GEOLOGICA CARPATHICA, 53, 2, BRATISLAVA, APRIL 2002
127 — 132
DETERMINATION OF INTERNAL SHEAR STRENGTH PARAMETERS
OF GEOCOMPOSITE CLAY LINERS
BILJANA KOVAČEVIĆ ZELIĆ
, DAVORIN KOVAČIĆ
and DOBROSLAV ZNIDARČIĆ
University of Zagreb, Faculty of Mining, Geology and Petroleum Engineering,
Pierottijeva 6, Zagreb, Croatia; email@example.com
BBR-CONEX, Kalinovica 3, Zagreb, Croatia; firstname.lastname@example.org
University of Colorado at Boulder, Department of Civil, Environmental, and Architectural
Engineering, Boulder, USA; email@example.com
(Manuscript received October 4, 2001; accepted in revised form December 13, 2001)
Abstract: Geocomposite clay liners (GCLs) are used in environmental, transportation and geotechnical engineering
applications. Determination of the internal and interface shear strength parameters of GCLs has a huge importance for
stability analyses. Therefore, the study of the internal shear strength of one type of nonreinforced GCLs was performed.
Laboratory testing programme consisted of five series of direct shear tests. Special attention was given to the influence
of the specimen hydration procedure and horizontal displacement rate on the shear test results. The analysis of the direct
shear tests, presented in the paper, clearly demonstrate that the measured values of internal shear strength depend on the
way of performing the laboratory tests. The internal shear strength envelopes for nonreinforced GCLs are proposed on
the basis of obtained results.
Key words: landfills, bentonite, geocomposite clay liners (GCLs), internal shear strength, direct shear test, nonreinforced
Geocomposite clay liners represent one type of the geosyn-
thetics that have been used more frequently since the 70’s. Ac-
cording to ASTM a geosynthetic is “a planar product manu-
factured from polymeric material used with soil, rock, earth,
or other geotechnical engineering related material as an inte-
gral part of a man-made project, structure, or system”. Koern-
er (2000) gives a broad overview of the various possibilities
for their engineering application.
Geocomposite clay liners (IGS 2000) or geosynthetic clay
liners (ASTM 1997) are manufactured hydraulic barriers con-
taining a layer of high-quality sodium bentonite clay attached
or adhered to geotextiles or a geomembrane. Numerous com-
mercially manufactured products of GCLs are available on the
worldwide market. They are used in environmental, transpor-
tation and geotechnical engineering applications. In environ-
mental applications GCLs act as a hydraulic barrier compo-
nent on waste disposal sites as a part of the liner and cover
systems. Because of their low permeability they serve very of-
ten as a replacement for low permeability soil or clay liners. In
comparison with classical clay liners they provide many ad-
vantages, including a greater resistance to differential settle-
ments, desiccation, and freeze-thaw deterioration. They also
have self-healing characteristics due to their swelling poten-
tial. It is also important that their installation is simple, easy
and less time-consuming.
From the engineering design point of view, stability analy-
ses of the hydraulic barriers at waste disposal sites are one of
the most important issues. Therefore, proper determination of
the shear strength parameters of all the barrier components
plays a major role in the design process. Internal shear strength
parameters of GCLs are generally obtained from laboratory di-
rect shear tests, but the amount of the published data is limited.
The principal issues for internal shear strength testing of GCLs
are: test configuration, gripping technology, specimen size, de-
gree of hydration and hydration liquid, normal stress range,
and shear strain rate. It was found, by reviewing the published
data, that there is a significant variability of the test procedures
used and the results obtained. Therefore, a study of the internal
shear strength of one type of nonreinforced GCL was per-
formed. Special attention was given to the determination of the
influence of two parameters: specimen hydration procedure
and shear strain rate. The laboratory testing program was
planned accordingly and consisted of five series of direct shear
tests. Two procedures of specimen hydration and four different
shear strain rates were investigated in the program.
The peak and residual shear strength envelopes were deter-
mined on the basis of test results. Influence of the hydration
procedure and shear strain rate is clearly shown. The appropri-
ate laboratory testing procedure is proposed for the determina-
tion of internal shear strength of GCLs. With this procedure
the two most relevant parameters are included in the testing
Internal shear strength testing of GCLs
There are a wide variety of commercially manufactured
products of GCLs on the market. From the shear strength point
of view, they are divided into two main groups: reinforced and
nonreinforced GCLs. In our investigation, one of the very few
nonreinforced GCLs known under commercial name Claymax
200R (CETCO, USA) is used. It consists of approximately
128 KOVAČEVIĆ ZELIĆ, KOVAČIĆ and ZNIDARČIĆ
of adhesive bonded natural sodium bentonite sand-
wiched between two lightweight woven geotextiles. A cross-
section sketch of the nonreinforced GCL is shown in the Fig.
1. In the other type of GCLs called reinforced GCLs, geotex-
tiles are held together, for example, by stitch bonding or nee-
dlepunching. The physical bonding of the geotextiles enhances
the internal resistance of GCLs to shearing.
Several reasons influenced the decision to test one type of
nonreinforced GCL in our investigation. It was our opinion
that in order to understand the shearing behaviour of GCLs, it
was necessary to investigate in detail the behaviour of bento-
nite itself. In the case of reinforced GCLs, their behaviour in
direct shear test is influenced by the presence of synthetic
yarns. Moreover, previous investigations (Gilbert et al. 1996)
prove that reinforced GCLs have larger peak strengths, but the
residual strength is the same as for nonreinforced ones. Final-
ly, the specimen size is not a critical parameter in testing the
nonreinforced GCLs, as is the case for reinforced ones. In the
case of reinforced GCLs large specimens are necessary. Large
samples are not practical for standard tests in common geo-
technical laboratories. Some investigators show that there is
also a problem of partial hydration of specimens that leads to
inaccurate results in shear (Gilbert et al. 1997).
GCLs are geocomposites consisting of geological material
(predominantly sodium bentonite) and synthetic materials
(geotextiles, geomembranes). Bentonite is a key component
because its function is to maintain low hydraulic permeability.
Therefore, it will be presented in some details.
Bentonite is a naturally occurring clay that is extremely hy-
drophilic (water attracting). In contact with water or even wa-
ter vapour bentonite attracts the water forming a complex con-
figuration that leaves little free-water space in the voids. This
fact explains the resulting low permeability of most GCLs, the
most important property of barrier layers.
Bentonites, which are used for the production of GCLs, con-
sist mainly of three-layer mineral montmorillonite of the
smectite group. Other ingredients like quartz, christoballite,
feldspars, muscovite/illite, and other clay and nonclay miner-
als are not important for the functionality in waste contain-
ment applications (Egloffstein 1997). Because of the high con-
tent of montmorillonite (60—90 %), bentonites have desirable
properties like swelling, high ion exchange capacity, adsorp-
tion capacity against heavy metals and very low permeability.
Most of the commercially available GCLs use sodium ben-
tonites. Calcium bentonites are rarely used because of lower
swelling potential and higher permeability. The water adsorp-
tion capacity of sodium bentonites is approximately 400—700 %
compared with the capacity of calcium bentonites of roughly
200 % (Egloffstein 1995). The permeability of calcium bento-
nites is 1—5
m/s and of sodium bentonites 1—3
(Egloffstein 1997). GCLs manufactured in the USA use natu-
ral sodium bentonites found in Wyoming. In many European
countries natural calcium bentonites are found (Koerner
Fig. 1. Cross-section sketch of nonreinforced GCL.
Montmorillonite content [%]
Specific surface area [m
Exchange capacity [meq/100g]
1997). In order to receive better swelling properties such calci-
um bentonites are activated with soda (soda activated bento-
nites). Some properties of natural sodium bentonites are given
in Table 1, compared to the properties of a calcium bentonite,
for example, from Bavaria.
In addition to the previously mentioned favourable proper-
ties of bentonites for the waste containment applications, there
are some critical issues that should be kept in mind. The ion-
exchange capability of bentonite under typical use conditions
can cause the transformation of original sodium to calcium
bentonites. As a consequence of that, some physical character-
istics will be changed, like swelling, permeability and self-
healing properties (Egloffstein 1997). This can be avoided by
the proper installation of GCLs. The design engineer should
consider the compatibility of GCL with the adjacent soils or
liquids with which it will come into contact. GCLs should not
be used if they can come into contact with limestone. Extreme
weather conditions (heavy raining, very dry areas) should also
be avoided (Mackey 1997).
The most interesting property of bentonites for our research
is their shear strength. Previous investigations show that they
have very low strength especially in free-swell conditions.
Mesri & Olson (1970) and Olson (1974) conducted consolidat-
ed-undrained triaxial tests with pore water pressure measure-
ments on homoionic sodium- and calcium-montmorillonites.
Gleason et al. (1997) performed consolidated-drained direct
shear tests on thin layers of bentonites according to ASTM
D3080. The obtained values of shear strength parameters of
montmorillonites and bentonites are shown in Table 2.
Direct shear tests were performed in a modified shear appa-
ratus. Some modifications of a standard shear box as used in
common soil mechanics laboratories were necessary. Standard
shear boxes have the following dimensions: specimen size 70
70 mm or 60
60 mm, specimens height 20 mm. GCLs have a
height of approximately 5 mm in as-received state, and their
height is variable depending on the degree of saturation.
Therefore, porous plates of different thickness were added to
the apparatus. Specimen size was enlarged to 100
100 mm. In
that way, the maximum horizontal displacement was enlarged,
too, and the measurement of residual strength was achieved.
Finally, gripping of the specimens is solved by using teethed
metal plates, containing 112 teeth on the size of 100
Details of the modified shear box are presented in Fig. 2.
Laboratory testing program
Published results of the internal shear strength parameters
for nonreinforced GCLs demonstrate a huge variety of data
Table 1: Properties of bentonites (Koerner 1997).
SHEAR STRENGTH PARAMETERS OF GEOCOMPOSITE CLAY LINERS 129
ing the hydration for 24 hours (normal consolidation) or 9
days (extended consolidation), vertical displacements were
measured and recorded continuously. Shearing of the speci-
mens with different rates of displacements begun at the end of
the hydration stage until the relative displacement of 15 %
was achieved. Depending on the rate of displacements, shear-
ing lasted from 17 minutes for series I and V to 9.5 days for
series IV (Table 3).
Stress and strain components
During the direct shear testing of nonreinforced GCL, shear
stress, vertical and horizontal displacements were measured
and recorded in an output file. Time intervals for recording
output data were adapted to the different test durations given
in Table 3. On the basis of this data, stress-displacement
curves were created for every series of testing program. Three
curves coresponding to three normal stress values (50, 100
and 200 kPa) for a series III are shown on Fig. 3. It can be
seen that nonreinforced GCLs demonstrate shearing behav-
iour similar to that of overconsolidated clay materials exibit-
ing peak and residual strengths. In our case, residual strengths
were determined at the displacement of 15 mm, that is at the
relative deformation of 15 %. Total values of peak and residu-
al shear strengths along with their ratios are given in Table 4.
By reviewing the data presented in Table 4, we can con-
clude the following:
– the obtained values of peak and residual strengths are low-
er for lower displacement rates (comparing series I to IV),
– strength reduction from peak to residual values are higher
for lower displacement rates,
– extended consolidation produces lower strengths, too
(comparing series I and V).
CU triaxial test
CD direct shear test
Liquid limit [%]
880 – 1160
190 – 220
Plasticity index [%]
160 – 190
Friction angle [°]
Table 2: Shear strength parameters of montmorillonites and bentonites (Mesri & Olson 1970; Gleason et al. 1997).
Fig. 2. Cross-section of the modified shear box.
(Daniel & Shan 1991; Daniel et al. 1993; Fox et al. 1998; Shan
1993). The range of the measured values of friction angle,
and cohesion, c, are the following:
= 22—37° c = 7—50 kPa,
= 0—27° c = 0.2—30 kPa.
There is obviously a huge scatter of the measured values of
shear strength parameters. However, by reviewing the pub-
lished data, it can be seen that laboratory procedures differ sig-
nificantly concerning the following issues: test configuration,
specimen size, hydration procedure, normal stress range, strain
rate, and maximum horizontal displacement. One of the rea-
sons for the scatter is that the established test methods or stan-
dards for the shear strength determination of GCLs did not ex-
ist at that time. We concluded that the two most important
parameters for which the influence should be clearly deter-
mined are: specimen hydration procedure and shear strain rate.
Our laboratory testing program was therefore planned accord-
ingly. It consisted of five series of direct shear tests (Table 3).
Two procedures for specimen hydration (series I and V) and
four different shear strain rates (series I—IV) were investigated
in the program.
Specimens were placed into the direct shear box during the
hydration stage. Immediately after the application of normal
stress on the specimen, water was added to the shear box. Dur-
Table 3: Laboratory testing program.
= 50 kPa
= 100 kPa
= 200 kPa
I-IV: Consolidation stage 24 hours.
V: Consolidation stage 9 days.
130 KOVAČEVIĆ ZELIĆ, KOVAČIĆ and ZNIDARČIĆ
* Series V – extended consolidation
Horizontal displacement [mm]
Shear stress [kPa]
Internal shear strength envelopes
Peak and residual shear strength envelopes are shown in
Figs. 4 and 5. Note that the scale for the x- and y-axes are dif-
ferent for clarity of presentation. Both figures show that differ-
ent internal shear strength envelopes will be obtained depend-
ing on the way of performing the direct shear test. The
strength envelopes for series I—III and V are almost parallel. It
means that they have similar friction angle but different cohe-
sion. Series IV with a lowest displacement rate shows signifi-
cantly smaller friction angle.
The obtained values of friction angle and cohesion for five
series of tests are extracted in Table 5. The range of obtained
values for friction angle and cohesion are:
= 11.5—16.4° c = 1—15.4 kPa,
= 5.1—11.2° c = 0—15.3 kPa.
According to the ASTM D3080 standard and our oedometer
test results, drained conditions were realized only for series IV.
Fig. 3. Shear stress vs. horizontal displacement.
Table 4: Results of the direct shear tests.
Fig. 5. Residual shear strength envelopes.
Normal stress [kPa]
Peak shear stress
Fig. 4. Peak shear strength envelopes.
Normal stress [kPa]
Although the range of obtained values is smaller than the pre-
viously mentioned range from the published data, it is obvious
that the influence of the displacement rate and the hydration
procedure should not be disregarded.
a. Influence of the displacement rate: The influence of a dis-
placement rate on the measured peak and residual shear
strength is demonstrated in Figs. 6 and 7. Note that the results
of the series V are represented on both figures by full squares,
but they were not included in the trend lines because of differ-
ent hydration procedure. It can be seen that higher values of
displacement rate cause larger values of measured strength. It
is interesting to compare the results for series IV and V (Table
5). They have almost identical total duration of test including
consolidation and shearing stage and therefore approximately
the same hydration conditions. Their displacement rates repre-
sent the minimal and maximal values of the applied range. It
can be seen that displacement rate influences the value of a
friction angle much more than the value of a cohesion. Look-
ing at the total values of measured shear strength (Table 4) for
series IV and V, one can see that in spite of the similar hydra-
tion, series IV shows lower strength due to a much lower dis-
placement rate. The only explanation for such results could be
found in the rheological properties of the material.
b. Influence of the hydration procedure: The influence of a
hydration procedure can be seen by comparing the results for
series I and V (Table 5). The samples from series I and V were
SHEAR STRENGTH PARAMETERS OF GEOCOMPOSITE CLAY LINERS 131
sheared with the same displacement rate, but their hydration
procedure was totally different. The standard hydration and
consolidation procedure of 24 hours duration was used for se-
ries I. For the series V extended hydration and consolidation
lasted 9 days. The cohesion for series V is lower by a factor of
2 or more when compared to the cohesion for series I, looking
to the peak and residual values. The friction angle in both cas-
es is somewhat larger for series V than for series I. It can be
concluded that hydration procedure affects much more the co-
hesion than the friction angle of GCLs.
By reviewing all data recorded during our investigation, the
influence of hydration procedure is demonstrated through the
influence of the final water content of samples on the mea-
sured shear strength. The final water content of samples de-
pends on the applied normal stress and the total test duration
(Fig. 8). Lower values of applied normal stress and longer test
duration cause larger final water content of samples. Achieved
values of the final water content affect the measured values of
shear strength. This fact is demonstrated in Fig. 9 through the
influence of the final water content on the peak shear strength.
The influence of the displacement rate, and test duration on the
results is also shown in the same figure.
The interpretation of Figures 8 and 9 took us to the follow-
– final water content is inversely proportional to the applied
normal stress, that is larger normal stress causes a lower final
– final water content is proportionally dependant on the test
duration, that is a longer test cause higher values of final water
– longer test duration results in lower shear strengths and this
is caused by higher values of the final water contents,
– higher displacement rates on the contrary cause higher
shear strengths, that is rate effects exist, too.
Table 5: Shear strength parameters.
Cohesion, c [kPa]
Cohesion, c [kPa]
Fig. 6. Peak shear strength vs. displacement rate.
Fig. 7. Residual shear strength vs. displacement rate.
y = 2,59Ln(x) + 67
y = 2,49Ln(x) + 43
y = 1,29Ln(x) + 23
Displacement rate [mm/min]
Peak shear stress
y = 3,7Ln(x) + 51
y = 3,15Ln(x) + 35
y = 1,48Ln(x) + 18
Displacement rate [mm/min]
Fig. 9. Peak shear strength vs. final water content.
= 356,5 σ
Final water content,
Fig. 8. Final water content of samples.
= 43000 w
Final water content, w
Peak shear stress,
acem ent rate
132 KOVAČEVIĆ ZELIĆ, KOVAČIĆ and ZNIDARČIĆ
Proposed internal shear strength criteria
As already mentioned in the introduction to this article, sta-
bility analyses of the hydraulic barriers at waste disposal sites
are one of the most important issues for design engineers.
Therefore, proper values of the shear strength parameters of all
barrier components including GCLs have to be chosen.
The analysis of the direct shear tests, presented in the paper,
clearly demonstrate that the measured values of the shear
strength depend on the way of performing the laboratory tests.
The main factors affecting the results are hydration procedure
and displacement rate. Therefore, we propose the shear
strength envelopes as shown in Fig. 10. The envelopes are cre-
ated on the basis of the results presented in Figs. 6 and 7. Peak
and residual strengths are obtained by the extrapolation of the
functions to the displacement rate of 0.001 mm/min for all
three normal stress values. These values are redrawn in Fig.
10, giving the proposed envelopes. The real values of mea-
sured peak and residual strengths are presented by full and
blank marks, respectively. The envelopes defined in that way
will give the opportunity for engineers to design waste dispos-
Acknowledgments: The work described in this paper is fund-
ed partly by the U.S.—Croatian Joint Board on Scientific and
Technological Cooperation JF 150 “Impervious barriers for
landfills in karst” and partly by the Croatian Ministry of Sci-
ence and Technology Project “Geotechnology for solid waste
landfills”. This support is gratefully acknowledged.
ASTM D 3080—90: Standard Test Method for Direct Shear Test of
Soils Under Consolidated Drained Conditions.
ASTM D 4439—97a: Standard Terminology for Geosynthetics.
Daniel D.E. & Shan H.-Y. 1991: Results of direct shear tests on hy-
drated bentonitic blankets. Project Report, Geotech. Engrg.
Ctr., The University of Texas, Austin, 1—13.
Daniel D.E., Shan H.-Y. & Anderson J.D. 1993: Effects of partial
wetting on the performance of the bentonite component of a
geosynthetic clay liner. Proceedings, Geosynthetics ‘93, Van-
couver, B.C., IFAI Publ., 1483—1496.
Egloffstein T. 1995: Properties and test methods to assess bentonite
used in geosynthetic clay liners. In: R.M. Koerner, E. Gartung
& H. Zanzinger (Eds.): Proceedings of an International Sympo-
sium: Geosynthetic Clay Liners, Nurnberg, Germany 14—15,
April 1994. Balkema, Rotterdam, 51—71.
Egloffstein T. 1997: Geosynthetic clay liners. Part six: Ion ex-
change. Geotechnical Fabrics Report, June-July, 38—43.
Fox P.J, Rowland M.G. & Scheithe J.R. 1998: Internal shear
strength of three geosynthetic clay liners. Journal of Geotech-
nical and Geoenvironmental Engineering 124, 10, 933—944.
Gilbert R.B., Fernandez F. & Horsfield D.W. 1996: Shear strength
of reinforced geosynthetic clay liners. Journal of the Geotech-
nical Engineering 122, 4, 259—266.
Gilbert R.B., Scranton H.B. & Daniel D.E. 1997: Shear strength
testing for geosynthetic clay liners. In: L.W. Well (Ed.): Testing
and acceptance criteria for geosynthetic clay liners. ASTM STP
Gleason M.H., Daniel D.E. & Eykholt G.R. 1997: Calcium and sodi-
um bentonite for hydraulic containment applications. Journal
of Geotechnical and Geoenvironmental Engineering 123, 5,
IGS 2000: Recommended descriptions of geosynthetics functions,
geosynthetics terminology, mathematical and graphical symbols.
Koerner R.M. 1997: Perspectives on geosynthetic clay liners. In:
L.W. Well (Ed.): Testing and acceptance criteria for geosyn-
thetic clay liners. ASTM STP 1308, 3—20.
Koerner R.M. 2000: Emerging and future developments of selected
geosynthetic applications (The Thirty-Second Terzaghi Lec-
ture). Journal of Geotechnical and Geoenvironmental Engi-
neering 126, 4, 293—306.
Mackey R.E. 1997: Geosynthetic clay liners. Part five: Design, per-
mitting and installation concerns. Geotechnical Fabrics Re-
port, January—February, 34—39.
Mesri G. & Olson R.E. 1970: Shear strength of montmorillonite.
Geotechnique 20, 3, 261—270.
Olson R.E. 1974: Shearing strength of kaolinite, illite, and montmo-
rillonite. ASCE Journal of the Geotechnical Engineering Divi-
sion 100, GT11, 1215—1229.
Shan H-Y. 1993: Stability of final covers placed on slopes contain-
ing geosynthetic clay liners. Dissertation, The University of
Texas at Austin, 1—296.
Fig. 10. Proposed shear strength envelope.
Normal stress [kPa]
Shear stress [kPa]
A series of direct shear tests were performed in the modified
shear box in order to determine the internal shear strength of
one type of GCLs. The results showed that by using the modi-
fied shear box, peak and residual strength could be obtained
for the nonreinforced types of GCLs. The key parameters for
performing the direct shear tests are the hydration procedure
and the applied displacement rate. The influence of these pa-
rameters on the results is demonstrated. On the basis of the ob-
tained results, peak and residual strength criteria are proposed.
As the waste disposal facilities have a huge impact on the en-
vironment, it is important for design engineers to have reliable
shear strength parameters. It is believed that the proposed
shear strength criteria enable designing of new facilities in a