal lamella to another. To illustrate such
linkages, Figure 3 shows a high magnification micrograph^ of tie molecules bonding together the
crystalline regions in HDPE. The tie molecules are the fundamental property which controls the
stress cracking behavior of the material, assuming other properties are equal.
The major points to be gained from these figures and from this initial discussion on stress
cracking are the following:
• The stress crack resistance of a semi-crystalline polymer like high density polyethylene is
primarily a function of the resin itself. Processing, additives, exposure, seaming, etc.,
by themselves cannot generate stress. cracking, they can only exacerbate the
phenomenon.
• On a molecular basis, the tie molecules hold a key role in the stress crack resistance of
semi-crystalline polymer resins. They physically tie, or bond, the crystalline regions of
the material into a coherent structural unit thus forcing a ductile, rather than brittle,
behavior of the material to occur when placed under tension.
• Since evaluation of polymer stress cracking on the microscopic scale is very difficult,
tedious and expensive to perform, laboratory scale experimental tests are usually used to
indirectly evaluate the stress cracking phenomenon.
1.2 TYPES OF STRESS CRACKING
When discussing the stress cracking behavior of semi-crystalline polymeric materials there
are two different phenomenon that may be involved. They have been classified in the literature and
are generally referred to as rapid crack propagation (RCP) and slow crack growth (SCO).
RCP is associated with stress cracking that occurs at very high velocities, generally over
300 m/sec. ® The extent of the cracking can be over hundreds of meters in length and it usually
occurs in a dendritic pattern when it happens in geomembranes, see Figure 4(a). The ambient
-------
Tie
Molecules
Cilia
Loose Loop
(a)
(a) Initial steps in the deformation of polyethylene
N *
(a) «»
(b) Steps in the ductile deformation of polyethylene
i (tr*i yi -Jr.. V
(c) Final step in the slow crack growth of polyethylene
Figure 1 - Conceptualization of ductile and brittle failure mechanisms in
semi-crystalline polymer materials, after Lustiger and Rosenberg(6)
-------
(a) Ductile failure (at 30 times magnification)
(b) Brittle failure (at 30 times magnification)
Figure 2 - Scanning electron micrographs of ductile and brittle failure
surfaces of HOPE geomembrane sheet.
-------
Figure 3 - Transmission electron micrograph showing tie molecule
linkages between crystalline regions of HOPE at 20,000 times
magnification, after Keith, et al. ^7\
-------
(a) Rapid crack propagation (RCP)
(b) Slow crack growth (SCG)
Figure 4 - Stress crack failures in HDPE geomembranes in the field.
-------
conditions of the few failures that have occurred in this mode have been on exposed
geomembranes when they were subjected to extremely cold night time temperatures, e.g., less than
-20°C. They are apparently initiated by some kind of dynamic or impact type of force.
In contrast to the above, SCO occurs at velocities less than 0.1 m/sec(9X Hence the extent
of cracking and the crack propagation growth can occur during the service lifetime of the material.
This phenomenon is of considerable interest. Figure 4(b) shows a slow crack growth failure
adjacent to HDPE geomembrane seam. The crack growth rate varies in accordance with the type
of polymer, applied stress and the temperature. From a stress viewpoint, SCO is always
associated with applied stresses lower than the yield strength of the material. Since SCG can occur
at applied stresses as low as 20% of yield, this particular mechanism is of concern with regard to
the long-term durability of the material, i.e., SCG could define a geomembrane's service
"lifetime".
For semi-crystalline geomembranes, such as HDPE, RCP is possibly installation related to
some degee. Sufficient slack in the geomembrane should be introduced to compensate for cold
weather contraction of the liner during the exposed period. This is evidenced by the observation
that the RCP failures studied here have occurred in Canada and the north central United States
during the intensely cold winter of 1989. Conversely, SCG is a long-term phenomenon which
might occur during the service lifetime of the geomembrane. Hence, an analysis of SCG behavior,
review of current test methods, possible upgrading of the existing test methods, and relationships.
to installation procedures should be evaluated. SCG will be understood to be the mechanism of
stress cracking addressed for the entirety of this report. While it is not known with certainty, it
may be the case that selection of the proper stress crack resistance resin to avoid SCG, coupled
with proper installation procedures, will mitigate RCP problems as well.
1.3 CRACK INITIATION AND CRACK GROWTH RATE
For stress cracking to occur in a semi-crystalline geomembrane material, like HDPE, the
material must fulfill two conditions. First, it must be under some form of stress and second it must
have a point at which a crack can initiate. The first condition is obvious because with no stress in
the geomembrane, there can be no stress cracking. The second condition of a crack initiation point
could be at an imperfection, scratch, carbon black agglomeration or other anomaly, but will most
likely be at the location of a seam. This is due to the natural discontinuity of the overlap
configuration used to seam geomembranes and also possibly due to overheating of fusion seams
and/or excessive grinding associated with extrusion flat or fillet seams. There is a recent EPA
Technical Guidance Manual focused on proper geomembrane seam fabrication and inspection
techniques which will hopefully avoid these types of field installation problems/10^
Once a stress crack has initiated, it grows perpendicular to the stress orientation via a series
of built-up "crazes". The periodical tension failure of fibrils in the craze permits the crack to
propagate. Figure 5 shows a micrograph of a stress crack, crazes and the related progress of the
stress cracking phenomenon. The rate of propagation of the crack in this manner is referred as its
crack growth rate, "da/dt", which is correlated to the material property known as fracture
toughness, "K". The empirical relationship can be expressed as follows:
-------
Fibril
Void
Figure 5 - Micrograph (at magnification of 100 times) and schematic diagram
showing crack and craze formation in HDPE geomembranes
-------
where:
K = fracture toughness (MPa-m0-5), see Table 1
da
-r- = crack growth rate (m/sec)
dt
p = constant (dimensionless)
[value ranges from 0.5 - 0.125 for PE materials]
q = constant with dimensions of [(MPa-m°-5)(m/sec)~P]
Table 1 - Fracture toughness (K) values of different polymers, after Williams^9)
Materials Fracture Toughness (MPa - m°-5)
polystyrene (PS) 0.7-1.1
polycarbonate (PC) 2.2
polyvinyl chloride (PVC) 2.0-4.0
polypropylene (PP) 3.0-4.5
polyethylene (PE) 1.0-6.0
polyamide (PA) 2.5-3.0
polyester (PET) ~ 5
1.4 CURRENT EVALUATION OF STRESS CRACKING IN HOPE GEOMEMBRANES
The current test method used to qualify semi-crystalline geomembrane sheet materials, like
HDPE, against stress cracking is ASTM D1693, commonly known as the "Bent Strip Test".(n)
The test is one in which a surface notched rectangular test specimen is bent in a 180° arc and placed
within the flanges of a small metal channel. The size of the test specimen is 38 mm long by 13 mm
wide. The notch is placed on the outer surface of the bend of the test specimen running parallel to
its length and is approximately 20% of the sheet material's thickness in its depth. Usually ten of
these test specimens are placed adjacent to one another in a channel holder and then the entire
assembly is immersed in a surface-active wetting solution which is maintained at an elevated
temperature. See Figure 6 for sketches of the test specimen, the channel holder and a typical
immersion test tube.
It should be noted that the bending moment placed in the specimen by positioning it in a
180° arc induces a hoop stress along the direction parallel to the notch. When cracks occur in the
-------
3
7
H
•t
on
I
CQ
en
Ox
vo
E
E
in
10
•8
1
u
(/3
*J
00
E
E
\
E
E
ro
E
E
03
to
10
-------
test specimens, they generally develop at the sides of the notch near the top of the arch, growing
towards the edge of the specimen at approximately right angles to the notch. At the same time the
cracks grow through the thickness of the specimen until it breaks. Several failed and unfailed
specimens can be seen in the specimen channel holder shown in Figure 7.
Although the Bent Strip Test is very simple to perform, it has two problems associated with
it. The first problem is that the test specimens (under their individual hoop stress conditions) will
tend to stress relax during the progress of the test The rate of stress relaxation is variable between
different materials and different thicknesses and is completely unknown as to its magnitude and
behavior. To merely require the test to be run for a long time period does not necessarily make the
test more challenging to the material. Due to the stress relaxation issue, a comparison of different
HDPE geomembranes is questionable and should be viewed with considerable caution. The
second problem has to do with the qualitative nature of the test results, i.e., either the specimen
cracks within the stipulated time or it does not. One has no meaningful way of assessing the
relative degree of stress cracking resistance of the different materials evaluated since there is no
meaningful way to quantify the results, and certainly no way to differentiate between the many
materials which perform well in this test.
Not withstanding the above comments, the Bent Strip Test is routinely used by the HDPE
geomembrane industry and is performed according to various documents as described in U.S.
EPA/600/2-88/052/12) The current trend is to use a specific surface active agent, i.e., Igepal (CO
630), in a 10% concentration mixed with 90% tap water. A constant temperature incubation is to
be maintained at 50°C for a test duration of 1500 hours. (Note that the recommended solution and
incubation times were recently changed by the geomembrane industry to the above values from
100% Igepal and 500 hours incubation time, both of which are less severe than the current
recommendations). Furthermore, most specifications require that ten test specimens be evaluated,
five in the machine direction and five in the cross machine direction, and that no failures occur in
these ten specimens within the recommended time period.
All manufacturers of HDPE geomembranes (as well as resin suppliers, consultants, test
organizations, owners, operators and regulators) use the above recommendations, i.e., ASTM
D1693 Bent Strip Test under the conditions just stated, as the current state-of-the-art for HDPE
geomembranes insofar as stress crack resistance testing is concerned.
It should also be noted that the plastic pipe industry (which has an on-going catalogue of
field pipe failures)^13) has a very different formalized test procedure for qualifying HDPE
resins/compounds used in gas transmission lines. In their test procedures, small lengths of capped
pipe sections are pressurized and placed in constant temperature baths until they fail. The time
required for failure can be years and major research institutions around the world have ongoing
tests of this type. The results of such tests can be used not only to qualify the polymeric materials
but also can be used for lifetime prediction procedures. Unfortunately, such tests can not be used
for geomembranes due to different geometry, different stress conditions, and different seam types.
Furthermore, one simply cannot conclude that "what is good for pipe, is good for geomembranes".
Some marked differences between HDPE pipe and HDPE geomembranes are listed in Table 2.
11
-------
(a) Test specimens before being placed into the incubation bath
(b) Test specimens after removal from the incubation bath,
showing cracked and uncracked specimens.
Figure 7 - ASTM D 1693 "bent strip" test specimens in a containment channel
before and after removal from the incubation bath.
12
-------
Table 2 - Selected differences between HDPE pipe properties and HDPE geomembrane properties
Item
HDPE Geomembranes
HDPE Pipes
Resin density
Melt flow index
Carbon Black
Additive Package
Processing
Thickness
Seaming
Stress Crack Test
Testing Duration
0.935 to 0.945 g/cc
0.08 to 0.40 g/10 min
1.5 to 3.0%
unknown
sheet casting, or
blown film extrusion
1.0 to 2.5 mm
overlap extrusion, or
overlap fusion
ASTMD1693
1500 hours
0.950 to 0.956 g/cc (typical)
0.15 to 0.6 g/10 min
0.0 to 2.5%, or other pigments
unknown
pipe extrusion
10 to 25 mm
butt fusion, or
mechanical
ISO/TC 138N1081<14>or ASTMD1598<15)
1 to 10 years
13
-------
SECTION 2
OCCURRENCES OF HDPE GEOMEMBRANE STRESS CRACKING IN THE FIELD
In order to establish the need for the experimental laboratory work to follow, a detailed
review of stress cracking occurrences of HDPE geomembranes that have occurred in the field is
necessary. This chapter is presented in two parts; first, the available literature on the subject and
second the field situations which have been investigated during the course of this study.
2.1 LITERATURE REVIEW
While the open literature is quite extensive on stress cracking of plastics (recall Section
1.1), it is rather limited on stress cracking of geomembranes. One reason, of course, is that semi-
crystalline geomembranes like HDPE have only been used since the early 1980's, but another
reason is that the occurrences have been relatively rare with respect to the amount of HDPE liner
that has been installed.
Interestingly, three state-of-the-art reviews on stress cracking in HDPE geomembranes, in
three different countries (USA, Canada and Germany), all appeared in the mid-1980's. Under
EPA sponsorship, Telles, et al.^ provided an annotated bibliography on the interrelated topics of
HDPE geomembrane stress cracking and various seaming methods. Not only was the ASTM
D1693 test referenced, but also a constant tensile load test designated as ASTM D2552^17^ and
known as the Lander's test was also reviewed.
The Canadian reference by Fruch, et al/18) was prepared for Alberta Environment and was
a very complete treatise on various test methods for polyethylene geomembranes. Included in the
stress cracking portion of the report was a description of the molecular structure of HDPE, its tie
molecule arrangements, the function of cracks and crazes, and related theories as to possible
mechanisms and behavior. They found that the current test methods were "in a state of disarray"
and that additional research and development were warranted.
Lastly, the Hoechst group in Germany under the leadership of Koch , have reviewed
(and greatly extended) lifetime prediction methods for HDPE geomembranes. Essential to their
extrapolation of accelerated temperature testing is the fundamental ductile-to-brittle behavior of
HDPE which holds the key toward the test method which will be developed later in this report.
A few isolated papers on the subject of HDPE geomembrane stress cracking have also
14
-------
appeared in the literature. Fisher^20), in 1989, reported on several field failures noting the
importance of manufacturing, installation and seaming in preventing the phenomenon from
occurring. Peggs^21), in 1989, reviewed several field failures along with the actual remedial
measures taken at the various sites/22) Finally, a survey made for the U.S. EPA by GeoServices,
Inc/23) resulted in the information provided in Table 3.
Table 3 - Summary of stress cracking survey, after GeoServices Inc/23)
(Parentheses indicate number of sites with stress cracks)
Mfgr. Type of Installation Environment
Exposed Covered Years in
Solid Liquid Liner Liner Service
Waste Waste Total Hot Cold Both
1 2 (0) 7 (3) 9 (3) 1 (0) 3 (3) 5 (0) 9 (3) 0 (0) 7-10
2 . 17 (2) 6 (2) 23 (4) 16 (3) 4 (1) 3 (0) 19 (4) 4 (0) 2-6
3 4(0) 14(10) 18(10) 11(9) 0(0) 7(1) 18(10) 0(0) 1-8
As seen in the above Table, seventeen (17) stress cracking incidents were discovered, all of which
were on exposed HDPE geomembranes. The conclusion of the GeoServices study (23\ as well as
Telles, et al.(16) and Fruch, et al/18\ was that more appropriate testing methods were needed to
evaluate stress crack resistance than currently exists. The recommendations for preventing the
phenomenon were focused into four distinct categories; (a) materials and seams, (b) design, (c)
construction, and (d) regulations. Included in the preparation of the GeoSyntec study of Table 3
were numerous conversations and data taken from this current research effort which is also under
the sponsorship of the U.S. EPA.
2.2 CURRENT STUDY
One of the tasks in the current study was to investigate stress cracking in the field at
available sites-of-opportunity. Only HDPE geomembranes are included in this data set although
stress crack failures were also investigated in two other geomembrane types.
Table 4 presents fifteen (15) case histories of stress cracking of HDPE geomembranes
which were investigated during the course of this study. Table 4(a) gives a general description of
the sites, Table 4(b) gives the mechanical and physical properties of the exhumed material and
Table 4(c) gives the stress cracking details. It is important to note that approximately seven (7) of
the sites listed in the GeoServices study of Table 3, are co-listed in this study of Table 4. For most
15
-------
Table 4 fa) General description of the failure sites evaluated in this study
(numbers are given throughout this table since site accessibility was given voluntarily
and the information is proprietary)
Site No. Region
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
SW
SW
NE
Purpose Thickness Mfgr.
(mils)
SI 80 #1
(Sludge)
SI 100 #2
(Raffinate)
SI 60 #4
(Leachate)
NW SI 60 #2
(Gypsum Sludge)
NW
SW
SW
SW
SW
SE
MW
SE
NW
(Can.)
NW
(Can.)
SW
SI 80 #3
(Leachate)
SI 100 #1
(Brine)
SI 100 #1
(Brine)
r
SI 80 #1
(Evap. Pond)
SI 60 #3
SI 60 #2
(Water)
SI 40 #5
(Sewage)
SI - #3
SI 60 #1
SI 80 #1
(Black Liq.)
80 #3
Install.
#1
#4
#8
, #2
#3
#1
#1
#1
#3
#6
#5
#3
#1
#7
#3
Resin Seam
type
#4 #1 & #2
#2
#4
#1 #2
#2 #3 & #2
#3 #1 & #2
#3 #1 & #2
#5 #1
#3
#2
#4
#3
#1
#2 & #3
#3 #2
Const.
Date
Fall
1981
June
1987
1986
Summer
1986
Summer
1988
Spring
1983
Spring
1982
Fall
1987
-
1987
1985
-
-
Summer
1991
Service
(years)
4 to 7
1
2
3 months
(Nov.-Jan.)
4 months
(Nov.-Feb.)
2 to 6
2 to 6 .
1
-
1
3
-
-
2
<1
Notes : Mfgr. = Manufacturer,
Const = Construction,
SI = Surface impoundment.
Install. = Installer,
Service = Length of service,
16
-------
Table 4 ffr) - Mechanical and physical properties of exhumed HOPE geomembranes
Site No. Density C.B. M.I. (E) FRR D 1693* NCTL*
(g/cc) (%) (g/lOmin.) (P/E) (hours) Tt(hr.) T(J(%CTy
1 0.952 2.5 0.23 5.00 600
(7/10 failed)
2 - - - - - - -
3 0.956 2.5 0.17 4.7 1000 97 35
(no fail)
4 0.962 2.4 0.27 5.00 400 -
(3/5 failed)
5 0.949 2.4 0.57 4.2 300 7 43
(2/5 failed)
6 0.954 2.2 0.19 6.4 700 4 41
(9/10 failed)
7 0.956 2.0 0.22 4.7 72
(10/10 failed)
v
8 0.955 2.4 0.28 4.2 1000 27 35
(1/10 failed)
9 .
SCTL*
) (hours)
#1 Seam :
91 at 35% CTy
#2 Seam:
66 at 31.5% CT
-
#4 Seam:
240 at 35% CTy
#2 Seam:
4 at 35% CTy
#3 +#2 Seams:
44 at 35% CTy
#3 Seam:
29 at 35% CTy
#1 + #2 Seams:
20 at 35% CT
33 at 31.5% CTy
#2 Seam:
15 at 35% CT
15at31.5%CTy
#1 + #2 Seams:
38 at 31.5% CTy
41at28%CT
#1 (Ballast)
42 at 35% CTy
65 at 3 1.5% CTy
#1 (side 1, 2)
246 at 3 1.5% CTy (1)
152 at 3 1.5% CTy (2)
-
(continued)
17
-------
Table 4 (b) - Continued
Site No. Density
C.B. M.I. (E) FRR
(%) (g/lOmin.) (P/E)
D 1693*
NCTL*
(hours) Tt(hr.)
SCTL*
) (hours)
10 0.958 2.2
0.18
4.7
48
(10/10 failed)
#2 Seam
14 at 35% or
18 at 31.5%
11
12
13 0.954 2.3 0.16 5.0
2000 55 30
(0/20)
14 0.948 2.3 0.56 to 2.9 to 2000 17 42 #2 Seam:
0.48 3.4 (7/10) 75 at 31.5% o\,
15 0.949 2.4
0.57
4.3
500
(10/10)
11 42.5
#3 Seam:
30 at 30%
Notes: C.B. = Carbon Black
M.I. = Melt Index, Condition E
FRR = Flow Rate Ratio (Condition P/Condition E)
* = tested at 50°C in 10% Igepal CO 630/90% Tap Water
18
-------
Table 4 fc) - Description of type and cause of stress cracking at field sites
Site No. Morph. Crack
Type
1 Fibrous Slow
2
3 Flake Rapid
4 Fibrous Slow
5 Fibrous Slow
& Flake then
Rapid
6 Fibrous Slow
7 Fibrous Slow
8 Fibrous Slow
9 Fibrous Slow
10 -
11 - Slow
Crack Crack Loc.
Severity at Seam*
Major LS & Ext.
(hundreds of cracks
short and long)
Moderate
Minor LS
(one-60 cm long
crack)
Major LS
(dozens of cracks,
short and long)
Major LS
(entire section
cracked spreading
throughout sheets
and seams)
Major LS & Ext.
(dozens of cracks,
short and long)
Moderate LS
(10 cracks 2 to
50 cm in length)
Major US
(many cracks, (due to
short and long bending)
Minor
Minor LS
(two-45 cm cracks)
Minor LS
(2 small seam
cracks)
Crack Loc. Cause
at Site"1" of Stress
2, 3 Thermal
2,3 Thermal (?)
3 Hydraulic
2, 3, 4 Thermal
3, 4 Thermal
2, 3 Thermal
2, 3 Thermal
6 Thermal
3
2,3
2 Thermal
Cause
of Failure
Heavy
grinding
High temp.
during Seaming (?)
Probably due to an
impact at the seam
Heavy grinding and
high temperature
during seaming
Extreme cold (-25°C)
initiated at seam
spread throughout
high temperature
seaming
Heavy grinding and
high temperature
during seaming
Heavy grinding and
high temperature
during seaming
Anchored in
position, heavy
grinding
-
-
Overheating and
generally poor
workmanship
(continued)
19
-------
Table 4 (cl - Continued
Site No. Morph. Crack Crack
Type Severity
Crack Loc. Crack Loc. Cause
at Seam* atSite+ of Stress
Cause
of Failure
12
Rapid
13 Damaged Rapid
surface
Major
(entire section
cracked spreading
throughout sheets
and seams)
Major
(entire section
cracked spreading
throughout sheets
and seams)
3,4 Thermal
4, 5, 6 Thermal
Extreme cold
Extreme cold
(-40°C)
14 Fibrous Slow
15 Flake Rapid
Major LS
(entire section
cracked spreading
throughout sheets
and seams)
Major LS and/or
(single crack US
~48 mlong)
4, 5, 6 Thermal
Intermittent hot
liquid at 65°C
created folds in the
liner. Cracks formed
at the apex of folds
Heap Leach Dynamic Loading
Loading during filling
Notes:
Morph. = Morphology
* Crack Location at Seam:
LS = lower sheet at seam
US = upper sheet at seam
EXT = through the seam extrudate
+ Crack Location at Site :
®
Vhigh ®/ - 1®
V low
20
-------
of the sites in Table 4 a complete forensic analysis was performed by GRI and close interaction
was maintained between the owner/operator and manufacturer/installer. Summary comments taken
from each successive column of the information provided in Tables 4(a), 4(b) and 4(c) follow.
• All geographic sections of the USA (and two in Canada) are included. Six sites were in
the Southwest, four in the Northwest and Canada, and the other,five were spread
throughout the Central, Southeast and Northeast regions.
• Due to their occurrence, only exposed geomembranes (either the exposed liner material of
surface impoundments above the liquid level, or during construction and before
backfilling) were involved in the data set. No opportunities were available to examine
buried HDPE geomembranes such as those used for landfill liners.
• Liner thickness varied from 1.0 to 2.5 mm (40 to 100 mils).
• All five major manufacturers of HDPE geomembranes have at least one installation
involved.
• In addition to direct installation by the major manufacturers, several cases occurred where
independent installation firms have been involved.
• In general, the specific type of resin is difficult for the authors to identify. Where it is
known, its identity is numerically listed in Table 4(a). This same comment applies to the
additive package as well.
• All four major HDPE seaming methods have been involved, i.e., flat extrusion, fillet
extrusion, hot wedge and hot air. However, the large majority of failures were associated
with the two extrusion types of seams (flat and fillet). In addition, five failures involved
double seaming, i.e., a fillet extrusion seam placed over a flat extrusion or hot wedge
seam.
• Manufacturing of the geomembrane and its field installation ranged from 1981 (which
coincides with the earliest HDPE installations) to 1991.
• The length of service time varied from 3 months to 7 years.
• Geomembrane densities, ASTMDISOS^24), ranged from 0.948 to 0.962 g/cc.
• Carbon black content, ASTM 1603<25), ranged from 2.0 to 2.5%.
• Melt flow index values, ASTM D 1238@$ condition E, ranged from 0.16 to
0.56 g/10 mm.
• Row rate ratio, ASTM 1238@^ condition P/condition E, ranged from 2.4 to 4.3.
• Bent strip test results, per ASTM D1693 (^ results ranged from all samples failing, to no
samples failing.
21
-------
• Notched constant tensile load tests (the proposed geomembrane sheet test method to be
described fully in Chapter 3) ranged in transition times from 4 to 97 hours and in
transition stress from 20% to 45% of yield stress.
• Seamed constant tensile load tests (the proposed geomembrane seam test method to be
described in Chapter 5) ranged in failure times from 4 hours to 240 hours.
• The surface morphologies of the failure surfaces (as viewed under scanning electron
micrographs) were either fibril or flake structures, see Reference 27 for descriptions of
the various types of failure morphologies.
• Both slow and rapid crack growth rates were involved in the various field sites. In Table
4(c) it is seen that four sites were of the RCP type; the others were SCO which is the type
of stress cracking of concern herein and the thrust of this study.
• The estimated severity of cracking is listed with values ranging from extremely small,
e.g., at Sites 10 and 11, to extremely large.
• The location of the cracks when they occurred at seams (which was the general situation
for SCO) was generally in the lower sheet.
• The location of the cracks within the facility itself was always in the exposed runout
length or along the side slopes above the liquid level for surface impoundments. For the
few cases where the cracks were at the bottom of the facility it was during construction
and before filling, i.e., the geomembrane was exposed.
• The cause of the tensile stresses was almost always thermally induced resulting in
expansion and contraction of the geomembrane. However, there are a few anomalies,
i.e., hydraulically induced or by a heavy and rapid loading.
• The cause of the initiation of the failure varied, but heavy grinding (for extrusion seams)
and/or high seaming temperatures (from double seaming) were generally involved.
In summarizing the findings presented in Tables 4(a), 4(b) and 4(c) it is seen that all field
failures investigated were associated with exposed geomembranes (although opportunities to
evaluate buried, or even covered, geomembranes were never made available). For atmospheric
exposed geomembranes, temperature and/or ultraviolet light induced stresses play a significant role
in stress cracking of HDPE geomembranes. Furthermore, the failures were generally of the slow
crack growth (SCO) type of mechanism. It is not known if Bent Strip Testing via the ASTM
D1693 test method was performed on the original and as manufactured sheet, but since the test has
been in existence for over 20 years it is likely that it was done and that the samples had performed
adequately.
From Table 4(c) it is also seen that the initiation of the cracks was generally at the
discontinuity formed by overlapping seams and was associated with heavy grinding and/or
overheating. Poorly constructed seams can initiate stress cracking. The tensile stresses were most
likely induced by contraction due to thermal cooling of the geomembrane which had insufficient
slack constructed in it in order to accommodate the contraction. Thus, proper CQC/CQA is also an
important factor in minimizing the occurrence of stress cracking.
22
-------
However, stress cracking is fundamentally associated with the polymer itself at the
molecular level, i.e., stress cracking is primarily a resin and geomembrane sheet issue. Hence, the
development of the best possible test method to assess and/or qualify various geomembrane resins
and sheet materials is the main finding of the literature search and field study presented in this
chapter. This topic is the subject of the following chapter and the remainder of this report.
23
-------
SECTION 3
NOTCHED CONSTANT TENSILE LOAD TEST
The notched constant tensile load (NCTL) test is the focus of this chapter. Its background,
development, details, procedures, mterlaboratory evaluations, correlation to laboratory and field
results, final recommendations as to the utilization of the test and correlation to the Bent Strip Test
will be presented.
3.1 BACKGROUND AND DEVELOPMENT
As reviewed in Section 1.4, the Bent Strip Test is the currently used test method to assess
HOPE geomembranes for their stress crack resistance. Its limitations and disadvantages were
described previously and will not be repeated here.
A different test to assess stress cracking in the form of constant tensile load test was
developed in 1960 by Lander^28) and was subsequently formalized by ASTM as D 2552(17), "Test
Method for Environmental Stress Rupture of High Density Polyethylene Under Constant Tensile
Load". The test used uniformly curved dumbbell shaped specimens, placed them under various
constant stresses, and then incubated them in a constant temperature bath containing a wetting
agent to reduce the surface tension of the liquid. Unfortunately, the test took very long to complete
(generally over 10,000 hours) and furthermore the test results were generally scattered. Therefore,
the test data required a statistical method for quantification. This statistical variation was due to the
random nature of the existence of a crack initiator within the narrow section of the test specimens.
If a surface scratch, or other anomaly existed in one specimen, its likelihood of being at the same
location, orientation and depth in another specimen (much less in all of the test specimens) was
very remote. Nevertheless, the concept of a constant tensile load test was of great interest.
Unfortunately, the test was discontinued by ASTM in 1986 due to inactivity by the manufacturing,
user and specifier communities. Shortly thereafter the test was adopted by GRI and modifications
have been advanced to arrive at the current state.
3.2 TEST DETAILS AND PROTOCOL
The notched constant tensile load (NCTL) test is performed in accordance with the
following test procedure. It will only be briefly described here since a copy of the complete testing
protocol exists as GRI Test Method GM-5(a) and is included as an Appendix to this Report. The
test and its variations is also under review by Task Group D35.02.89.01 within ASTM's
Committee on Geosynthetics chaired by Dr. Hsuan, a co-author of this Report.
'24
-------
3.2.1 TEST SPECIMEN DETAILS
In conducting NCTL tests, the test specimens are taken parallel to the cross machine
direction of the geomembrane sheet. This will also be the direction of general stress application
which is the most sensitive orientation of an extruded geomembrane sheet for stress cracking. If
this is in doubt, test specimens can also be taken along the machine direction of the sheet. The test
specimens are cut out of the full sheet or representative sample thereof using a ASTM D1822
die^X This produces a dumbbell shaped specimen with a constant width central section 3.2 mm
wide and 10.0 mm long, see Figure 8. Metal grommets are then placed in the enlarged end tabs of
the specimen for eventual application of a tensile load.
3.2.2 NOTCHING DETAILS
Using a single edge razor blade, a notch is made in the surface of the test specimen to a
controlled depth. The notch is located at the center of the constant width section as shown in
Figure 9. The razor blade is held in the upper grip of a compression testing machine. The rate of
advancement of the razor blade into the specimen is 0.25 mm/min (0.010 in/min). A new razor
blade is used after notching each set of 20 test specimens.
The depth of the notch is such that 80% of the nominal sheet thickness remains as an intact
section of the material which will be referred to as its "hinge". Thus the actual thickness of each of
the test specimens must be individually measured with a micrometer so as to compensate for
geomembrane sheet thickness variations between the different test specimens.
3.2.3 TEST PROCEDURE
After preparing and notching a group of test specimens they are then mounted in a
stationary tensioning device via hooks which are inserted into the grommets at the bottoms of the
specimens (the fixed end) and at the tops of the specimens (the movable end). The movable end is
actually a pivot arm which has a known mass on its opposite end. Usually a mechanical advantage
is utilized so as to minimize the amount of dead load required to reach a desired stress level. The
type of device just described is shown in Figure 10 which (with the notable exception of test
specimen size and preparation) is similar to the Lander's device mentioned earlier.
A specific amount of dead weight (typically lead shot) is added to a can at the back end of
the load arm so as to reach the desired stress acting on the test specimen. In the calculation of the
desired applied stress, the cross sectional area is the hinge area remaining after notching. Also the
stress level is based on the yield stress of the material at standard temperature and humidity, i.e.,
23°C and 50% relative humidity in accordance with ASTM D 638 (30\ This is the conventional
test method used for determining the stress-vs-strain behavior of HOPE geomembranes.
The required applied stress levels to be evaluated for the individual test specimens are from
20% to 65% of the yield stress of the material in increments of 5%, or less. It is sometimes
desirable to use 2.5% stress increments in the vicinity of the transition portion of the response
25
-------
25.00
60.00
r'
.50
8.75
i 4
^
Figure 8 - Dimensions of a ASTM D 1822 dumbbell test specimen
O
hinge thickness
0.60 t
t = norminal thickness
of geomembrane
Notch
// \
O
Figure 9 - Schematic diagram of a notched ASTM D 1822 test specimen
26
-------
Micro Svitchto
Timers
Joints
Dead
Weight
20 Positions
75.20cm
(30") •
Specimen
SIDE VIEW
Tray|Tiav8dupand
dovvn on rack and
pinion arrangement
FRONT VIEW
(a) Schematic diagram
of NCTL test device
(fromRef. (17))
(b) Front view of
the test device,
showing the timing
clocks.
(c) Back view of
the test device,
showing the
loading cans.
Figure 10 - Photograph and schematic diagram of twenty stations
notched constant tensile load (NCTL) test device.
27
-------
curve to better define the onset of brittle behavior.
Three specimens should be tested at each individual stress level and the average of the
failure times of these three tests used to plot a single data point on the resulting response curve. If
the standard deviation of the three values is greater than 15% of the mean value, the three tests at
this particular stress level must be repeated. Thus, a minimum number of 30 test specimens are
required to obtain a complete response curve (10 stress increments at 3 specimens for each
increment). The test device shown in Figure 10 can handle 20 test specimens simultaneously.
An immersion bath properly sized to enclose all of the test specimens is prepared with the
required wetting solution. This solution must be 10% Igepal (CO-630) mixed with 90% tap water.
(If disputes arise in the testing program, the tap water can be replaced with deionized water). The
immersion bath is raised so as to completely surround the test specimens and is then brought from
room temperature up to 50°C + 1°C. This incubation temperature is maintained throughout the test.
Constant circulation of the solution is required as well as periodic replacement of the water that is.
lost due to evaporation.
3.2.4 DATA COLLECTION AND PRESENTATION
Each specimen should be equipped with a timing clock which runs continuously until its
corresponding test specimen fails. Failure comes about by either a limiting ductile elongation or by
brittle failure. In either case, the load arm moves through a distance of 20 mm (3/4 in.) which trips
a switch thereby stopping the timing clock associated with each test specimen.
The times for failure of the three test specimens evaluated at the same stress level are then
averaged and the standard deviation is calculated. As described above, if the standard deviation is
less than 15% of the mean value, the data are accepted.
When sufficient average failure times at stress levels varying from 20% to 65% of the yield
stress are obtained, the data sets are plotted on a percent yield stress versus failure time graph on a
log-log scale. Shown in Figure 11 are the three types of general response curves resulting from
successfully completed NCTL tests.
3.2.5 BEHAVIOR OF RESPONSE
The response curves of Figure 11 clearly show at least two distinct regions. In the initial
region (i.e., at high stress levels) the test specimens respond in a ductile manner. Note that the test
can be terminated by either ductile failure (i.e., complete separation) or ductile yielding (i.e.,
elongated necking) as determined by the positioning of the mechanical lever arm. The response
can be visually seen as the test specimens elongate and neck-down in their behavior. In the
terminal region (i.e., at low stress levels) the test specimens fail in a brittle manner. This is
visually seen as the test specimens separate (i.e., they "crack") with no elongation except possibly
at the back end of the specimen where the stress on the reduced cross section becomes excessively
high. Photographs of ductile and brittle test specimens after test completion are seen in Figure 12.
28 '
-------
100 -1
Percent SO -
Yield
Stress
25 —J
10
'Knee"
r
1 10 100
Failure Time (hr.)
1000 10,000
100 —
50 -
Percent
Yield
Stress 25 —
10
o.i
(a) Ri-Linear for "Rnee'"> Response Curve
"Nose"
T
T
10
—T"
100
1000
—r~
10,000
Failure Time Response Curve
Percent
Yield
Stress
100 -
50 -
25 -
10
0.1
"Step"
1
10
I
1000
1 10 100
Failure Time (hr.)
(cl Tri-Linear (or "Step1") Response Curve
10,000
Figure 11 - Typical response ccurves resulting from a complete
notched constant tensile load (NCTL) test
29
-------
(a) Ductile behavior specimens
(b) Brittle behavior specimens
Figure 12 - Photographs of ductile and brittle behavior of NCTL
test specimens.
30
-------
The morphology of these different failure surfaces can also be viewed via scanning electron
micrographs. Shown in Figure 13(a) is a micrograph in the ductile region (showing the classic
"pullput" behavior) and in Figure 13(b) is a micrograph in the brittle region (showing the classic
"fibril" behavior). A complete discussion is available showing these different surface
morphologies and also subtle variations of the transitional morphologies/31)
The region between the ductile and brittle behavioral trends is called the "transition" region
and can either be very abrupt, have a "nose" shape or have a "step" shape. Each has precedence in
the polymer literature.(31) Different morphologies of failed specimens in this transition region are
shown in the micrographs of Figure 14. Here different surface characteristics and idiosyncrasies
can be observed.
3.2.6 ANALYSIS OF RESPONSE
At least four quantitative test parameters can be obtained from the NCTL response curves
shown in Figure 11. They are the following:
• the slope of ductile (initial) portion of the curve,
• the slope of brittle (terminal) portion of the curve,
• the transition time, and
• the transition stress level.
The two values of slope can be utilized in describing a failure time as follows:
tf~a- • (2)
where
tf = failure time (hrs.)
a = applied stress (percent of yield or actual stress level)
s = slope (either in the ductile or brittle region)
Values of "s" are typically -20 to -50 in the ductile region, i.e., indicating gentle slopes, while they
are -2 to -15 in the brittle region, i.e., indicating steep slopes.
Of significantly greater importance than the slopes, however, is the information regarding
the transition point. Shown on the behavioral curves of Figure 11 are the specific targeted points
for transition time "Tt" and transition stress "T0", respectively. As can be suspected from the
discussion presented to this point, the longer the transition time and the lower the transition stress,
the better is the HDPE geomembrane's stress crack resistance. The recommended numeric values
31
-------
pullout
fibril
(a) Surface morphology in the ductile region
(at 40 times magnification)
jullout
fibril
15KU X48
0002
DUMAT
(b) Surface morphology in the brittle region
(at 40 times magnification)
Figure 13 - Scanning electron micrographs of ductile (pullout) and
•brittle (fibril) failure surfaces resulting from the NCTL test
32
-------
pullout
fibril
notch
pullout
fibril
notch
Figure 14 - Various surface morphologies of NCTL test specimens in the
transition regin illustrating a combination of pullout and fibril type
of morphologies (both micrographs are at 40 times magnification)
33
-------
for the transition time will be given later in Section 3.5.
3.3 LABORATORY TESTING PROGRAM
A number of replicate tests and interlaboratory tests have been performed in the course of
the development of the NCTL test.
3.3.1 Reproducibilitv Tests
Using a single type of HDPE geomembrane sheet, a series of three duplicate tests were
performed at GRI to investigate initial test reproducibility. The experimental curves are seen in
Figure 15 from which the resulting test parameters are summarized in Table 5.
Table 5 - NCTL reproducibility tests
Test
No.
1
2
3
Ave.
S.D.
Ductile
Slope
(%/hr.)
-0.035
-0.046
-0.047
-0.043
15%
Brittle
Slope
(%/hr.)
-0.58
-0.37
-0.40
-0.45
25%
Transition
Time
(hrs.)
11.0
9.1
11.8
10.6
12%
Transition
Stress
(%
-------
H
s
-------
'3.3.2 Intel-laboratory Test #1
With the cooperation of four outside laboratories in addition to GRI, a geomembrane sheet
sample was sent to each for evaluation via the NCTL test. The written test method GRI GM5(a)
given in the Appendix accompanied the sheet material. The different laboratories did their own
sample preparation, notching, testing, etc. They then sent the results back for plotting and
comparison. Figure 16 shows the graphic behavior, and Table 6 compares the numeric results
from these curves for the major features under investigation.
Table 6 - NCTL interlaboratory test #1 results
Laboratory
No.
1
2
3
4
5
Ave.
S.D.
Ductile
Slope
(%/hr.)
-0.047
-0.044
-0.096
-0.037
-0.056
-0.056
42%
Brittle
Slope
(%/hr.)
-0.40
-0.39
-0.45
-0.45
-0.30
-0.40
15%
Transition
Time
(hrs.)
12
10
16
16
16
14
20%
Transition
Stress
(% ay)
35
35
35
37
28
33.4
11%
The results compare well with one another (particularly the transition time and stress) and despite
the variations inherent in different laboratories and personnel performing the tests, the results show
encouragement as to further evaluation of the test method.
36
-------
r«
ft,
bo
•S
3 >
|£
1
PI«!A
37
-------
3.3.3 Intel-laboratory Test #2
Perhaps one of the reasons for the close agreement of the previous interlaboratory tests is
that the geomembrane sheet itself had a relative low value of transition time, i.e., it was somewhat
sensitive to stress cracking. Thus a second interlaboratory test was performed which used a
geomembrane sheet that was more resistant to stress cracking, i.e., it had a longer transition time.
Samples of this second geomembrane sheet were sent to the same laboratories that had participated
previously (four outside laboratories plus GRI), who cut their own test specimens, notched and
tested them, and then sent the results back for comparison purposes. The graphic behavior is
shown in Figure 17. From these curves the significant test results are tabulated in Table 7.
Table 7 - NCTL interlaboratory test #2 results
Laboratory
No.
1
2
3
4
5
Ave.
S.D.
Ductile
Slope
(%/hr.)
-0.042
-0.075
-0.098
-0.063
-0.110
-0.078
35%
Brittle
Slope
(%/hr.)
-0.32
-0.26
-0.13
-0.12
-0.36
-0.24
45%
Transition
Time
(hrs.)
153
60
100
70
210
119
53%
Transition
Stress
(%<*y)
40
40
30
35
35
36
12%
The resulting comparison shows a significantly longer transition time for this sample by all
participating laboratories and somewhat more spread insofar as the standard deviations are
concerned. This was anticipated since the transition time is on a logarithmic scale and as it
becomes longer the variation between laboratories should also increase.
A further aspect of the test to be investigated, however, was that variations in test specimen
notching techniques could be the factor causing the high standard deviation of the transition time.
38
-------
o
o
o
o
o
H
ts I
i- CM CO «* 10
ij JS .£> JQ A
CO CO (0 (0 CB
Q
O O
-------
3.3.4 Intel-laboratory Test #3
Samples from the same geomembrane sheet just described in interlaboratory test #2 (i.e.,
the geomembrane with the longer transition time) were used in this third interlaboratory test. The
difference between interlaboratory tests #2 and #3 was that all of the test specimens were cut,
prepared and notched by one organization. The organization was not GRI, but a firm which has a
specially designed notching tool. Three outside laboratories (one withdrew from the testing
program) in addition to GRI participated and the results are shown graphically in Figure 18. From
these curves the numeric test data of Table 8 were obtained.
Table 8 - NCTL interlaboratory test #3 results
Laboratory
No.
1
2
3
4
Ave.
S.D.
Ductile
Slope
(%/hr.)
-0.071
-0.066
-0.052
-0.051
-0.060
17%
Brittle
Slope
(%/hr.)
-0.37
-0.33
-0.40
-0.23
-0.33
22%
Transition
Time
(hrs.)
43
85
100
100
82
33%
Transition
Stress
35,
32.5
35
32.5
33.7
4%
Table 8 shows that the values of standard deviation are somewhat better than the results of the
previous interlaboratory test where each laboratory did their own notching. This appears to
indicate that consistent notching is one of the important factors in obtaining reproducible results.
The importance of careful notching has been demonstrated by others in related studies of this type,
e.g., see Lu, et.al. @2\ in addition, the improvement in the standard deviation value of transition
time may also contributed by two other factors. One is that Laboratory No.l had improved their
test device. The second is the withdrawal of Laboratory No.5 which gave very different results
than the other participants in the previous interlaboratory test
40
-------
. o
JEI
I
• MM
H
I
1
SS9JJS PP!
41
-------
3.3.5 Laboratory Testing Program
Based on the reproducibility and interlaboratory test data just presented, it is felt that the
NCTL test is a candidate test method to evaluate stress cracking of HDPE geomembrane sheet
materials. The test information presented so far also suggests that of the four resulting parameters
obtained from the experimental NCTL test curves, the transition time has the greatest variability
from material to material. Thus it is perhaps the major target value for assessing variations in
resin, additive packages, manufacturing methods, etc. To determine the variation of the transition
time value (and the other three parameters as well), as many commercial samples of HDPE
geomembrane sheet as could be obtained were evaluated at GRI. This was not a controlled
experimental design, but is actually an assessment of various sheet materials that are currently
available. Table 9 presents the results of NCTL tests on eighteen (18) different commercially
available HDPE geomembranes for the four major parameters that we are considering. The entire
stress versus failure time curves from which this data was obtained is included in Appendix "B".
Figure 19 presents a graph of the above data showing the relative positioning of the
transition time and transition stress for each sample. It clearly illustrates that the transition time has
tremendous variation from material to material, i.e., the transition time varies approximately 500
times (from a low of 10 hours to a high of 5000 hours) for the 18 commercially available HDPE
geomembrane materials that were tested. Truly, the transition time parameter varies greatly from
material to material when evaluated in this particular test method!
In addition to the transition time the data shows that in thirteen of the eighteen samples, the
transition stress is less than or equal to 35%. This observed feature will be tied into the
methodology and recommendations of a single point test to be presented later.
42
-------
Table 9 - NCTL test results of various HDPE geomembrane samples
Sample
No.
1
. 2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Ductile
Slope
(%/hr.)
-0.047
-0.062
-0.055
-0.037
-0.032
-0.019
-0.047
-0.049
-0.043
-0.063
-0.052
-0.038
-0.054
-0.053
-0.061
-0.021
-0.049
-0.037
Brittle
Slope
(%/hr.)
-0.16
-0.17
-0.19
-0.32
-0.20
-0.43
-0.40
-0.50
-0.33
-0.11
-0.10
-0.33
-0.15
-0.064
-0.24
-0.39
-0.48
-0.33
Transition
Time
(hrs.)
400
70
115
30
50
70
10
5000
60
300
500
70
600
600
100
110
17
200
Transition
Stress
(%
-------
1
B
B
B
i B
B
O
O
- O
U
e
Q
** 05
§§
ON
S
SS3JJS
44
-------
3.3.6 Summary of NCTL Tests
The value of having independent laboratories do parallel evaluation for a new test method
and procedure cannot be overemphasized. Differences in equipment, personnel, environment,
devices, etc., often take a procedure, initially considered to be very consistent, and turn it into
chaos. Clearly this did not occur with the NCTL test procedure presented herein, as evidenced by
the data presented in this chapter.
• Regarding the general behavior of the resulting applied stress versus failure time curves, all
show clearly defined ductile and brittle regions. Both visual and microstructural examinations
verify and strengthen the existence of these two regions for all HDPE geomembrane sheet samples.
Regarding the transition region itself, three possible behavioral patterns are shown in Figure 11.
In fact, all three types have been observed as indicated by the NCTL test curves of Appendix "B".
From the resulting NCTL test curves, four parameters are obtained; the two slopes of the
linear portions of the curves and the two coordinates of the transition point. Of these, the transition
time and corresponding transition stress are the most significant. Between these two, the transition
time is the most important since its variation from material to material is seen to be so large, recall
Table 9 and Figure 19. For the data presented thus far, the variation in transition time is from 10 to
5000 hours, while the variation in transition stress is only from 28 to 40 percent of yield stress.
Due to the high sensitivity of transition time, this value will now be the major focus of the
study. More specifically, the question of a recommended minimum transition time will now be
addressed.
45
-------
3.4 CORRELATION BETWEEN LABORATORY AND FIELD TESTS
In order to define a minimum transition time for stress crack resistant HDPE
geomembranes, the data for the field retrieved and laboratory samples have been further analyzed
and compared to one another.
For the HDPE geomembrane lined sites where failure occurred (recall Section 2.2 and
Table 4 in which all were indeed stress crack problems), NCTL tests were performed on field
retrieved samples at GRI. For some of the field cases listed in Table 4 there was insufficient
material available to do the complete NCTL test. Recall that a minimum of 30 test specimens are
required. Nevertheless, at seven (7) of the fifteen (15) failure sites there was adequate material for
evaluation purposes. The stress versus failure time curves of those seven field retrieved
geomembranes are given in Appendix "B". From the curves, the usual four test parameters are
obtained and the data is listed in Table 10.
Table 10 - NCTL Test results on field retrieved HDPE geomembranes that failed in a
stress cracking mode
Site No.
(refer to Table 4
for description)
3
5
6
8
13
14
15
Ductile
Slope
(%/hr.)
-0.045
-0.036
-0.072
-0.058
-0.047
-0.048
-0.046
Brittle
Slope
(%/hr.)
-0.17
-0.36
-0.54
-0.42
-0.28
-0.34
-0.36
Transition
Time
(hrs.)
97
7
4
27
55
17
11
Transition
Stress
(% CTy)
35
43
41
35
30
42
42.5
The corresponding transition time and transition stress from these field retrieved samples can now
be plotted as shown in Figure 20. The data are shown as solid diamonds. Added to this same
figure are the 18 transition points from the laboratory samples which were reported in Table 9 and
on Figure 19.
If one examines all of the transition points on Figure 20 it is seen that:
• Transition times from field retrieved samples are all less than 55 hours, except for one
point which is 97 hours. The average for the seven field geomembranes evaluated was 31
hours.
• Transition times of laboratory samples are spread from 10 to 5,000 hours, but appear to
be somewhat separated by the 100 hour value in that 10 of the laboratory samples were
46
-------
B
1
a
s
tuo
-O
H
|
'3
<4H 00
o
-------
equal or higher than this value and 8 were lower.
Taking all of these test results (field and laboratory) in mind it appears as though a minimum
transition time of 100 hours can be set, with two important considerations included. First, the data
indicates that all field failures are considered, i.e., the failure times of all seven (7) field sites are
less than the 100 hour transition time. Second, that there are a number of commercially available
HDPE geomembrane products that are currently available that meet the 100 hour criterion, i.e., 10
out of 18 exceed the 100 hour transition time.
With respect to the transition stress, it is seen that three out of seven field samples have a
stress less than or equal to 35% yield stress. Also to be noted is that most of the laboratory
samples which gave greater than 100 hour transition time were also at, or below, 35% yield stress.
This suggests that brittle failure would be mostly likely to occur at applied stress less than 35%
yield stress. This finding will have significant influence on the development of a single point
NCTL test and a seam constant tensile load (SCTL) test which will be discussed in Sections 4 and
5, respectively.
3.5 NCTL TEST RECOMMENDATIONS
Based on the work presented in this section of the report, the criterion that can be
recommended as far as a transition test time obtained from a complete NCTL test on HDPE
geomembrane sheet is as follows :
• Transition time > 100 hours
The test procedure used to generate the above data must follow the GRI Standard Test Method
GM5(a) given in Appendix "A" of this report for the recommendation to be applicable.
3.6 CORRELATION BETWEEN NCTL TEST AND BENT STRIP TEST
Since beginning this report, where it was stated that the currently used test method for
qualifying a HDPE geomembrane sheet material against stress cracking is the ASTM D1693 Bent
Strip Test, no comparison with the NCTL test has been presented. Such a comparison, however,
is readily achievable since the currently recommended procedure for the Bent Strip Test utilizes
50°C incubation temperature and a 10% Igepal CO 630/90% tap water immersion medium. These
being identical values to the NCTL test conditions, allows one to place the Bent Strip Test
specimens in their channel holder (recall Figures 6 and 7) directly into the NCTL test incubation
bath. Thus any subtle variations in temperature or solution concentration are automatically
compensated for in a comparison testing program. Table 11 provides the results of such a
comparison between fourteen (14) laboratory samples and seven (7) field retrieved samples insofar
as Bent Strip Testing and NCTL testing is concerned. (Bent strip tests have not been performed on
Laboratory S amples L-15 to L-18)
48
-------
Table 11 - Comparison of ASTM D1693 bent strip test results with
notched constant tensile load (NCTL) test results
Bent Strip Test
Notched Constant Tensile Load
Test No.*
L-l
L-2
L-3
L-4
L-5
L-6
L-7
L-8
L-9
L-10
L-ll
L-12
L-13
L-14
F-3
F-5
F-6
F-8
F-13
F-14
F-15
ASTMD1693
no. failed/no, tests
at 1500 hours
0/10
0/10
0/10
0/10
0/10
0/10
10/10
0/10
0/10
0/10
0/10
0/10
0/10
0/10
0/10
10/10
10/10
1/10
0/10
10/10
7/10
NSF54
Recommendation:
pass (P) or Fail (F)
P
P
P
P
P
P
F
P
P
P
P
P
P
P
P
F
F
P
P
F
F
NCTL
transition
(hours)
' 400
70
115
30
50
70
10
5000
60
300
500
70
600
600
97
7
4
27
55
11
17
GRI Recommendation
Tt>100hrs.
pass (P) or Fail (F)
P
F
P .
F
F
F
F
P
F
P
P
F
P
P
F
F
F
F
F
F
F
*L- refers to laboratory sample as given in Table 9
F- refers to field retrieved sample as given in Table 4
The data of Table 11 shows clearly that the NCTL test is far more rigorous in assessing the
potential stress crack performance of HDPE geomembranes than is the currently used Bent Strip
Test. For example, note the following:
• 13 of 14 (93%) laboratory samples pass the Bent Strip Test, while in the same group
only 7 of 14 (50%) pass the NCTL test criteria.
• 3 of 7 (43%) field retrieved samples pass the Bent Strip Test, while in the same group 0
of 8 (0%) pass the NCTL test criteria.
The further significance of the NCTL test is clearly seen in the data of Table 11 in that every test
that failed the Bent Strip Test also failed the NCTL test recommendations. More importantly, a
large number of other samples which passed the Bent Strip Test actually failed the NCTL test.
Thus it is seen that HDPE geomembrane sheet materials which pass the Bent Strip Test but fail the
49
-------
NCTL Test criteria are lower in their stress crack resistance when compared to samples which pass
the NCTL test criteria.
In light of the above comparison of the two test methods, a further recommendation of this
study is that the ASTM D 1693, "Bent Strip Test" method be replaced as the primary qualifying test
for acceptable stress crack resistance of HDPE geomembrane sheet materials by the NCTL test as
described herein. It should continue to be a manufacturers option as to whether, or not, to use the
Bent Strip Test and report its findings along with that of the recommended NCTL test results. It
should be recognized, however, that the NCTL test is a very long and involved test thus it should
be considered to be primarily a research and development test for resin and/or geomembrane
processing development studies. Hence the development of a complimentary NCTL test is
required which is focused toward an index or qualifying test.
50
-------
SECTION 4
SINGLE POINT-NOTCHED CONSTANT TENSILE LOAD (SP-NCTL) TEST
The previously described NCTL test requires the generation of the entire applied stress
versus failure time curve. For acceptable samples with greater than 100 hour transition time, the
establishment of the brittle region of the curve can take well over 1000 hours (41+ days) to obtain.
Furthermore, if the statistical averages are not reasonably well behaved, some of the data points
must be retested. Considering such a long time frame, the thought of developing a testing protocol
based on a unique point, i.e., a single point - notched constant tensile load (SP-NCTL) test, was
investigated. This chapter describes the development of a SP-NCTL test and presents
recommendations as to the utilization of the test data.
4.1 TEST DETAILS AND PROTOCOL
The concept of a SP-NCTL test is to select a stress level near the initial portion of the brittle
region of the applied stress versus failure time curve and see that the failure time of a test specimen
exceeds a specified value. Consider the idealized curve of Figure 21 to illustrate the concept. The
transition point is taken to be located at the recommended minimum transition time value of 100
hours, and at a transition stress of 35% of the yield stress. This latter value is the transition stress
value of most of the laboratory qualified samples. Using the brittle slope of all qualified samples, a
zone of acceptability can be obtained, see Figure 21. This brittle zone is now used to illustrate the
recommendation for the single point test. Assume that the highest applied stress level that can
assure brittle behavior is 30% of the yield stress. Therefore, a straight line drawn at 30% a will
intersect the minimum and maximum boundaries of the brittle zone at 120 and 200 hours,
respectively. Any failure time in excess of the maximum value (i.e., 200 hours) would indicate
that the test specimen has a better stress cracking resistance than those which minimally qualify
through the full NCTL test. In other words, a sample with greater than 200 hour failure time in a
SP-NCTL test at 30%
-------
I
C/5,
a
00
§
3
£
O
a
•—I
I-H
o
2
'S
(In
1
1
i
"8
§
1
52
-------
is considered to be the primary QC/QA qualifying test regarding the stress crack resistance of the
material and should be performed on a more regular basis than the full NCTL test. The test method
is included in Appendix "A" of the report and is identified as GRI Test Method GM5(b). The test
can be performed on specimens taken directly from geomembrane sheet or from plaques made
from resin pellets in the manner prescribed in the Test Method. ,
Furthermore, to gain some type of statistical validity, a minimum of five (5) replicate tests
at 30% yield stress should be performed. Within this set, all five test specimens must show
performance time greater than 200 hours, i.e., there can be no failures of any of the test specimens
within 200 hours. If this SP-NCTL test criterion is not met, one needs to perform the entire NCTL
test. If the resulting transition time is 100 hours, or longer, the sheet material is acceptable under
the full NCTL test procedure and its recommendations as stated in Section 3.5 of this report. Thus
performance of the complete NCTL test takes precedence over this SP-NCTL test if a dispute
arises.
53
-------
SECTIONS
SEAM CONSTANT TENSILE LOAD (SCTL) TEST
While most of the stress crack failures observed in the field have initiated at seams (recall
Table 4), it must be recognized that stress cracking is fundamentally a materials issue rather than a
seaming issue. A properly selected resin and additive package along with proper manufacturing of
the sheet will result in a stress crack resistant material, at least to the extent of avoiding failures
such as those described in Table 4 of this Report. However, if the resin, formulation and/or
manufacturing is not adequate in this regard and stress cracking does occur, its likely initiation
point will be at a seam location. This is due to the naturally occurring discontinuity at seams (i.e.,
the necessary overlap geometry), possible excess grinding (for fillet or flat extrusion seams), and
possible excessive heat (for thermal fusion seams). Thus an effort to adapt the NCTL concept to
field seams was undertaken. This section presents the development of a seam constant tensile load
(SCTL) test and presents final recommendations as to the utilization of the test data.
5.1 TEST DETAILS AND PROTOCOL
The purpose of the SCTL test is to evaluate the quality of a HDPE geomembrane seam to
see whether it has the same stress cracking resistance as the sheet material. Direct comparison can
be made between the sheet material and seam if their stressed geometries are similar. Hence, a
notched seam specimen will be used in the proposed SCTL test.
The size of the dumbbell shaped specimen is to be according to ASTM D1822. The notch
is introduced on the sheet surface opposite to the seam junction, as shown in Figure 22. The notch
depth is such that 80% of the nominal sheet thickness remains intact. Thus the test specimen shape
and notching is the same as with the NCTL test, except now the edge of a seam is included.
The details of the testing procedure follow in Appendix "A" to this Report as GRI Test
Method GM5(c). In this procedure, five (5) replicate test specimens are utilized for each seam test
sample that is to be evaluated. The applied load is to be 30% of the yield stress of the sheet
material which is measured based on the ASTM D638 tensile tesjt performed at standard laboratory
temperature and humidity, i.e., 23°C and 50% RH. The incubation medium is 10% Igepal CO 630
and 90% tap water maintained in a constant temperature bath held at 50°C. Since the test
conditions are the same as that used in SP-NCTL test, the failure time limits of the five test
specimens should be set at the same value as the SP-NCLT test, namely 200 hours. This means
that if the seam sample meets, or exceeds, the 200 hour time limit, it should have equal stress
cracking resistance as the sheet material.
54
-------
Notch
Notch
Notch
Fillet Extrusion Flat Extrusion
Fusion
(Hot Wedge or Hot Ah-)
Figure 22 - Seam constant tensile load (SCTL) notched test specimens
55
-------
5.2 SCTL TEST RECOMMENDATIONS
For the seam constant tensile load (SCTL) test an applied stress level of 30% of yield stress
is subjected to five (5) notched test specimens as shown in Figure 22. The test temperature and
incubation medium is the same as with the NCTL and SP-NCTL test procedures. The failure time
for each test specimen must be 200 hours, or longer, regardless of the mode of the failure. The
SCLT is not meant to be a routine CQC/CQA method to assess field seaming. It is meant to be a
research and development test to evaluate variations in seaming and newly developed seaming
techniques.
56
-------
SECTION 6
RELATED CONSIDERATIONS
The type of constant tensile load test just described has great use in other applications of
polymer science and engineering. For example, Lu and Brown@3) have performed accelerated
temperature testing and lifetime prediction in HOPE pipe by using a tensile test specimen in a bar
shape which is notched on three sides. Additionally, different immersion media have been used by
researchers showing interesting failure behavior/34) In this section, some parallel data will be
reported using the NCTL test procedure.
6.1 EFFECT OF DIFFERENT TEMPERATURES
The NCTL test recommended and reported herein uses a constant temperature immersion of
50°C. Its selection is somewhat arbitrary and it is an intermediate value, below which the test takes
excessively long to perform and above which problems of high water evaporation and possible
polymer modifications begin to appear. However, to illustrate how the ductile-to-brittle behavioral
curves change with different temperatures, consider the experimental curves of Figure 23. For this
particular HDPE geomembrane sample, tests were initially performed at 50°C and were then
repeated at 40°C. Both curves show the anticipated behavior. Using data points from these two
curves in their ductile regions, and then again in their brittle regions, additional curves at different
temperatures can be analytically predicted. The prediction uses the following equation:
(3)
where
tf = failure time (hours)
a = applied stress (Mpa)
T = temperature (°K)
Ag, A! and A2 are constants
The constants A0, A j and A2 are obtained by solving three equations simultaneously.
Separate sets of constants are calculated for the ductile and brittle regions. Using the 50°C and
40°C curves of Figure 23, two sets of three simultaneous equations can be obtained, from which
the constants are calculated. For the ductile region, the constants are listed below:
57
-------
AQ = -37.18, A! = 18774 and A2 = -21.2
By substituting these constants back into Equation 3, the following equation results for predicting
the ductile portion of a curve at any desired temperature (T):
log tf=-37.18 + 18774/T - 21.2 log a/T (4)
For the brittle region, the constants are obtained similarly and are listed below:
. Ao = -11.8, Aj = 4853 and A2-2.5
Therefore, the equation for predicting the brittle portion of a curve at any desired temperature (T) is
as follows:
»
logtf=-11.8 + 4853/T-2.51ogo/T (5)
To illustrate use of the concept we have chosen 25°C as the temperature for which a new
curve is desired, and apply it into Equations 4 and 5 from which the calculated curve at 25°C is
obtained. This curve is shown by open triangles connected by dashed lines in Figure 23. To
corroborate this calculated behavior, a series of NCTL tests were performed at 25°C incubation.
The experimental results at 25°C are given as solid diamonds connected by solid lines where an
excellent agreement appears between calculated and experimental behavior. This type of
calculation, called "rate process method (RPM)" in the literature^35), is routinely used by polymer
researchers in extrapolating a limited set of test data to lower, or even higher, temperatures.
6.2 EFFECT OF DIFFERENT INCUBATION MEDIA
The NCTL test recommended in this report uses an incubation medium of 10% Igepal
(CO 630) in 90% tap water. In comparison to other fluids, this combination will be seen to be a
very aggressive incubation liquid.
Tests on the same sample of HDPE geomembrane sheet material were performed at 50°C in
three different incubation media.
• 10% Igepal (CO 630)/90% tap water
• 100% tap water
• air (i.e., no liquid immersion)
58
-------
e
o
o
o
o
i
H
i
1
59
-------
The results are given in Figure 24 and are tabulated in Table 12.
Table 12 - Effect of different immersion media on NCTL test results
Sample Ductile Brittle Transition Transition
Immersion Slope Slope Time Stress
.(%/hr.) (%/hr.) (hrs.) (% ay)
10% Igepal (CO 630) -0.047 ^O40 10 ~~35
790% tap water
100% tap water
air
-0.031
-0.030
-0.36
-1.47
400
1000
40
30
Clearly illustrated is the sensitivity of the transition time to the type of immersion media. It is seen
to vary from 10 hours to 1000 hours depending upon the type of immersion medium. Also to be
noted is that the slope of the brittle portion of the air immersion response is very steep.
6.3 OTHER TASKS IN THIS RESEARCH REPORT
\
During the course of this three year research effort, tasks other than those described
previously were pursued and evaluated. They will be briefly described in this section. It will be
noted that these efforts were generally of limited value in arriving at definitive conclusions, or they
resulted in information which is only peripherally related to the thrust of this report. Where
possible, reference will be made to the availability of additional information.
6.3.1 Density Variation in the Seamed Area
Using the ASTM D1505 density gradient method, a number of HDPE seams were
sectioned into 1 mm wide by 1 mm high by 2 mm length test specimens and measured for their
individual densities. Accuracy in this method is probably 0.0005 g/cc which allows for close
comparison of the individual pieces. By plotting these individual density results on an enlarged
view of the seam's cross section, contours of equal density were mapped for possible variations
and trends.
However, the results show a relatively small variation in density when evaluating thermal
seams (hot wedge or hot air), and extrusion seams where the density of the extrudate was the same
as the geomembrane sheets. The method did bring out those seams where the extrudate density
was less than that of the individual sheets being seamed together.
60
-------
r O
4>
H
1
1
•^
I
IB
PPIA
61
-------
6.3.2 Residual Stress Determination by Heating
An attempt was made to determine the residual stresses in HDPE geomembrane seams by
uniformly increasing the heat in a forced air oven until the polymer samples began to melt. The
idea was to photographically track the possible uneven deformations (i.e., contraction or
expansion) to qualitatively determine which production method and/or seaming method gave rise to
nonuniform behavior. It was found that if such deformations occurred, they were too small for us
to measure, hence no additional studies were undertaken.
6.3.3 Residual Stress Determination by Strain Gage Rosettes
A residual stress quantification technique used in metals and many composite materials is to
attach strain gage rosettes to the material in question, drill a hole in the center of the rosette and
monitor the relaxation of strains via the associated stress release in the material. We attempted to
duplicate this method for HDPE geomembranes and seams.
After developing the technique of attaching the strain gage rosettes to the surface of the
geomembrane and the drilling of the material under investigation to release the stress, we did obtain
some tangible results. Indications are that HDPE geomembrane sheets can have up to 10% of their
yield stress contained within the sheets in the form of residual stresses. Problems encountered
with this part of the study, however, were the very poor reproducibility of the measurements and
the nature of the results (which was often indicating an expansive residual stress). Nevertheless,
the technique could be of academic interest and a technical paper was written to describe the
methodology and our preliminary results/3^
6.3.4 Failure Surface Morphology
The different modes of HDPE geomembrane sheet and seam failures can be analyzed and
characterized by the use of a scanning electron microscope (SEM). At magnifications of 20 to 200
times, the surface patterns are very effective in performing a forensic analysis of various failure
patterns. The different morphologies revealed the following different features.
• Short Fibrous — caused by slow crack growth failure at low stresses.
• Long Fibrous — caused by slow crack growth failure at moderate stresses.
• Flake — resulting from fast, brittle failure at high stresses.
• Hackle — resulting from global plastic failure at high stresses.
• Lamellar—due to a combination of slow crack growth together with local plastic failure
at low stresses.
Further details are found in Reference (27). Note that the above fracture surface descriptions were
used in identifying the different failures listed in Table 4.
62
-------
6.3.5 Accelerated Laboratory Degradation
Several efforts were initiated in the course of this study to attempt to quantify the aging
characteristics of HDPE geomembranes. Accelerated degradation is being performed on 1.5 mm
thick HDPE geomembranes under compression and in a relaxed state. The incubation liquid is tap
water and the test temperatures are 85°, 75°, 65° and 55°C. These devices are complete at this time
and samples are being routinely removed and evaluated. The work is anticipated to take 3 to 5
years to complete. A description of the methodology of the degradation quantification and lifetime
prediction is given in Reference 37.
63
-------
SECTION 7
SUMMARY AND CONCLUSIONS
Stress cracking of polymeric materials, commonly called "environmental stress cracking" in
the polymer literature, has long been known to occur under varying conditions and to varying
degrees in a wide range of polymers. It is no surprise that it occurs in geomembranes and is most
commonly associated with semi-crystalline geomembranes. Within this category, HDPE has the
greatest crystallinity and has been the focus of this particular research and development effort.
This study has included the retrieving and subsequent analysis of several field sites-of-
opportunity which had geomembranes with some degree of stress cracking. Fifteen (15) HDPE
geomembrane lined sites were investigated, sampled and evaluated. They had a wide geographic
distribution and included a variety of resins and sheet manufacturers. All were exposed
geomembranes and most of the failures were of the slow crack growth variety. Furthermore, the
cracking generally was located at field seams. A complete forensic analysis was done for each site
wherever sufficient material for sampling could be retrieved. Both conventional Bent Strip Tests
(ASTM D1693) and the newly developed notched constant tensile load (NCTL) tests were
performed. Of significance, a number of these sites still had acceptable ASTM D1693 test results,
yet clearly showed signs of stress cracking in the field, while the NCTL test predicted all of the
failures.
The currently used ASTM D1693 Bent Strip Test was reviewed, its deficiencies noted and
it has been compared to the proposed Notched Constant Tensile Load (NCTL) test. Clearly, the
Bent Strip Test (ASTM D1693) is seen to be deficient in such a comparison. As a result, it is
recommended to use the NCTL to assess the stress cracking resistance of HDPE geomembranes.
This decision is based upon the following reasons:
• The NCTL test is reproducible to ± 15% of the mean value in its transition time and
transition stress when evaluated within a single laboratory.
• Various interlaboratory evaluations have been encouraging.
• Correlation to the field retrieved samples which actually had stress cracks is excellent.
• Testing on a variety of commercially available HDPE geomembrane samples gives a wide
range from poor to acceptable stress cracking performance. However, Bent Strip Testing
shows all of them to be acceptable.
The essence of a NCTL test is to notch a series of ASTM D1822 dumbbell shaped test
specimens from the same material, place them under various stress levels from 20% to 65% of
64
-------
yield stress in a 50°C incubation bath which contains an aggressive wetting agent, and noting the
times to a limiting deformation or failure. Plotting of these times at the different stress levels
evaluated allows for the generation of a piecewise linear ductile-and-brittle curve when plotted on a
log-log axis. The onset of the brittle portion of the response curve is critical. This point defines
the transition failure time and its associated transition stress level. Based on testing of both field
retrieved samples and laboratory samples, the acceptable criteria of an HOPE geomembrane when
tested in a NCTL test is for the transition time to be equal or greater than 100 hours. The
conducting of such a complete NCLT test is suggested as being a research and development tool
for assessing new, or different resins, additives and processing variables.
Since the complete NCTL test curve is time consuming to perform, a single point test,
designated as SP-NCTL has been developed. In this test the evaluation of five test specimens,
each of which is subjected to 30% yield stress, is recommended. None of the test specimens
should fail within 200 hours. The criterion is slightly more conservative than the NCTL test
criterion stated above. However, the test takes much less time and effort to perform, and should
be used as a quality control and/or quality assurance test in assessing resin shipments and finished
sheet geomembrane. If the test is used as QC/QA for checking the resin pellets, a plaque (i.e., cast
sheet) is first necessary to make, from which test specimens can be taken.
Since field seams are often associated with stress cracking, a seam constant tension load
test, designated as SCTL, has also been developed. In this test, five seam test specimens are
notched on the opposite side of the seam (i.e., on the surface of the geomembrane facing
downward in the field) at the edge of the seam. The five test specimens are placed under 30%
yield stress of the sheet in a constant temperature bath at 50°C containing 10% Igepal (CO 630) and
90% tap water. As with the single point test, none of the specimens should fail within 200 hours.
The test is suggested as being a research and development tool for assessing new, or different,
seaming methods.
Other variations of NCTL testing were also explored. The. effect of different incubation
temperatures (and subsequent prediction of the anticipated response to untested temperature
responses) and of different immersion media were also evaluated. Both variations strongly
suggest that the NCTL test must be conducted according to a very prescribed testing procedure. A
GRI Test Standard has been developed for the NCTL, SP-NCTL and SCTL tests and should serve
until such time that ASTM adopts (and/or adapts) the procedure for its own Standards. The current
testing protocol for these tests are included as an Appendix "A"to this report and are designated as
GRI Test Methods GM5(a), GM5(b) and GM5(c), respectively.
65
-------
REFERENCES
1. Koerner, R. M., Lord, A. E., Jr. and Halse-Hsuan, Y. Degradation of Polymeric Materials
and Properties. In : Proc. Inti. Symp. on Research Developments for Improving Solid Waste
Management, U.S. EPA, Cincinnati, OH, 1991.
2. , ASTM D883 - Definitions of Terms Relating to Plastics. Vol. 08.01.
3. Choi, S. W. and Broutman, L. J. Ductile-Brittle Transitions for Polyethylene Pipe Grade
Resins. Eleventh Plastic Fuel Gas Pipe Symposium, San Francisco, CA, 1989, pp. 296-
320.
4. Chan, M. K. V. and Williams, J. G. Slow Stable Crack Growth in High Density
Polyethylenes. Polymer, Vol. 24, February, 1983, pp. 234-244.
5. Lu, X. and Brown, M. Unification of Ductile Failure and Slow Crack Growth in an
Ethylene-Octene Copolymer. J. of Materials Science, 26, No. 3, 1992, pp. 612-620.
6. Lustiger, A. and Rosenberg, J. Predicting the Service Life of Polyethylene in Engineering
Applications. Li: GRI-Seminar n, Durability and Aging of Geosynineties. Ed. by Koerner,
R. M., Published by Elsevier Applied Science, 1988, pp. 212-229.
7. Keith, H. D., Padden, F. J. and Vadimsky, R. G. Intercrystalline Links: Critical Evaluations.
Jour, of Appl. Physics, Vol. 42, No. 12, 1971, pp. 4585-4592.
8. , Plastic Pipe Line, published by Gas Research Institute, Vol. 3, No. 2, Dec. 1982.
9. Williams, J.G. Fracture Mechanics of Polymers. John Wiley and Sons, New York, 1987.
10. , Inspection Techniques for the Fabrication of Geomembrane Field Seams. EPA/530/
SW-91/051, Technical Guidance Document, U.S. EPA, Cincinnati, OH, May 1991, 174 pgs.
11. , ASTM D1693. Test Method for Environmental Stress-Cracking of Ethylene Plastics.
\bl. 08.02.
12. , Lining of Waste Containment and Other Impoundment Facilities. EPA/600/2-88/052.
13. , Field Failure Reference Catalog for Polyethylene Gas Piping. Gas Research Institute-
84/0235.1, 1st Edition, Jan. 1980-Dec. 1984.
14. , ISO/TC 138. Plastic Pipes and Fittings for Transport of Fluids. Working Group 5,
Test Methods and Specifications, Document 747, "Standard Extrapolation Methods".
15. , ASTM D1598, Test Method for Time-to-Failure of Plastic Pipe Under Constant
Internal Pressure. 08.04. .
66
-------
16. Telles, R. W., Lubowitz, H. R. and Unger, S. L. Assessment of Environmental Stress
Corrosion of Polyethylene Liners in Landfills and Impoundments. Report on Contract 68-
03-3218, U.S. EPA, 1984, Cincinnati, OH, 89 pgs.
17.' , ASTM D2552. Test Method for Environmental Stress Rupture of Type III
Polyethylenes Under Constant Tensile Load. 08.02 (Discontinued 1986).
18. Fruch, B., Dixon, P. and Hunt, L. W. Review and Development of Physical Test Methods
for Polyethylene Geomembranes. Alberta Environmental Research Trust, Nov. 1986, Final
Report by Hanson Materials Engineering, Edmonton, Alberta, 150 pgs. N
19. Koch, R. Gaube, E., Hersel, J., Gondro, Ph.C. and Heil, H. Long-Term Resistance of
Dump Sealing Sheets of Polyethylene. Report TR-88-0051, 0054-0054, Translated by
Sitran, Santa Barbara, CA, 1988.
20. Fisher, G. E. Controlling Thermal Damage in Flexible Membrane Liners. Geotechnical
Fabrics Report, IFAI, Vol. 7, No. 2, March/April, 1989, pp. 39-41.
21. Peggs, I. D. and Carlson, D. S. Stress Cracking of Polyethylene Geomembrane Seams:
Field Experience. In : GRI-Seminar n, Durability and Aging of Geosynthetics., R. M.
Koerner Editor, Elsevier Science Publishers Ltd., London, 1989, pp. 195-211.
22. Peggs, I. D. Failure and Remediation of a Geomembrane Lining System. Geotechnical
Fabrics Report, Vol. 6, No. 6, November, 1988, pp. 13-16.
23. , Preliminary Assessment of Stress Cracking of Polyethylene Liners. Draft Report to
U.S. EPA, Task 6, Work Assignment 68-3, Under NUS Corp., by GeoServices, Inc., 17
March 1989, 27 pgs. (GeoServices, Inc., is now GeoSyntec, Inc.,)
24. , ASTM D1505. Test Method for Density of Plastics by the Density-Gradient Technique.
08.01.
25. , ASTM D1603. Test Method for Carbon Black in Olefin Plastics. 08.02.
26. , ASTM D1238. Test Method for Flow Rates of Thermoplastics by Extrusion
Plastometer. 08.01.
27. Halse, Y, Lord, A. E., Jr. and Koerner, R. M. Stress Cracking Morphology of HOPE
Geomembrane Seams. I. D. Peggs Editor, ASTM, STP 1076, 1990, pp. 78-89.
28. Lander, L. L. Environmental Stress Rupture of Polyethylene. SPE Journal, December,
1960, pp. 1329-1332.
29. , ASTM D1822. Test Method for Tensile Impact Energy to Break Plastics and Electrical
Insulating Materials. 08.02.
30. ,ASTMD638. Test Method for Tensile Properties. 08.01.
31. Halse, Y H., Lord, A. E., Jr. and Koerner, R. M. Ductile-to-Brittle Transition Time in
Polyethylene Geomembrane Sheet. Geosynthetic Testing for Waste Containment
Applications, ASTM STP 1081, Koerner, R. M. Editor, ASTM, Philadelphia, PA 19104,
1990, pp. 95-109.
67
-------
32. Lu, X., Qian, R. and Brown, N. Notchology - The Effect of the Notching Method on the
Slow Crack Growth Failure in a Tough Polyethylene. Journal of Materials Science, Vol. 26,
No. 4, 1991, pp. 881-888.
33. Lu, X. C., Brown, N. The Ductile-Brittle Transition in Polyethylene Copolymer. Journal of
Materials Science, Vol. 25, No. 1 A, 1990, pp. 29-34.
34. Shanahan, M. E. R. and Schultz, J. Environmental Stress Cracking of Polyethylene: Criteria
for Liquid Efficiency. J. of Polymer Sci., Polymer Physics Edition, Vol. 17, 1979, pp. 705-
713.
35. Bragaw, C. G. Service Rating of Polyethylene Piping Systems by the Rate Process Method.
Eighth Plastic Fuel Gas Pipe Symposium, 1983, pp. 40-47.
36. Lord, A. E., Jr., Koerner, R. M. and Wayne, M. H. Residual Stress Measurements in
Geomembrane Sheets and Seams. Proceedings Geosynthetics '91, IFAI, St. Paul, MN,
1991, pp. 333-350.
37. Koerner, R. M., Lord, A. E., Jr. and Hsuan, Y. Degradation of Polymeric Materials and
Products. Proc. Intl. Symp. on Research Developments for Improving Solid Waste
Management," Cincinnati, OH, EPA, Feb. '91, pp. 1-11.
68
-------
APPENDIX "A"
• GM5(a) - NCTL Test
GM5(b) - SP-NCTL Test
• GM5(c) - SCTL Test
69
-------
GEOSYNTHETIC RESEARCH INSTITUTE
Drexel University
West Wing - Rush Bldg. #10
Philadelphia, PA 19104
CGeogrldi Q GeocompwitM
-**-
(Revised My 31,1992)
GRI Test Method GM5(a)
Standard Test Method for
I i
"Notched Constant Tensile Load (NCTL) Test for Polyolefin Resins or Geomembranes"
1. Scope
1.1 This test method is used to develop data from which one can evaluate the
susceptibility of polyolefin resins or geomembrane sheet materials to stress cracking under a
constant tensile load condition and an accelerated environmental condition.
1.2 This test method measures the limiting deformation or failure time associated with a
given test specimen at a specified tensile load level. Results from a series of such tests utilizing a
range of load levels can be used to construct an applied stress vs. failure time plot on a log-log
axis. This test is called a notched constant tensile load (NCTL) test and is designated as GM5(a).
1.3 Closely related to the NCTL test is a test protocol to evaluate resin or sheet material at
a single specific value of stress. This test is called a single point - notched constant tensile load
(SP - NCTL) test and is designated as GM5(b).
1.4 Related to the SP-NCTL test is a test protocol to evaluate geomembrane seams at a
single specific value of stress. This test is called a seam constant tensile load (SCTL) test and is
designated as GM5(c).
1.5 The values stated in SI units are to be regarded as the standard. The values stated in
parentheses are provided for information only.
1.6 This standard may involve hazardous materials, operations and equipment. This
standard does not purport to address all of the safety problems associated with its use. It is the
responsibility of the user of this standard to establish appropriate health and safety practices and
determine the applicability of regulatory limitations prior to use.
70
-------
2. Referenced Documents
2.1 ASTM Standards:
D 638 Standard Test Method for Tensile Properties of Plastics
D 883 Definitions of Terms Relating to Plastics
D 1822 Test Method for Tensile-Impact Energy to Break Plastics and Electrical
Insulating Materials
D 4354 Practice for Sampling of Geosynthetics for Testing
D4433 Terminology for Geosynthetics
D4491 Test Methods for Water Permeability of Geotextiles by Permittivity
3 Definitions
3.1 geomembrane, h - an essentially impermeable geosynthetic composed of one or
more synthetic sheets (ASTM D 4833).
3.1.1 Discussion - In geotechnical engineering, "essentially impermeable" means that
no measurable liquid flows through a geosynthetic when tested in accordance
with ASTM D 4491.
3.2 stress crack, n - an external or internal crack in a plastic caused by tensile
stresses less than its short-time mechanical strength (ASTM D 883).
3.2.1 Discussion - The development of such cracks is frequently accelerated by the
environment to which the plastic is exposed. The stresses which cause cracking
may be present internally or externally or may be combination of these stresses.
4. Summary of Method
4.1 The method consists of subjecting a dumbbell shaped notched test specimen from a
polyolefin sheet to a constant tensile load in the presence of a surface-active agent and at an
elevated temperature. The time to a limiting deformation or failure of the test specimen is
recorded. The results of a series of such tests conducted at different stress levels are presented
by plotting stress level against average test or failure time for each stress level on a log-log axis.
5. Significance and Use
5.1 The test method does not purport to interpret the resulting response curve. Such
interpretation is left to the parties involved in the commissioning and reporting of the test results.
5.2 This method is intended as an index test and may be used hi a research manner for
evaluating polyolefin resin and geomembrane sheet processing in regard to stress crack
sensitivity.
5.2.1 Conditions which can affect stress cracking include: level of loading, test
temperature and environment, microstructure, polymer additive package,
processing history, and thermal history.
5.3 The single point test procedure designated as GRI Test Method GM5(b) may be used
as a quality control or quality assurance procedure to rapidly compare the performance of a
71
-------
plaque made from the candidate resin or acceptance of a geomembrane sheet material.
5.4 The seam constant tension load test procedure designated as GRI Test Method
GM5(c) may be used as a research tool to investigate the stress creak susceptibility of different,
or modified, field seaming methods.
6. Apparatus
6.1 Blanking Die - A die suitable for cutting test specimens to the dimensions and
tolerances shown in Figure 1 which is in accordance with ASTM D 1822.
Note 1 - The length of the specimen can be changed to suit the design of the test
apparatus. However, there should be a constant neck section with length at least 13
mm (0.5 inch) long. The width should be 3.17 mm (0.125 inch).
Note 2 - The tab widths may be enlarged to accommodate grommets of different
sizes with which to attach hooks for the purpose of loading.
6.2 Notching Device - A device or machine which can produce a consistent notch depth1.
Note 3 - An evaluation of the depth produced by the notching technique can be
performed by quenching a notched specimen in liquid nitrogen and then fracturing
it. The notch depth can readily be measured by examining the fracture surface
under a reflected light microscope.
6.3 Stress Cracking Apparatus - Equipment suitable for subjecting multiple test
specimens to a tensile force of up to 13.8 MPa (2000 lb/in2). The specimens shall be maintained
at a constant temperature of 50°C ± 1 °C while being totally immersed in a surface-active agent.
The solution should be constantly agitated to provide an uniform concentration throughout the
bath.
Note 4 - The apparatus2 shown in Figure 2 is one type which has been used and is
capable of testing up to 20 specimens at a time. This equipment uses a lever system
with a mechanical advantage (MA) of three to impose the desired loading on each
specimen. The surface-active agent in which the specimens are immersed is
contained in an open stainless steel tank. A submersion heater and controller are
used to maintain the test temperature. A pump keeps the fluid in a constant state of
agitation. A timing clock for each test specimen is also provided to record
automatically the termination test time of the specimens to the nearest 0.1 hour.
Note 5 - If on/off switches are used to control the timing clock, the switch must be
sensitive enough to be turned off under 200 g of force.
6.4 Compression Molding Press - The press should consist of two platens that can be
heated to at least 200 °C and also are capable to mobilize a force up to 222 KN (50,000 Ib).
1. Notching equipment is available through REMCO Industrial Machine Co., Manville, NJ 08835
2. This equipment is available through Custom Scientific Instruments, Co., Cedar Knolls, NJ 07927,
andBT Technology Inc., 613 W. Clinton St Rushville, IL 62681
72
-------
25.00
60.00
.50
8 75
^
r i
©.
• — '
^
12.70
.
Figure I - Dimensions of the D 1822 type "L" specimens
(dimensioned in mm to an accuracy of 0.02 mm)
Micro Svitchto
Timers
Pin
Joi nts
Shot
Can
760mm
20 Positions {30")
Specimen
SIDE VIEW
"Traymovedupand
dow on rack end
pinion arrangement
FRONT VIEW
O=CN
Figure 2 - Constant stress loading apparatus consisting of twenty specimen
test positions (the number of positions in the test frame is optional)
73
-------
6.5 Compression Molding Die - A single cavity square frame mold with area not less than
0.032 m2 (49 in2) and depth of 20 mm (0.79 inch). The entire steel mold should be a polished
finish without any surface defects. It should also be strong enough to resist warping or
distortion under the stated molding conditions.
Note 6 - There should be a hole in the upper and lower plates of the mold for
placing of a thermocouple in order to monitor the temperature of the plates.
7. Sampling
7.1 Lot Sample - Divide the product into lots and take the lot sample as directed in
Practice D 4354.
7.2 Laboratory Sample - As a laboratory sample, lake a full-width swatch approximately
1 m (40 in) long in the machine direction from each roll in the lot sample. The sample may be
taken from the end portion of a roll provided there is no evidence that it is distorted or different
from other portions of the roll.
7.3 Test Specimens - At least thirty test specimens are cut from each swatch in the
laboratory sample. All specimens must taken from one direction for each set of tests.
Note 7 - Quite often the test is required to challenge the weakest direction of the
sheet material. If this is the cross machine direction, the test specimens should be
cut in this direction. Hence the notch is placed in the machine direction so that the
specimens are stressed in the desired cross machine direction.
Note 8 - It has been found that inserting a grommet or eyelet in the two holes at the
end tabs of the test specimen helps to reduce the number of "grip failures" or
failures occurring outside of the neck section of the specimen.
Note 9 - If specimens larger than ASTM D 1822 type are used in the test, extra care
must be exercised to insure that the specimen is properly aligned to avoid load
eccentricity.
8. Reagent
8.1 The reagent should consist of 10% surface-active agent with 90% tap water. The
surface-active agent is Igepal CO-6301 which is a nonylphenoxy poly(ethylene-oxy)ethanol. The
reagent should be stored in a closed container. The reagent in the bath should be replaced every
two weeks to maintain a constant concentration.
Note 10 - In case of dispute, the water should be either distilled or deionized which
is at the discretion of the parties involved.
Note 11 - Other incubation solutions may also be used in the test, provided that the
parties involved mutually agree to the changes and state the specific details in the
final report.
1. Igepal CO-630 may be obtained from Rhone-Poulenc, CN 7500, Prospect Plains Road, Cranbury,
NJ 08512-7500
74
-------
9. Compression Molding Procedure
9.1 Weigh out the amount of resin pellets necessary to fill the volume of the mold based
on the density of the resin and a thickness of 2 mm.
9.2 Place the weighed resin pellets into the mold and place the mold into the press. Once
the resin reaches 160 °C, a pressure of 3.5 MPa (507 lb/in2) is alternately exerted and released
for three cycles. The same process is to be repeated at 180 °C. Finally the mass is solidified
under a pressure of 1.1 MPa (156 lb/in2) and cooled slowly from 180 °C to room temperature.
Note 12 - The mold should be suitably coated or covered so that the finished sample
is not damaged during its removal.
10. Procedure
10.1 Measure the thickness of each individual test specimen at its minimum cross section
to the nearest 0.01 mm (0.0005 inch).
10.2 Cut a controlled depth imperfection (notch) into each test specimen on one surface
as shown in Figure 3. The depth of the notch should produce a hinge thickness of 80% of the
nominal thickness of the specimen.
Note 13 - Using this procedure the actual notch depth will vary in accordance with
the actual thickness of the test specimen. For example, a sheet of nominal thickness
of 80 mil might have thicknesses ranging from 78 to 82 mil. To obtain a constant
hinge thickness of 64 mil, the notch depth would .vary from 14 to 18 mil,
depending upon on the actual thickness of the individual test specimen.
10.3 Conditioning the specimens in ice water for 30 minutes before notching can help
produce a consistent notch depth. This procedure is particularly useful for low density
polyolefin materials.
10.4 Inspect the edge of the blade for scratches and burrs under normal vision prior to
the cut. No single blade shall be used for notching more than 20 test specimens.
10.5 Test specimens are loaded at various percentages of their room temperature yield
stress. The applied stress levels should range from approximately 65% to 20% at maximum
increments of 5%. Three replicate specimens are tested at each stress level to produce a
statistically significant result
Note 14 - Tq develop the entire curve in a single direction at the recommended
values listed above will require 10 increments at 3 specimens each, or 30 individual
tests. If both directions are to be challenged, the entire test will require twice as
many test specimens.
10.6 The yield stress of the material should be measured according to ASTM D 638
(Type IV) in the direction which the specimens are to be tested.
10.7 Calculate the tensile stress to be applied to each individual specimen from the
equation given below:
75
-------
o
t = ligament thickness
L =0.80t
• Notch
t = nominal thickness
of geomembrane
\
O
Front view of the cut test specimen
Notch
0.801
60 mm (2.4")
Side view of the test specimen
Figure 3 - Front and side views of the notched test specimen of NCTL test
76
-------
Force = (A) (cry ) (W) (t) (1/M.A.)
where:
Force = force to be applied to achieve a specific percentage of the yield stress (N)
A = percentage of yield stress to be achieved (as a ratio).
ay = the yield stress of the material at room temperature, N/m2 (lb/in2).
W = width of the neck of the test specimen (3.17 mm (0.125 in) for ASTM
D 1822 dumbbell specimens).
t = the hinge thickness of the test specimen, i.e. recommended to be 80% of the
nominal thickness, m (in).
M. A. = mechanical advantage of the test apparatus, which is 3.0 for the apparatus
shown in Figure 2.
10.8 Fill the test bath in its lowered position with reagent, and adjust the temperature to
50 ± 1°C.
Note 15 - Other temperatures may be used when conducting this test. However, it
must be mutually agreed upon by parties involved and the test temperature must be
stated in the final report.
10.9 Attach the test specimens to the hooks of the test apparatus.
10.10 Adjust the distance between the lever arm and the switch to a dimension equal to 20
mm (0.75 in) for test specimens of the length indicated in this test standard.
Note 16 - For samples longer than those produced by an ASTM D 1822 die, the
distance can be changed proportionally.
10.11 Immerse the test specimens and allow temperature equilibrium to be reached. The
minimum time is 30 minutes.
10.12 Prepare the weight of lead shot (or other) required for each individual test specimen.
10.13 Load each individual specimen with its respective weight and record the elapsed
time to the limiting deflection or failure to the nearest 0.1 hour.
Note 17 - Expanded polystyrene or other types of insulation can be placed on top of
the liquid to minimize the evaporation of water and oxidation of liquid.
Note 18 - The liquid level in the bath can be maintained by using an automatic water
feeder. This will minimize fluctuation of the reagent concentration.
11. Presentation of Results
11.1 The test data should be presented in graphic form by plotting the logarithm of
percentage yield stress versus the logarithm of the average test or failure time for each stress
level. Three possible types of curves can result, see Figure 4.
11.2 Identification of the various portions of the resulting curves shown in Figure 4 are not
interpreted in this Test Standard, nor are recommended limiting performance values.
77
-------
100
Percent 50 -
Yield
Stress 25-
10
"Knee"
0.1
1 10 100
Failure Time (hr.)
1000 10,000
(a) A Bi-Linear (or "Knee") Response Curve
100 ~
Percent
Yield
Stress
50 ~
25-
10
"Nose"
0.1
10 100
Failure Time (hr.)
1000 10,000
(b) A Overshoot (or "Nose") Response Curve
100
Percent
Yield 25 _
Stress
10
0.1
1
10 100
Failure Time (hr.)
(c) A Tri-Linear (or "Step") Response Curve
1000 10,000
Fig. 4 - Possible response of curves resulting from a complete
notched constant tensile load (NCTL) test
78
-------
12. Report
12.1 The report shall include the following:
12.1.1 Complete identification of material tested.
12.1.2 Method of conditioning used in both molding and in conducting the
test, if different from that specified herein.
12.1.3 A plot of the percentage yield stress against test or failure time on log-
log axes.
79
-------
GEOSYNTHETIC RESEARCH INSTITUTE
Drexel University
West Wing - Rush Bldg. #10
Philadelphia, PA 19104
(215)895-2343
ubc (215)895-2785
F« 015)895-1437
(Revised July 31,1992)
GRI Test Method GM5(b)
Standard Test Method for
"Single Point Notched Constant Tensile Load (SP - NCTL) Test
for Polyolefin Resin or Geomembranes"
1. Scope
1.1 This method can be used as a quality control (QC) test or quality assurance (QA) test
on polyolefin resin or finished geomembrane. Its intent is to obtain an indication of the
sensitivity of the particular resin or final geomembrane sheet to stress cracking without
performing the complete series of tests as described in the NCTL test of Test Method GM5(a).
2. Summary of Method
2.1 The method includes the compression molding of an 2 mm (0.08 inch) thick plaque
made from the proposed polyolefin resin pellets. Five test specimens are to be cut out of the
molded plaque and notched as previously described. The notched specimens are all subjected to
the same value of constant tensile load while immersed in a surface-active agent at an elevated
temperature. The time to failure of each test specimen is recorded.
2.2 The method can also be utilized directly on finished geomembrane sheet in an
abbreviated form of the NCTL test, see GRI Test Method GM5(a).
2.3 The test method does not purport to interpret the resulting value. Such interpretation
is left to the parties involved in the commissioning and reporting of the test results.
3. Significance and Use
3.1 This method can be used for quality control evaluation of polyolefin resins in regard
to stress cracking as well as an abbreviated form of the NCTL test on finished geomembrane
sheet samples.
80
-------
4. Apparatus
4.1 The blanking die, notching device, and stress cracking test apparatus are the same as
those described in the previous Standard, see Sections 6.1,6.2, and 6.3, respectively.
4.2 Compression Molding Press - The press should consist of two platens that can be
heated to at least 200 °C and also are capable to mobilize a force up to 222 KN (50,000 Ib).
4.3 Compression Molding Die - A single cavity square frame mold with area not less than
0.032 m2 (49 in2) and depth of 20 mm (0.79 inch). The entire steel mold should be a polished
finish without any surface defects. It should also be strong enough to resist warping or
distortion under the stated molding conditions.
Note 1 - There should be a hole in the upper and lower plates of the mold for
placing of a thermocouple in order to monitor the temperature of the plates.
5. Compression Molding Procedure
5.1 Weigh out the amount of resin pellets necessary to fill the volume of the mold based
on the density of the resin and a thickness of 2 mm.
5.2 Place the weighed resin pellets into the mold and place the mold into the press. Once
the resin reaches 160 °C, a pressure of 3.5 MPa (507 lb/in2) is alternately exerted and released
for three cycles. The same process is to be repeated at 180 °C. Finally the mass is solidified
under a pressure of 1.1 MPa (156 lb/in2) and cooled slowly from 180 °C to room temperature.
Note 2 - The mold should be suitably coated or covered so that the finished sample
is not damaged during its removal.
6. Test Specimens
6.1 Five specimens are cut from the molded plaque in random directions using an ASTM
D 1822 die. No specimens can be taken within 25 mm (1.0 in) from the edges of the plaque.
6.2 Test specimens taken from the finished geomembrane sheet should follow Section 7
of the NCTL Test Method designated as GRI GM5(a).
7. Reagent
7.1 The reagent used in this test should be identical as that described previously in
Section 8.1 of the NCTL Test Method designated as GRI GM5(a).
8. Procedure
8.1 The test specimen notching procedure should follow Sections 10.1 to 10.4 of the
NCTL Test Method designed as GRI GM5(a).
8.2 All test specimens are stressed at a single value which is 30% of the yield stress of the
material at room temperature.
81
-------
Note 3 - The yield stress should be evaluated according to ASTM D 638 (Type IV).
Three specimens should be cut from the molded plaque in random directions and
tested accordingly. The average value is then used for the calculations. If sheet
geomembrane is being used, test specimens must be taken in the same direction that
the SP-NCTL test specimens will be evaluated.
8.3 The calculation of the actual applied force is according to the equation in Section
10.7, of the NCTL Test Method designated as GRI GM5(a).
8.4 Condition and loading the test specimens should follow Sections 10.8 to 10.13, of
the NCTL Test Method designated as GRI GM5(a).
9. Report
9.1 The report shall include the following:
9.1.1 Complete identification of material tested.
9.1.2 Method of conditioning used in both molding and in conducting the test, if it
is different from that specified herein.
9.1.3 The test or failure times of the five test specimens evaluated.
82
-------
GEOSYNTHETIC RESEARCH INSTITUTE
Drexel University
West Wing - Rush Bldg. #10
Philadelphia, PA 19104 A
(215) 895-23)3
life (215)895-2785
Fax: (215) 895-1437
(Revised July 31,1992)
GRI Test Method GM5(c)
Standard Test Method for -
"Seam Constant Tensile Load (SCTL) Test for Polyolefin Geomembrane Seams"
1. Scope
1.1 This method can be used as a research and development test on polyolefin
geomembrane seams. Its intent is to obtain an indication of the sensitivity of the geomembrane
seam to stress cracking.
2. Summary of Method
2.1 The method includes fabricating or retrieving seam samples from field sites. Any
seam type can be utilized, e.g., fillet extrusion, flat extrusion, hot wedge or hot air. Five test
specimens are to be cut out from the seam and notched. The notched specimens are all subjected
to the same value of constant tensile load while immersed in a surface-active agent at an elevated
temperature. The time to failure of each test specimen is recorded.
2.2 The test method does not purport to interpret the resulting test values. Such
interpretation is left to the parties involved in the commissioning and reporting of the test results.
3. Significance and Use
3.1 This method can be used as a research and development test in the evaluation of
modified procedures or new methods for fabrication of polyolefin geomembrane seams in regard
to their stress cracking sensitivity.
4. Apparatus
4.1 The blanking die, notching device, and stress cracking test apparatus are the same as
those described in the NCTL Test Method GM5(a), see Sections 6.1, 6.2, and 6.3, respectively.
83
-------
5. Seam Samples
5.1 Seam samples can be specially fabricated for investigative purposes or can be taken
directly from the field. Furthermore, they can be sampled from test strips or be part of the
production seam test samples for investigative purposes.
6. Test Specimens
6.1 Five specimens are to be died out of the seam sample. The edge of the seam should
be located at the center of the neck section of the specimen, i.e., the test specimens are taken in
the direction perpendicular to the seam. The general configuration can be seen in Figure 1.
7. Reagent
7.1 The reagent used in this test should be identical as that described previously in the
NCTL Test Method GM5(a) in Section 8.1.
8. Procedure
8.1 Locate the edge of the seam under a light microscope with magnification of lOx and
mark it with a white pen. Extend the marked line around all four edges of the specimen.
8.2 Measure the thickness of the single sheet adjacent to the seam.
8.3 Cut into each specimen a control imperfection, i.e., a notch. The notch should be
located right at the marked location (i.e. edge of the seam) but on the surface opposite to the seam
overlap, as shown in Figure 2. The depth of the notch should be 20% of the nominal sheet
thickness. As with the GRI GM5(a) Test Method, the intent is to have an 80% hinge of un-
notched sheet thickness remaining for the test.
8.4 All test specimens are loaded at a single stress level of 30% of the yield stress of the
sheet material at standard temperature and humidity.
Note 1 - The yield stress of the sheet material should be evaluated according to
ASTM D 638 (Type IV). Three specimens should be cut from the geomembrane
sheet at the area adjacent to the seam, and they should be taken in the direction
parallel to the SCTL test specimen
8.5 The calculation of the actual applied force is according to the equation given in GRI
GM5(a) Test Method in Section 10.7.
Note 2 - The hinge thickness (t) should be 80% of the nominal sheet thickness
8.6 Condition and loading the test specimens should follow Sections 10.8 to 10.13 of the
NCTL Test Method of GRI Standard GM5(a).
9. Report
9.1 The report shall include the following:
9.1.1 Complete identification of material tested.
9.1.2 Method of conditioning used in conducting the test if it is different from that
specified herein.
84
-------
Fillet Extrusion
Flat Extrusion Fusion
(Hot Wedge or Hot Air)
Figure 1 - The configuration of the test specimens of
seam constant tensile load (SCTL) test
85
-------
Notch
Notch
Notch
Fillet Extrusion
Flat Extrusion Fusion
(Hot Wedge or Hot Air)
Figure 2 - Seam constant tensile load (SCTL) notched test specimens
86
-------
APPENDIX "B"
Notched Constant Tension Load (NCTL) Curves of
Eighteen Laboratory Samples
and
Seven Field Samples
Generated and Analyzed in This Report
87
-------
ON
6
o>
"5.
CO
,2
(B
"S
Q
i=
2
889.118 PI9IA »USOJ9d
a.
I
o
(Q
•s
(Q
o
o
o
o
ssaj;s
SS9JJS PI9IA JU90J3d
88
-------
(O
6
.2
a
f
o
I
8
0>
o §
PI»!A
Q.
i
o
1
(0
•5
§
o
o
o
o
o
w
0 .i
" 1
i
PI9IA
89
-------
V
Q.
i?
I
3
5
CB
8
: o
CM
d
£
Q,
I
o
A
3
o
IB
SS8JJS
8
ssaj)s
O)
6
«
a
i
o
CD
§
0>
i=
I
i
i=
8
PI9IA
PI9IA
90
-------
I
O,
OT
O
CO
o ^
§
o
o
o
0>
o E
i- H
0)
)U80J8d
ssejjs PI9IA
a
a
(O
£>
o
2
o
o
a
O £
*" a>
el
a.
o
I
o •£"
0 £
" «
H
= £
ssans PI»!A
91
-------
00
6
a.
I
o £,
1
o
i
Q
I
I
^
o
I
^
2
•5
(0
8
e
0)
§
92
-------
CO
2
I
§
PI»!A
eo
6
I
(0
2
E
"5
a
§
ro|
2
£
SS8JIS
SS3JIS
94
-------
a.
2
s
F
I
,8
2
o
a
«
8
• o
0
o E
SS3J)S PI9IA
at
6
to
a.
o
JO
2
•5
a>
a.
8
p
£
PI9IA ;uaoj8d
ssajjs
90
-------
•5
a
§
- O JC
i
tO
«
o.
i
£
o
o
o
o
0)
o E
i- H
£
J3
'(5
s
luaoJ9d
PIS!A
&
Q.
i
o
S
s
. o
m
d
.
i
£•
1
•s
CD
o
o
V o .5,
" «
§
PI9!A
91
-------
00
V"
I
o
o
CO
£
I
i
Q
669J1S PieiA
IN
•^
O
2
*
D.
CO
I
O
(0
«
Q
£
o>
PI9IA
92
-------
o
o
}ueojad
o>
i/>
^
o
E
•s
(O
PI3!A
PI9IA
93
------- |