Geotextiles and Geomembranes 49 (2021) 789–803

Contents lists available at ScienceDirect

Geotextiles and Geomembranes

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

Failure mechanisms of geosynthetic clay liner and textured geomembrane

composite systems
Shahin Ghazizadeh a, Christopher A. Bareither b, *
a
Wood Environment and Infrastructure Solutions Inc., 2000 S Colorado Blvd # 2-1000, Denver, CO, 80222, USA
b
Civil & Environmental Engineering, Colorado State University, 1372 Campus Delivery, Fort Collins, CO, 80523, USA

A R T I C L E I N F O A B S T R A C T

Keywords: The objective of this study was to evaluate shear behavior and failure mechanisms of composite systems
Direct shear comprised of a geosynthetic clay liner (GCL) and textured geomembrane (GMX). Internal and interface direct
Geomembrane shear tests were performed at normal stresses ranging from 100 kPa to 2000 kPa on eight different GCL/GMX
Geosynthetic clay liner
composite systems. These composite systems were selected to assess the effects of (i) GCL peel strength, (ii)
Shear strength
geotextile type, (iii) geotextile mass per area, and (iv) GMX spike density. Three failure modes were observed for
the composite systems: complete interface failure, partial interface/internal failure, and complete internal fail­
ure. Increasing normal stress transitioned the failure mode from complete interface to partial interface/internal
to complete internal failure. The peak critical shear strength of GCL/GMX composite systems increased with an
increase in GMX spike density. However, the effect of geotextile type and mass per area more profoundly
influenced peak critical shear strength at normal stress > 500 kPa, whereby an increase in geotextile mass per
area enhanced interlocking between a non-woven geotextile and GMX. Peel strength of a GCL only influenced the
GCL/GMX critical shear strength when the failure mode was complete internal failure.

1. Introduction strength of the reinforcement fibers and connection strength between

the fibers and carrier geotextile. The fibers can be thermally fused to the
Geosynthetic clay liners (GCLs) are used in waste containment bar­ carrier geotextile to create a heat-treated NP GCL or if the fibers are not
rier systems due to their hydraulic performance, economic advantages, thermally fused the GCL is a non-heat-treated NP GCL.
ease of installation, and self-healing properties (Robbins et al., 1995; Internal and interface shear strength of GCLs are commonly evalu­
Sivakumar et al., 2001; Guyonnet et al., 2009; Fox and Stark 2015; Rowe ated via direct shear and torsional ring shear tests (Fox and Stark 2015).
et al., 2019; Rowe 2020). In side-slope and base liner systems of Researchers have evaluated the effects of hydration and consolidation
containment systems, GCLs are often combined with textured geo­ procedures, normal stress, displacement rate, reinforcement type (i.e.,
membranes (GMX). A GCL/GMX composite system is exposed to shear needle-punching versus stitched bonding), heat-treatment method,
and normal stresses that must be resisted internally within the GCL and specimen size, peel strength, dynamic loading, temperature, hydration
along the interface between the GMX and geotextile of the GCL to pre­ with non-standard solutions, and creep on the internal shear strength of
vent failure of the barrier system (Fox and Stark 2015). Determining the GCLs (Gilbert et al., 1996; Hewitt et al., 1997; Siebken et al., 1997;
critical shear strength (τcr) of the GCL/GMX composite system, which is Trauger et al., 1997; Fox et al. 1998, 2006; Eid et al., 1999; Koerner
defined herein as the minimum of the GCL internal shear strength and et al., 2001; Zornberg et al., 2005; Müller et al., 2008; Ross 2009; Fox
GCL/GMX interface shear strength, is important for the design of barrier et al., 2010; Fox and Ross 2011; Ross and Fox 2015; Ghazizadeh and
systems. Bareither 2018a, 2018b, 2020; Bareither et al., 2018; Sawada et al.,
The internal shear strength of GCLs often is enhanced via needle- 2019; Thielmann et al., 2016; Feng et al., 2020). In previous studies on
punching, whereby fibers from a non-woven cover geotextile are GCL/GMX interface shear strength, researchers have evaluated the ef­
punched through the bentonite layer and entangled into the carrier fects of geotextile characteristics, GMX characteristics, surface rough­
geotextile of the GCL. Internal shear strength of needle-punched rein­ ness, monotonic loading, bentonite extrusion, polymer type,
forced GCLs (NP GCLs) increases due to contributions from tensile displacement rate, and out of plane deformation (Gilbert et al., 1996;

* Corresponding author.
E-mail addresses: [email protected] (S. Ghazizadeh), [email protected] (C.A. Bareither).

https://doi.org/10.1016/j.geotexmem.2020.12.009
Received 1 July 2020; Received in revised form 16 December 2020; Accepted 22 December 2020
Available online 13 January 2021
0266-1144/© 2021 Elsevier Ltd. All rights reserved.
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

Fig. 1. Schematic shear and normal stresses within a composite system composed of a textured geomembrane (GMX) and needle-punched geosynthetic clay
liner (GCL).

Hewitt et al., 1997; Hillman and Stark 2001; Olsta and Swan 2001; (Jones and Dixon, 1998; Frost et al., 2001; Hebeler et al., 2005; Ross
Triplett and Fox 2001; Chiu and Fox 2004; Stark et al., 2004; Li and 2009; Li and Gilbert 2006; Bacas et al. 2011, 2015; Fox and Stark 2015).
Gilbert 2006; Fox and Kim 2008; Vukelić et al., 2008; McCartney et al., The interlocking strength is a function of geotextile properties (e.g.,
2009; Zornberg and McCartney 2009; Chen et al., 2010; Bacas et al., tensile strength of fiber), geotextile type (i.e., woven vs. non-woven),
2011; Eid 2011; Fox and Ross 2011; Bacas et al., 2015; Hanson et al., and GMX spike interbedding within the geotextile.
2015; Ross and Fox 2015; Stark et al., 2015; Thielmann et al., 2016; Shear resistance mechanisms controlling the interface shear strength
Khilnani et al., 2017). and magnitude of τD are dependent on σn. At low σn, interlocking occurs
Shear testing to evaluate the internal shear strength of GCLs forces superficially between GMX asperities and geotextile fibers, and τD pri­
failure to occur internally within the GCL. However, internal failure of a marily is frictional. As σn increases, the geotextile of the GCL achieves
GCL in a GCL/GMX composite system may not occur if internal shear more intimate contact with the GMX (i.e., interbedding) and inter­
resistance is larger than the mobilized shear resistance along the GCL/ locking of the GMX occurs within the geotextile structure. Interbedding
GMX interface (Stark et al., 2015). In contrast, studies focused on of the geotextile with the GMX can also increase frictional resistance due
GCL/GMX interface shear strength may not observe internal shear to increased contact area. The improved interlocking and increased
deformation and failure of GCLs. Shear behavior of GCL/GMX composite frictional resistance increases τD as σn increases (Bacas et al., 2015).
systems via evaluation of both GCL internal and GCL/GMX interface Under a certain σn, τ and τD increase to a point at which three sce­
shear strength have only been the focus of select studies (e.g., Ross 2009; narios may occur that correspond to distinct failure modes in a GCL/
Fox and Ross 2011; Thielmann et al., 2016). Therefore, there was a need GMX composite system. In Scenario 1, τD reaches a maximum mobilized
to further evaluate shear behavior and failure of GCL/GMX composite shear stress (τDMAX) at the GCL/GMX interface and slippage initiates
systems. between the GMX and geotextile of the GCL. Assuming that the peak
The objective of this study was to improve our understanding of the internal GCL shear strength (τP-IN) is greater than τDMAX, negligible in­
shear mechanism of GCL/GMX composite systems via evaluating GCL ternal deformation of reinforcement fibers develops such that the failure
internal shear strength and GCL/GMX interface shear strength for mode of the GCL/GMX composite system is complete interface failure.
different composite systems and normal stress to define potential failure In Scenario 2, τDMAX is mobilized in a similar manner to Scenario 1
mechanisms. Direct shear tests were conducted to assess (i) geotextile such that slippage initiates between the GMX and the GCL. However, in
mass per area, (ii) geotextile type (i.e., woven versus non-woven), (iii) Scenario 2 τDMAX is sufficient to yield some internal GCL deformation
GCL peel strength, and (iv) GMX spike density. A total of 17 internal and potentially led to failure of some reinforcement fibers. Despite this
shear tests and 38 interface shear tests were performed on composite internal GCL deformation, τDMAX remains lower than τP-IN, and complete
systems comprising of three non-heat treated NP GCLs with different GCL internal failure does not occur. The failure mode in Scenario 2 is
peel strength and geotextile characteristics, and two GMXs with partial interface/internal failure.
different spike densities. In Scenario 3, τD increases to become comparable to τP-IN of the GCL
such that internal GCL failure occurs before full mobilization of τDMAX
2. Background along the GCL/GMX interface. Thus, negligible slippage occurs between
the GCL and GMX and the failure mode is complete internal failure.
A schematic of a GCL/GMX composite system subjected to shear The shear behavior and failure mode of a GCL/GMX composite sys­
stress (τ) and normal stress (σn) is shown in Fig. 1. Stresses applied to the tem depends on factors affecting GCL internal shear strength and GCL/
GMX lead to a developed normal stress (σD) and shear stress (τD) along GMX interface shear strength, such as σn, geotextile and GMX charac­
the GCL/GMX interface. Shear resistance of the GCL/GMX interface teristics, and GCL peel strength. Past research has shown a transition
develops from (i) frictional resistance between the GMX and geotextile from complete interface, to partial interface/internal, to complete in­
of the GCL and (ii) interlocking (i.e., hook and loop mechanism) be­ ternal failure of GCL/GMX composite systems with increasing normal
tween the geomembrane asperities (e.g., spikes) and geotextile fibers stress (e.g. Gilbert et al., 1996; Fox and Ross 2011; Theilmann et al.,

790
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

Table 1 4. Experimental procedure

Characteristics of the geosynthetic clay liners.
Category Characteristic GCL1 GCL2 GCL3 Experiments were performed with a displacement-controlled direct
a shear apparatus shown in Fig. 2. The apparatus is capable of both GCL
General Peel strength (N/m) 980 2180 3850
Heat treatment method NHT NHT NHT internal and GCL/GMX interface shear tests with σn up to 500 kPa for
Geotextile Polymer type PP PP PP 300 mm by 300 mm specimens and σn up to 2000 kPa for 150 mm by
Carrier geotextile type NW W NW 150 mm specimens. Gripping systems for the geosynthetics included
Cover geotextile type NW NW NW pyramid-tooth plates with 2-mm-tall teeth for σn < 300 kPa and 1-mm-
Carrier geotextile mass/area (g/ 260 130 360
m2)b
tall teeth for σn ≥ 300 kPa. The pyramid-tooth plates were designed
Cover geotextile mass/area (g/ 230 230 280 based on recommendations in Allen and Fox (2007) and compression
m2)b tests on representative GCLs such that during shear testing there is no
Bentonite Bentonite type Granular Granular Granular interference between the upper and lower pyramid-tooth plates. Addi­
Bentonite mass per area (g/m2)c 4220 4910 5570
tional details on the displacement-controlled direct shear machine and
Note: NHT = Non-heat treated; PP = Polypropylene; W = woven; NW = non- effectiveness of this gripping systems is in Bareither et al. (2018).
woven. All internal and interface shear tests were performed on 150 mm by
a
Reported by manufacturers, ASTM D6496/6496M. 150 mm GCL specimens to perform experiments at σn > 500 kPa. ASTM
b
Based on 20 measurements, ASTM D5261. D6243 allows specimens smaller than 300 mm × 300 mm provided that
c
Based on 20 measurements, ASTM D5993.
the specimen dimensions do not affect results (e.g., Thielmann et al.,
2016). A comprehensive discussion on verification that the 150 mm test
specimens yield similar shear behavior to larger specimens is in Bare­
Table 2
Characteristics of the textured geomembranes.
ither et al. (2018).
The GMX and GCL specimens were cut along machine direction ac­
Characteristic Standard GMX1 GMX2
cording to ASTM D6072. The GCL specimens were cut to initial di­
Manufacturing Process – Structured Structured mensions of 150-mm wide by approximately 200-mm long in machine
Polymer Typea – LLDPE LLDPE direction to facilitate gripping the GCL. The GMX specimens were cut to
Average Core Thickness (mm)a ASTM D5994 1.5 1.5
Average Asperity Height (mm)a ASTM D7466 0.5 0.5
dimensions of 175 mm by 300 mm, with the longer dimension in the
Spikes Density (No. Spikes/100 cm2) – 340 840 direction of shear. The longer GMX specimens ensured that the GCL
a remained in contact with the GMX throughout shear displacement.
Reported by manufacturers.
Specimen hydration and consolidation were performed according to
a 2-stage procedure (e.g., Fox et al., 1998; Bareither et al., 2018). In the
2016; Khilanin et al., 2017). Gilbert et al. (1996) reported complete hydration stage (i.e., first stage), GCL specimens were sandwiched be­
internal failure in GCL/GMX interface shear tests at σn as low as 13.8 tween two perforated PVC plates and hydrated in a plastic bin with
kPa. In contrast, internal GCL failure was not reported until σn exceeded de-ionized water (DIW) for 48 h under σn = 20 kPa. The hydration stage
300 kPa in Eid (2011), 1300 kPa in Fox and Ross (2011), and never was similar for all GCL specimens tested in the internal and interface
developed in Stark et al. (2015) for σn = 2070 kPa. These differences in shear tests.
failure modes were attributed to the differences in geotextile and GMX At the end of the hydration, specimens were transferred to the direct
properties as well as GCL peel strength. shear box to start consolidation (i.e., second stage). Internal shear
specimens were sandwiched between pyramid-tooth plates for consoli­
3. Materials dation and shearing. Extra geotextile in the shear direction was clamped
to both ends of the pyramid-tooth plates. The GCL specimens for inter­
3.1. Geosynthetic clay liners specimens face shear tests were secured to a single pyramid-tooth plate and placed
in contact with a GMX. The GMX was sandwiched between a GCL and a
Characteristics of the three non-heat treated, needle-punched GCLs 300-mm by 300-mm pyramid-tooth plate. Pyramid teeth gripped into
used in this study are in Table 1. All GCLs were commercially available and secured the GMX in place during shear. Consolidation initiated in­
with polypropylene geotextiles and granular bentonite. However, the side the shear box via reapplying σn = 20 kPa, and subsequently
GCLs had unique peel strength (PS) and geotextile characteristics. Peel doubling σn every 4–6 h to the target normal stress for shearing (σn-s).
strength was provided by the manufacturer and ranged from 980 N/m Once σn-s was obtained, test specimens were equilibrated in the shear
(GCL1) to 3850 N/m (GCL3). The carrier and cover geotextiles were all box for at least 24 h.
non-woven with different mass per area, except for the carrier geotextile All internal and interface shear tests were conducted at a horizontal
of the GCL2, which was woven (Table 1). Carrier and cover geotextile displacement rate of 0.1 mm/min to a horizontal displacement (δh) ≈ 70
mass per area and bentonite mass per area were evaluated on 20, 10 cm mm. This displacement was sufficient to capture peak shear behavior in
× 10 cm samples cut from each GCL roll. GCL3 had the highest geotextile all internal and interface shear tests, but may underestimate residual
mass per area and GCL2 the lowest. Bentonite mass per area ranged from shear strength of NP GCLs in the internal shear tests (Fox and Stark
4220 g/m2 to 5570 g/m2 (Table 1). 2015). Therefore, the large-displacement shear stress at δh = 70 mm (τld)
is used herein for analysis of the internal and interface shear behavior of
GCLs.
3.2. Textured geomembrane specimens The shear area of GCL specimens during internal shear tests
decreased with horizontal displacement due to offset between the upper
Characteristics of GMX1 and GMX2 are in Table 2. Both GMXs had and lower shear platens. Prior to reaching peak shear strength, needle-
spike-structured texturing and were made of linear low density poly­ punched fibers remain in contact with the geotextiles and an area
ethylene (LLDPE). They had the same manufacturer-reported core correction is not required to analyze peak internal shear strength.
thickness (1.5 mm) and asperity height (0.5 mm), but had different spike However, once peak shear strength was reached, failure of reinforce­
spacing in the machine and cross-machine direction that yielded ment fibers initiated and increased with increasing displacement.
different spike densities. Spike density was computed on 10 cm × 10 cm Therefore, area-corrected normal stress and large-displacement shear
areas, and GMX1 had a spike density ≈ 340 spikes/100-cm2 and GMX2 stress were computed at large displacement (Bareither et al., 2018).
had a spike density ≈ 840 spikes/100-cm2. Nominal stresses (i.e., non-area-corrected) were used for peak and large

791
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

Fig. 2. Schematics of (a) cross section and (b) plan view of the direct shear apparatus for displacement-controlled internal and interface shear testing (Bareither
et al., 2018).

displacement shear strength for all interface shear tests because the full inspection also was performed after interface shear tests on both GMX
shear plane area of the GCL remained in contact with the GMX. and GCL to help identify the failure mode of the GCL/GMX composite
All GCLs were visually inspected at the end of the internal shear tests systems.
to ensure internal shear failure. Any signs of stress localization, geo­
textile tearing/elongation, or slippage at the gripping surface indicated
unsuccessful internal shear and the tests were repeated. Visual

792
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

Table 3
A summary of GCL internal direct shear tests.
GCL σn-s (kPa) τp (kPa) φs-p (◦ ) δh-p (mm) AC σ70 (kPa) AC τ70 (kPa) φs-ld (◦ ) Rτ Test Status

GCL1 100 141.0 54.7 25.2 183.1 23.4 7.3 0.09 Successful Internal shear failure
300 233.7 37.9 23.9 549.4 39.7 4.1 0.09
500 295.5 30.6 21.9 915.7 78.6 4.9 0.15
1000 413.7 22.5 21.4 1831.4 130.9 4.1 0.17
2000 570.5 15.9 17.1 3662.7 199.1 3.1 0.19
GCL2 100 164.2 58.7 20.7 183.1 16.8 5.3 0.06 Successful Internal shear Failure
300 293.3 44.4 15.8 549.4 51.3 5.3 0.10
500 416.2 39.8 16.7 915.7 84.6 5.3 0.11
1000 541.2 28.4 17.0 1831.4 122.2 3.8 0.12
2000 700.6 19.3 15.6 3662.7 145.6 2.3 0.11
GCL3a 80 199.7 – – – – – Unsuccessful Internal shear Failure
160 279.5 – – – – –
250 294.2 – – – – –
500 462 – – – – –
1000 626.5 – – – – –
1500 690.4 – – – – –
2000 934.3 – – – – –

Notes: σn-s = shearing normal stress; τp = peak internal shear strength; φs-p = secant friction angle for peak shear strength; δh-p = horizontal displacement at peak shear
strength; AC σ70 = area-corrected shearing normal stress; AC τ70 = area-corrected shear strength at horizontal displacement (δh) = 70 mm; φs-ld = secant friction angle
for large-displacement shear strength; Rτ = post-peak strength reduction ratio (Eq. (1)).
a
Internal shear failure of GCL3 was not successful. Values of τp represent the maximum shear stress measured and not necessarily peak shear strength; thus, φs-p, δh-p,
AC σ70, AC τ70, and Rτ are not defined for GCL3.

5. Results internal shearing of GCL1 and GCL2.

Contrary to GCL1 and GCL2, irregular shear behavior was observed
A summary of the GCL internal direct shear tests is in Table 3, and in GCL3 (Fig. 3c), which had the highest peel strength GCL3 (Table 1).
includes σn-s, τp, secant friction angle (φs) at peak strength, horizontal For example, peak strength and post-peak strength reduction were not
displacement at peak shear strength (δh-p), area-corrected normal stress observed for σn-s ≤ 500 kPa. Although peak strength and post-peak
at δh = 70 mm (AC σ70), area-corrected large-displacement shear strength reduction were observed for σn-s ≥ 1000 kPa, τ-δh relation­
strength at δh = 70 mm (AC τ70), and post-peak strength reduction ratio ships at higher normal stress exhibited abnormal undulations not
(Rτ). The Rτ for the internal shear tests was computed based on Eq. (1). observed in GCLs undergoing complete internal failure. Irregularities in
the τ-δh relationships of GCL3 were indications of unsuccessful internal
ACτ70 /ACσ70
Rτ = / (1) shearing, as suggested by Fox and Kim (2008) and Fox and Stark (2015).
τp σn− s
Post shear pictures of GCL3 specimens from the internal shear tests
Strength parameters were not defined for GCL3 (Table 3) due to the are shown in Fig. 4a for σn-s = 160 kPa and in Fig. 4b for σn-s = 1000 kPa.
unsuccessful internal shearing (described subsequently). For σn-s ≤ 500 kPa, geotextile elongation and tearing along the surface in
A summary of the GCL/GMX interface direct shear tests is in Table 4. contact with the pyramid-tooth plate were observed with limited in­
There were eight different GCL/GMX combinations, which are shown ternal shear deformation. For σn-s ≥ 1000 kPa, varying amounts of fiber
schematically in Table 4 along with the following parameters: σn-s, τp, bundle disentanglement indicated some internal deformation occurred.
φs, δh-p, large-displacement shear strength at δh = 70 mm (τ70), Rτ, and However, there were also indications of geotextile elongation, specimen
observed failure mode. The Rτ for the interface shear tests was computed necking, and geotextile separation (near the clamping area), which
based on Eq. (2). implied that complete internal failure was not successful. Although the
gripping system used in this study surpassed requirements in ASTM
τ70
Rτ = (2) D6243, and was successful in shearing GCL1 and GCL2, the mobilized
τp
interface friction between GCL3 and the gripping system was insufficient
Different definitions of Rτ for the internal and interface shear tests to achieve successful internal shearing of GCL3. This resulted in local
were due to the use of nominal versus area-corrected shear and normal elongation of geotextiles, which indicated slippage between the gripping
stresses. plate and GCL3. Unsuccessful internal failure of high peel strength NP
GCLs has also been reported in (ASTM D6243, Fox et al. (1998), Fox and
5.1. GCL internal shear tests Ross (2011), Fox and Stark (2015), and Bareither et al. (2018).

Relationships of shear stress versus horizontal displacement (τ-δh) 5.2. GCL/GMX interface shear tests
for internal shear tests on GCL1, GCL2, and GCL3 are shown in Fig. 3.
Shear stress measured for GCL1 and GCL2 increased to a well-defined Relationships of shear stress versus horizontal displacement for
peak that coincided with maximum internal shear resistance of the interface shear tests on GCL1, GCL2, and GCL3 are shown in Figs. 5–7,
GCL, and then decreased with increasing displacement, which is typical respectively. All interface shear tests were performed at σn-s = 100, 300,
shear behavior for reinforced GCLs (e.g., Gilbert et al., 1996; Fox et al., 500, 1000, and 2000 kPa except for the cover geotextile of GCL3 with
1998; Zornberg et al., 2005; Fox and Ross 2011; Bareither et al., 2018). GMX2 (Fig. 6b), which was only tested at σn-s = 100, 300, 500 kPa due to
Post-peak strength reduction in NP GCLs is due to decreasing internal limited resources. Each interface shear test exhibited peak strength and
shear resistance as fiber bundles disentangle form the carrier geotextile post-peak strength reduction, which is common behavior of GCL/GMX
and/or undergo tensile rupture. Post shear inspection of GCL1 and GCL2 interfaces (e.g. Triplett and Fox 2001; McCartney et al., 2009; Fox and
specimens revealed fiber disentanglement, fiber rupture, or a combi­ Stark 2015; Thielmann et al., 2016). However, τ-δh data were unique for
nation of both without any sign of stress-localization, geotextile tear­ different GCL/GMX combinations and σn-s, which was due to whether
ing/elongation, or slippage at the gripping surface. The τ-δh the GCL/GMX experienced complete interface failure, partial inter­
relationships and post-shear specimen inspection support successful face/internal failure, or complete internal failure.

793
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

Table 4
Summary of GCL/GMX interface direct shear tests.
GCL GMX GT σn-s (kPa) τp (kPa) φs-p (◦ ) δh-p (mm) τ70 (kPa) Rτ Failure Mode GCL/GMX Schematica

GCL1 GMX1 CO 100 51.1 27.1 9.9 36.7 0.72 IF

300 128 23.1 10 76.2 0.60 IF
500 163 18.1 10.9 100.6 0.62 IF
1000 255 14.3 10.8 167.1 0.66 IF
2000 462.3 13 13.9 320.2 0.69 IF/IN
CA 100 47.7 25.5 9.9 40.8 0.86 IF
300 116.1 21.2 10.5 77 0.66 IF
500 178.8 19.7 12.3 137 0.77 IF
1000 315.8 17.5 13.7 125.7 0.40 IF/IN
2000 516.3 14.5 15.5 190.2 0.37 IN
GMX2 CO 100 115.8 49.2 16.7 43.9 0.38 IF/IN
300 208.1 34.8 15.7 82.9 0.40 IF/IN
500 284.1 29.6 17.8 116.6 0.41 IF/IN
1000 391.7 21.4 16.5 72.05 0.18 IN
2000 560.5 15.7 17.7 167.1 0.30 IN
GCL2 GMX1 CA 100 51.6 27.3 8.4 35.5 0.69 IF
300 108.2 19.8 8.5 68.4 0.63 IF
500 154.6 17.2 8.6 107 0.69 IF
1000 249.3 14 9.3 152 0.61 IF
2000 423.9 12 11.2 283 0.67 IF
GMX1 CO 100 55.9 29.2 8.3 40.1 0.72 IF
300 126.3 22.8 9.2 76.7 0.61 IF
500 183 20.1 8.8 114.9 0.63 IF
1000 297.2 16.6 8.6 179.3 0.60 IF
2000 500.2 14.0 13.5 297.7 0.60 IF
GMX2 CO 100 93.4 43.1 9.8 55.4 0.59 IF
300 249.1 39.7 13.6 104.3 0.42 IF/IN
500 328.5 33.3 14.6 134.4 0.42 IF/IN
1000 549.2 28.8 14.9 195.4 0.36 IF/IN
2000 679.7 18.8 16.4 183.1 0.27 IN
GCL3 GMX1 CA 100 60.8 31.3 9.4 48.3 0.79 IF
300 159.4 28 12.3 119.3 0.75 IF
500 196.8 21.5 11.2 145.3 0.74 IF
1000 329.9 18.3 12.2 216.6 0.66 IF/IN
2000 712.3 19.6 16.2 566.4 0.79 IF/IN
GMX2 CO 100 119.8 50.1 14.3 65.4 0.54 IF
300 277.6 42.8 16.7 115.7 0.42 IF
500 337.3 34 18.6 133.5 0.39 IF/IN

Note: GT = geotextile of the GCL in contact with GMX, whereby CA = carrier and CO = cover geotextile; σn-s = shearing normal stress; τp = peak interface shear
strength; φs-p = secant friction angle for peak shear strength; δh-p = horizontal displacement at peak shear strength; τ70 = area-corrected shear strength at horizontal
displacement (δh) = 70 mm; Rτ = strength reduction ratio (Eq. (2)); IF = complete interface failure; IF/IN = partial interface/internal failure; IN = complete internal
failure.
a
Cover geotextiles shown in grey; carrier geotextiles shown in black; geotextile mass per area increases with thickness; GCL peel strength increases with number of
reinforcement fibers; and geomembrane texturing increasing with number of spikes.

Failure modes of the GCL/GMX composite systems were determined geotextile of GCL1 sheared against GMX2 at σn-s = 300 kPa. Parallel lines
via post-shear specimen inspection and are summarized in Table 4. across the geotextile indicate slippage between the GCL and GMX,
Photographs that demonstrate examples of (i) complete interface failure, similar to the observation for complete interface failure (Fig. 8a).
(ii) complete internal failure, and (iii) partial interface/internal failure However, indications of reinforcement fiber failure and internal defor­
are shown in Fig. 8. Complete interface failure in Fig. 8a is for the cover mation were also observed, which rendered the failure mode partial
geotextile of GCL1 sheared against GMX1 at σn-s = 500 kPa. Parallel lines interface/internal. Shear behavior characteristic of partial interface/
across the specimen in the direction of shear developed from the GMX internal depended on the amount of interface relative to internal failure,
spikes as slippage occurred between the GCL and GMX. Shear behavior which changed the observed shear behavior. Thus, successful identifi­
characteristic of complete interface failure was a relatively low hori­ cation of the partial interface/internal failure mode relied on post-shear
zontal deformation at peak shear strength, low post-peak shear stress specimen inspection.
reduction, and large-displacement shear strength greater than a corre­
sponding large-displacement shear strength for internal shear. 6. Analysis
Complete internal failure is shown in Fig. 8b for the carrier geotextile
of GCL1 sheared against GMX1 at σn-s = 2000 kPa. In post-shear GCL/ Differences between the failure modes of GCL/GMX composite sys­
GMX specimens where complete internal failure was identified, there tems were attributed to variation in GCL peel strength, geotextile type,
were no signs of slippage between the GCL and GMX and the carrier geotextile mass per area, and GMX spike density. Effects of the afore­
geotextile completely separated from the GCL as reinforcement fibers mentioned parameters on the failure mode of a GCL/GMX composite
either ruptured or disentangled. Shear behavior characteristic of com­ system was evaluated via grouping different GCL/GMX interface shear
plete internal failure for a GCL/GMX composite system was identical to tests to isolate specific variables. The subsequent analysis of GCL/GMX
that described for internal GCL shear tests (i.e., well-defined peak shear shear behavior includes the following relationships: (i) peak shear
stress + pronounced post-peak stress reduction). strength – τp versus σn-s; (ii) large-displacement shear strength – τ70
Partial interface/internal failure is shown in Fig. 8c for the cover versus σn-s (interface shear) or AC τ70 versus AC σn-s (internal shear); (iii)

794
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

Fig. 4. Photographs of post-shear specimens of GCL3 from internal shear tests

conducted under shearing normal stresses (σn-s) of (a) 160 kPa and (b)
1000 kPa.

Fig. 3. Relationships between shear stress (τ) and horizontal displacement (δh) at low σn-s to internal rupture of fibers at high σn-s (Gilbert et al., 1996;
for direct shear tests on the internal shear strength of GCLs: (a) GCL1, (b) GCL2,
Bacas et al., 2013; Thielmann et al., 2016; Bareither et al., 2018). The
and (c) GCL3.
δh-p of GCL1 was always higher than GCL2 for a given σn-s, and showed a
slight reduction as σn-s increased from 1000 to 2000 kPa (Fig. 9b). These
shear displacement at peak strength – δh-p versus σn-s; and (iv) post-peak differences in trends of δh-p may be attributed to differences in peel
shear reduction – Rτ versus σn-s. In addition, post shear photographs of strength or carrier geotextile (woven vs. non-woven) between the two
select specimens from the different GCL/GMX composition systems an­ GCLS, but additional testing is needed to more specifically determine the
alyses are included in Figs. E1 through E6 (Electronic Annex) to support mechanism. The inflection point for the bi-linear strength envelopes at
observations of the identified failure modes (Table 4). σn-s = 500 kPa maybe an artifact of the normal stresses used in this
study. Identification of a real inflection point was not the focus of this
6.1. Internal shear behavior of needle-punched GCLs study and requires a larger number of shear tests at varying normal
stress.
Internal shear behavior of GCL1, GCL2, and GCL 3 is shown in Fig. 9. Large-displacement shear behavior of GCL1 and GCL2 (Fig. 9c) are
Data from GCL3 are only included for the τp versus σn-s relationship shown with bi-linear strength envelopes with respect to σn-s ≤ 500 kPa
(Fig. 9a) due to unsuccessful internal failure. In addition, τp for GCL3 and σn-s ≥ 500 kPa. The bi-linear large-displacement shear strength
represent the maximum measured shear stress, which should not be envelopes for GCL1 and GCL2 were based on area corrected shear and
assumed the internal peak shear strength. normal stresses and φs-ld decreased with increasing σn-s.
Bi-linear shear strength envelopes for τp were defined for GCL1 and The residual shear strength of NP GCLs is controlled by bentonite
GCL2 with respect to σn-s ≤ 500 kPa and σn-s ≥ 500 kPa. Friction angles shear strength. This is also true, for the most part, at large shear
(φp) and cohesion intercepts (cp) for peak strength are included in displacement as failure of reinforcement fibers initiated at lower shear
Fig. 9a. An increase in peel strength from GCL1 (PS = 980 N/m) to GCL2 displacement (i.e., at or around peak shear strength). The shear strength
(PS = 2180 N/m) increased peak shear strength for a given normal of hydrated bentonite is affected by numerous factors, including
stress. Although internal failure of GCL3 was unsuccessful, the mineralogical properties of bentonite (i.e., percentage of sodium-
maximum shear stress for GCL3 was higher than the peak shear strength montmorillonite), chemical properties of hydration solution and clay
of GCL1 and GCL2. The increase in shear strength with increasing peel pore fluid, hydration and consolidation procedure prior to shearing,
strength agrees with previous research (e.g., Athanassopoulos and Yuan moisture content of the bentonite at shearing, shearing normal stress,
2011). temperature, etc. (Mesri and Olson, 1970; Müller-Vonmoos, and Løken,
The δh-p versus σn-s relationships in Fig. 9b depict a change at σn-s ≈ 1989; Anson and Hawkins, 1998; Dellisanti et al., 2018). Therefore,
500 kPa, whereby δh-p decreases for σn-s ≤ 500 kPa and then is these factors are also expected to affect the large-displacement and re­
approximately constant at higher σn-s. This bi-linear behavior coincided sidual shear strength of GCLs.
with the change in slope of the strength envelopes at σn-s ≈ 500 kPa and Secant friction angles for large-displacement shear strength of GCL1
was attributed to a change from disentanglement of reinforcement fibers and GCL2 ranged between 2.3◦ and 7.3◦ (Table 3) and, which is within

795
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

Fig. 5. Relationships between shear stress (τ) and horizontal displacement (δh) Fig. 6. Relationships between shear stress (τ) and horizontal displacement (δh)
for direct shear interface tests on (a) cover GT of GCL1 and GMX1, (b) carrier for direct shear interface tests on (a) carrier GT of GCL2 and GMX1, (b) cover
GT of GCL1 and GMX1, and (c) cover GT of GCL1 and GMX2. GT of GCL2 and GMX1, and (c) cover GT of GCL2 and GMX2.

the range of φs-ld reported in previous studies (e.g., Fox et al., 1998; Fox
and Stark 2015) on GCLs with similar materials and experimental pro­
cedure. The bi-linear large-displacement strength envelopes in Fig. 9c
indicate that GCL2 exhibited a more pronounced decrease in φs-ld at σn-s
≥ 500 kPa. The lower φs-ld was attributed to presence of a non-woven
geotextile (Table 1), which likely produced a bentonite-geotextile
interface that had lower friction resistance relative to GCL1.
The shear strength reduction ratio plotted in Fig. 9d represents the
amount of post-peak strength reduction in NP GCLs. The higher Rτ of the
GCL1 was attributed to the comparable area-corrected large displace­
ment shear strength for both GCLs, but lower peak shear strength for
GCL1 compared to GCL2.

6.2. Effect of geotextile mass per area on GCL/GMX shear behavior

The effect of geotextile mass per area on the shear behavior of GCL/
GMX composite systems was evaluated via interface shear tests with
GCL1 and GMX1. The non-woven carrier geotextile of GCL1 had mass
per area = 260 g/m2, whereas the non-woven cover geotextile of GCL1
had mass per area = 230 g/m2. Interface shear tests were conducted
with the carrier and cover geotextile of GCL1 in contact with GMX1 to
isolate mass per area of a non-woven geotextile.
Fig. 7. Relationships between shear stress (τ) and horizontal displacement (δh)
Shear behavior interface shear tests on GCL1 with the carrier and for interface shear tests on (a) carrier GT of GCL3 and GMX1, and (b) cover GT
cover geotextile in contact with GMX1 are shown in Fig. 10. Peak in­ of GCL3 and GMX2.
ternal shear strength of GCL1 was higher than peak shear strength
measured in the interface shear tests on GCL1 with either the cover or
σn-s = 2000 kPa where internal failure was observed. The slightly
carrier geotextile in contact with GMX1. This indicates that the critical
smaller τp measured for the interface shear test involving the carrier
shear strength of the GCL1/GMX1 composite system depended on the
geotextile of GCL1 and GMX1 relative to the internal τp of GCL1 at σn-s ≤
interface. Post-shear specimen inspection revealed that either complete
2000 kPa may be attributed to the lower stiffness of the GMX1 spikes
interface or partial interface/internal occurred for all test specimens,
compared to the metal teeth of the pyramid-tooth plates.
with exception of the carrier geotextile of GCL1 in contact with GMX1 at

796
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

At low normal stress, interlocking between the non-woven geotextile

and GMX spikes was superficial, such that frictional resistance between
the GMX and geotextile was the dominant shear resistance mechanism.
Therefore, differences in mass per area of the non-woven geotextiles in
contact with the GMX did not result in considerably different shear
resistance. As normal stress increased, interlocking between the non-
woven geotextile and GMX spikes was more effective, and an increase
in non-woven geotextile mass per area enhanced interlocking. This
enhanced interlocking due to an increase in mass per area of a non-
woven geotextile increased interface shear resistance that ultimately
transitioned the failure mode to complete internal failure of the GCL at
σn-s = 2000 kPa.

6.3. Effect of geotextile type on GCL/GMX shear behavior

The effect of geotextile type on shear behavior of GCL/GMX com­

posite systems was evaluated using GCL2 and GMX1. The cover geo­
textile of GCL2 was non-woven, whereas the carrier geotextile was
woven. Interface shear tests were conducted with the cover geotextile
and carrier geotextile of GCL2 in contact with GMX1 to isolate the
variable of geotextile type.
Shear behavior for interface shear tests on GCL2 and GMX1 are
shown in Fig. 11. The failure mode for the GCL2/GMX1 interface shear
tests was complete interface failure for both geotextiles at all normal
stresses (Table 4). Therefore, critical shear strength for both composite
systems was controlled by the interface shear strength between the
carrier or cover geotextile of GCL2 and GMX1.
The interface shear tests on GCL2/GMX1 with the cover and carrier
geotextiles in contact with GMX1 yielded nearly identical shear
behavior (i.e., τp, τld, δh-p, and Rτ) for σn-s ≤ 500 kPa (Fig. 11). Negligible
differences in shear behavior at low normal stress with respect to the
non-woven versus woven geotextiles was due to superficial interlocking
between the GMX and geotextiles such that the dominant shear resis­
tance mechanism was friction. However, for normal stress ≥1000 kPa, τp
and τld were larger for the GMX1 interface with the non-woven geo­
textile compared to the woven geotextile. The increase in shear strength
Fig. 8. Photographs of post-shear GCL specimens from interface shear tests that was attributed to enhanced interlocking between the non-woven cover
show examples of (a) complete interface failure, (b) complete internal failure, geotextile and GMX1 as normal stress increased.
and (c) partial interface/internal failure. Materials tested and shearing normal Shear resistance for the interface between GMX1 and the woven
stress (σn-s) are identified with each photograph. geotextile of GCL2 was frictional for the entire range of normal stress.
The predominance of frictional resistance yielded essentially linear re­
Shear behavior of the GCL1/GMX1 composite systems with either lationships of τp and τld as a function of σn-s, as well as nearly constant
the carrier or cover geotextile in contact with GMX1 was similar for σn-s values of δh-p and Rτ as a function of σn-s. Thus, for a given normal stress
≤ 500 kPa, and all tests exhibited complete interface failure (Table 4). the mobilized shear stress was proportional to frictional resistance at the
However, shear behavior of these two composite systems was different interface, required the same amount of deformation to mobilized peak
relative to the internal shear behavior. Lower τp and δh-p as well as strength, and yielded the same post-peak strength reduction. Assuming
higher τld and Rτ were observed for composite systems relative to in­ complete interface failure for a given GCL/GMX composite system, as
ternal shear. The lower δh-p was due to quick mobilization of the normal stress increases, a non-woven geotextile in contact with the GMX
maximum developed shear stress at the GCL/GMX interface that allowed appears to lead to higher critical shear strength relative to a woven
interface slippage relative to the available internal peak strength of geotextile. A non-woven geotextile has a greater ability to interlock with
GCL1. The higher τld of the composite systems relative to τld of GCL1 was asperities of a textured geomembrane, which contributed to higher
due to large-displacement shearing at the interface of the geotextile and interface shear resistance observed for the non-woven cover versus
GMX1 versus shearing between hydrated bentonite and geotextile. The woven carrier geotextiles in the GCL2/GMX1 composite systems.
higher τld and lower τp of the composite systems resulted in higher Rτ
relative to internal shear strength of GCL1.
6.4. Effect of GMX spike density on GCL/GMX shear behavior
The influence of geotextile mass per area on the shear behavior of
GCL1/GMX1 was most observed for interface shear tests at σn-s = 1000
The effect of GMX spike density on the shear behavior of GCL/GMX
kPa and 2000 kPa. At σn-s = 1000 kPa, complete interface failure was
composite systems was evaluated in two cases: Case 1 – cover geotextile
identified for the lower mass per area cover geotextile in contact with
of GCL1 sheared in contact with GMX1 and GMX2; and Case 2 – cover
GMX1 as opposed to partial interface/internal failure for the higher
geotextile of GCL2 sheared in contact with GMX1 and GMX2. Although
mass per area carrier geotextile. Due to the transition in failure mode to
GMX1 and GMX2 had the same spike height, polymer type, and nominal
partial interface/internal failure, the large-displacement shear strength
thickness, GMX1 spike density = 339 spikes/100-cm2 and GMX2 spike
envelope that included the carrier geotextile of GCL1 approaches the
density = 841 spikes/100-cm2. The effect of GMX spike density was
large-displacement strength envelope of GCL1 at σn-s ≥ 1000 kPa
evaluated in two cases because spike density has slightly different effects
(Fig. 10c).
on composite systems comprising different GCLs.

797
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

Fig. 9. Shear strength and shear behavior for internal shear tests on GCL1, GCL2, and GCL3: (a) peak shear strength (τp) versus shearing normal stress (σn-s), (b)
horizontal displacement at peak shear stress (δh-p) versus σn-s, (c) area-corrected large-displacement shear stress at δh = 70 mm (AC τ70) versus area-corrected large-
displacement normal stress at δh = 70 mm (AC σ70), and (d) strength reduction ratio (Rτ) versus σn-s.

Fig. 10. Shear strength and shear behavior for internal shear tests on GCL1 and interface shear tests involving the carrier and cover geotextiles of GCL1 sheared
against GMX1: (a) peak shear strength (τp) versus shearing normal stress (σn-s), (b) horizontal displacement at peak shear stress (δh-p) versus σn-s, (c) large
displacement shear stress (τld) versus large-displacement normal stress (σld), and (d) strength reduction ratio (Rτ) versus σn-s.

6.4.1. Case 1: GCL1/GMX1 and GCL1/GMX2 kPa and complete internal failure for σn-s ≥ 1000 kPa. In contrast, τp for
Shear behavior for interface shear tests involving the cover geotextile GCL1/GMX1 consistently plotted below peak internal shear strength.
of GCL1 sheared against GMX1 and GMX2 are shown in Fig. 12. Peak Post-shear inspection indicated that failure of the composite system was
shear strength of the GCL1/GMX2 composite system yielded comparable complete interface failure for σn-s ≤ 1000 kPa and partial interface/in­
τp to GCL1 internal shear strength. Post-shear specimen inspection ternal failure at σn-s = 2000 kPa.
revealed that failure was partial interface/internal failure for σn-s ≤ 500 Linear peak and large-displacement shear strength envelopes for

798
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

Fig. 11. Shear strength and shear behavior for internal shear tests on GCL2 and interface shear tests involving the carrier and cover geotextiles of GCL2 sheared
against GMX1: (a) peak shear strength (τp) versus shearing normal stress (σn-s), (b) horizontal displacement at peak shear stress (δh-p) versus σn-s, (c)) large
displacement shear stress (τld) versus large-displacement normal stress (σld), and (d) strength reduction ratio (Rτ) versus σn-s.

Fig. 12. Shear strength and shear behavior for internal shear tests on GCL1 and interface shear tests involving the cover geotextile of GCL1 sheared against GMX1
and GMX2: (a) peak shear strength (τp) versus shearing normal stress (σn-s), (b) horizontal displacement at peak shear stress (δh-p) versus σn-s, (c) large displacement
shear stress (τld) versus large-displacement normal stress (σld), and (d) strength reduction ratio (Rτ) versus σn-s.

GCL1/GMX1, as well as nearly constant values of δh-p and Rτ as a An increase in spike density yielded higher peak shear strength of
function of normal stress, indicate that the failure mode of GCL1/GMX1 GCL1/GMX2 compared to GCL1/GMX1 due to an enhanced friction and
remained predominantly the same for the range of normal stress. interlocking between the cover geotextile of GCL1 and GMX2. The
Although interlocking between the GMX1 spikes and non-woven geo­ higher interface friction and interlocking in the GCL1/GMX2 composite
textile of GCL1 was likely enhanced at higher σn-s, the shear resistance system resulted in a comparable shear behavior measured in the inter­
from friction and interlocking was insufficient to yield complete internal face shear tests at σn-s ≥ 1000 kPa as was measured in the internal shear
failure. Interlocking increased at σn-s = 2000 kPa based on partial tests on GCL1.
interface/internal failure in the post-shear specimen and slight increase
in δh-p, which supports increased interface resistance and a transition 6.4.2. Case 2: GCL2/GMX1 and GCL2/GMX2
towards shear resistance from internal reinforcement fibers. Shear behavior for interface shear tests involving the cover geotextile

799
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

Fig. 13. Shear strength and shear behavior for internal shear tests on GCL2 and interface shear tests involving the cover geotextile of GCL2 sheared against GMX1
and GMX2: (a) peak shear strength (τp) versus shearing normal stress (σn-s), (b) horizontal displacement at peak shear stress (δh-p) versus σn-s, (c) large displacement
shear stress (τld) versus large-displacement normal stress (σld), and (d) strength reduction ratio (Rτ) versus σn-s.

of GCL2 sheared against GMX1 and GMX2 are shown in Fig. 13. The
peak shear strength of the GCL2/GMX1 composite system was consis­
tently lower than interface peak shear strength of GCL2/GMX2 and in­
ternal shear strength of GCL2. Post-shear specimen inspection revealed
that failure of the composite system of GCL2 in contact with GMX1 was
complete interface failure for all normal stress. In contrast, GCL2/GMX2
yielded comparable peak shear strength to GCL2 internal shear strength
at σn-s ≥ 1000 kPa. Post-shear analysis revealed that the GCL2/GMX2
composite system transitioned from complete interface failure at σn-s =
100 kPa, to partial interface/internal failure for 300 kPa ≤ σn-s ≤ 1000
kPa, and to complete internal failure for σn-s = 2000 kPa.
The linear relationships of peak and large-displacement shear
strength as a function of normal stress, as well as the nearly constant
values of δh-p and Rτ as a function of normal stress, for GCL2/GMX1
indicate that the failure mode remained the same at all normal stress.
However, the transition from complete interface failure to partial
interface/internal failure in GCL2/GMX2 at 300 kPa ≤ σn-s ≤ 1000 kPa,
was characterized by a transition of τp, δh-p and Rτ towards GCL2 in­
ternal shear behavior. Complete internal failure in GCL2/GMX2 at σn-s
= 2000 kPa resulted in comparable shear behavior to the GCL2 internal
shear tests.
The high spike density GMX2 in contact with GCL1 (Fig. 12) and
GCL2 (Fig. 13) resulted in a higher peak shear strength measured in the
interface shear tests regardless of the normal stress. This was due to the
enhanced friction and interlocking between the GMX and non-woven
geotextiles of the GCLs. The shear resistance of GCL1 and GCL2 in
contact with GMX2 ultimately transitioned to complete internal failure
at high σn-s, by which shear strength of the composite system then relied Fig. 14. Shear strength relationships for interface shear tests involving the
on the internal shear strength of the GCL. cover geotextiles of GCL1, GCL2, and GCL3 sheared against GMX2: (a) peak
shear strength (τp) versus shearing normal stress (σn-s), and (b) large-
displacement shear strength at δh = 70 mm (τld) versus σn-s.
6.5. Effect of GCL peel strength on GCL/GMX shear behavior
Limiting all variables in Case 1 and Case 2 to only peel strength was
The effect of GCL peel strength on the shear behavior of GCL/GMX
not possible due to variability in geotextile characteristics. In Case 1,
composite systems was evaluated in two cases: Case 1 – non-woven
non-woven cover geotextiles of GCL1 and GCL2 were the same and mass
cover geotextiles of GCL1, GCL2, and GCL3 sheared in contact with
per area of the geotextiles were comparable to the non-woven cover
GMX2; and Case 2 – non-woven carrier geotextiles of GCL1 and GCL3
geotextile of GCL3 (230 g/m2 in GCL1 and GCL2 versus 260 g/m2 in
and the non-woven cover geotextile of GCL2 sheared in contact with
GCL3). The effect of GCL peel strength on shear behavior of GCL/GMX
GMX1.

800
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

for GCL1/GMX2 as σn-s increased from 500 kPa to 1000 kPa supports the
transition to complete internal failure, whereas the higher large-
displacement shear strength of GCL2/GMX2 supports that complete
internal failure did not occur at σn-s = 1000 kPa. Full mobilization of
GCL internal shear resistance for both GCL1/GMX2 and GCL2/GMX2
was observed at σn-s = 2000 kPa, which corresponded to complete in­
ternal failure. The higher peak shear strength with GCL2 was attributed
again to higher peel strength. Finally, the large-displacement shear
strength of both composite systems was comparable at σn-s = 2000 kPa,
which supports complete internal failure and comparable large-
displacement shear strength.

6.5.2. Case 2: GMX1/GCL1, GMX1/GCL2, and GMX1/GCL3

Similar to Case 1, the failure mode in Case 2 changed with increasing
normal stress. GCL1 exhibited complete interface failure for σn-s ≤ 500
kPa, partial interface/internal failure at σn-s = 1000 kPa, and complete
internal failure at σn-s = 2000 kPa. GCL2 exhibited was complete
interface for all σn-s. GCL3 exhibited complete interface failure for σn-s ≤
500 kPa and partial interface/internal failure for σn-s ≥ 1000 kPa
(Table 4 and Fig. 15).
Peak shear strength envelopes of all composite systems were com­
parable at σn-s ≤ 500 kPa, and the peak shear strength envelopes of GCL1
and GCL2 sheared against GMX1 were comparable at all normal stress
(Fig. 15a). Similarity of τp for all three composite systems in Case 2 at σn-
s ≤ 500 kPa indicates similar shear resistance developed at the interface.
GMX1 had a low spike density and the geotextiles of GCL1, GCL2, and
GCL3 used in Case 2 varied with respect to mass per area; thus, shear
resistance was predominantly frictional and negligible GMX-geotextile
interlocking developed with the increased mass per area geotextiles
under low normal stress. The different peel strengths of the GCLs nor the
Fig. 15. Shear strength relationships for interface shear tests involving the mass per area of the geotextiles yielded different shear strength in
carrier geotextiles of GCL1 and GCL3, and cover geotextile of GCL2 sheared
composite systems with GMX1 at σn-s ≤ 500 kPa.
against GMX1: (a) peak shear strength (τp) versus shearing normal stress (σn-s),
At σn-s = 1000 kPa, peak shear strength of the composite systems in
and (b) large-displacement shear strength at δh = 70 mm (τld) versus σn-s.
Case 2 remained the same despite the difference in observed failure
modes. An increase in σn-s to 2000 kPa resulted in higher τp and τld for
composite systems was isolated to the extent possible in Case 1. In
GCL3/GMX1 compared to GCL1/GMX1 and GCL2/GMX1 (Fig. 15). The
contrast, in Case 2 the mass per area of non-woven geotextiles ranged
higher τp and τld of the GCL3/GMX1 composite system suggests that the
from 230 g/m2 in GCL2 up to 360 g/m2 in GCL3. In this case, the
larger mass per area of the non-woven carrier geotextile of GCL3 in
evaluation focused more on the effect of varying GCLs on the shear
contact with GMX1 supported interlocking between the geotextile and
behavior of GCL/GMX composite systems considering that peel strength
GMX to increase interface shear resistance at σn-s = 2000 kPa.
and geotextile mass per area varied.
Variability in large-displacement shear strength at σn-s ≥ 1000 kPa of
the three GCL/GMX composite systems evaluated in Case 2 (Fig. 15b)
6.5.1. Case 1: GCL1/GMX2, GCL2/GMX2, and GCL3/GMX2
supports the transition in failure mode that highlights this influence of
Failure modes of the three GCLs sheared in contact with GMX2
peel strength. GCL1 had the lowest peel strength and gripping between
changed with increased normal stress (Fig. 14). GCL1 exhibited partial
the carrier geotextile of GCL1 in contact with GMX1 yielded that lowest
interface/internal failure for σn-s ≤ 500 kPa and complete internal fail­
τld that coincided with a transition to partial interface/internal failure at
ure at σn-s ≥ 1000 kPa. GCL2 exhibited complete interface failure at σn-s
σn-s = 1000 kPa and to complete internal failure at σn-s = 2000 kPa. In
≤ 300 kPa, partial interface/internal failure for 500 ≤ σn-s ≤ 1000 kPa,
contrast shearing along the interface between the cover geotextile of
and complete internal failure at σn-s = 2000 kPa. GCL3 exhibited com­
GCL2 and GMX1 yielded interface failure for all normal stress and higher
plete interface failure for σn-s = 100 kPa and partial interface/internal
τld relative to the τld of GCL1 that had lower peel strength. Finally, the
failure for σn-s = 300 and 500 kPa (Table 4).
higher peel strength combined with higher mass per area of the carrier
Relatively linear peak and large-displacement shear strength enve­
geotextile of GCL3 supported the highest τld for σn-s ≥ 1000 kPa due to
lopes were observed for σn-s ≤ 500 kPa for all composite systems in Case
enhanced shear resistance via a combination of friction, interlocking,
1, and values of τp and τld were comparable. This similarity was
and internal strength, mechanisms that are supported by the partial
attributed to complete interface or partial interface/internal failure
interface/internal failure of this composite system.
modes, which did not fully mobilize internal shear resistance of the
GCLs. Therefore, for the composite systems with GMX2, the increase in
6.5.3. Synopsis on effect of peel strength
the GCL peel strength from GCL1 to GCL2 to GCL3 did not contribute
The GCL peel strength could affect the peak and large-displacement
considerably to increased peak shear strength at σn-s ≤ 500 kPa.
shear strength of a GCL/GMX composite system only if interface shear
In contrast, the higher peel strength of GCL2 relative to GCL1
resistance is sufficient to engage internal reinforcement of the GCL such
increased peak shear strength for the composite systems with GMX2 at
that either partial interface/internal or complete internal failure
σn-s ≥ 1000 kPa. The lower internal shear resistance of GCL1 (PS = 980
develop. Peel strength appears to have negligible influence on peak and
N/m) resulted in complete internal failure at σn-s = 1000 kPa, whereas
large-displacement shear strength of a GCL/GMX composite system
higher internal shear resistance of GCL2 (PS = 2180 N/m) led to partial
when shear deformation and failure occurs along the interface between
interface/internal failure and development of a higher peak shear
the geotextile and GMX in a purely frictional manner. Therefore, the
strength (Fig. 14a). The decrease in large-displacement shear strength
effect of peel strength on shear behavior and shear strength of a GCL/

801
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

Fig. 16. Comparison between critical shear strength (τcr) and normal stress (σn) for the eight GCL/GMX composite systems evaluated via interface shear tests in
this study.

GMX composite system is related to characteristics of a GMX and geo­ • In composite systems with high peel strength GCLs, complete inter­
textile of the GCL that can increase interface shear resistance substan­ nal failure may not develop if the carrier and cover geotextile are not
tially to engage internal reinforcement of the GCL. sufficiently strong to transfer shear stress applied to the GCL/GMX
composite system into the internal region of the GCL.
7. Design consideration • Increase in the GMX spike density (i.e., texturing) can increase crit­
ical shear strength of GCL/GMX composite systems at all normal
Relationships of peak critical shear strength versus normal stress for stress unless the failure mode is complete internal failure. Spike
all eight GCL/GMX composite systems are shown in Fig. 16. Differences density enhances both friction and interlocking mechanisms to in­
in GCL peel strength, GMX spike density, and geotextile characteristics crease the developed interface shear stress between the GMX and
resulted in a broad range of peak critical shear strength. In particular, GCL.
the overal lowest and highest strength envelopes were observed for • Increase in the mass per area of a non-woven geotextile of the GCL in
composite systems with GCL2, whereby the lowest strength envelope contact with the GMX can increase critical shear strength of a GCL/
corresponded to the woven carrier geotextile of GCL2 with GMX1 and GMX composite system if interlocking is the predominant mecha­
the highest strength envelope corresponded to the non-woven cover nism to develop interface shear stress. The effect of geotextile mass
geotextile of GCL2 with GMX2. per area on critical shear strength is more pronounced at higher
In this study, evaluation of GCL and GMX physical properties on normal stress.
shear behavior and critical shear strength of GCL/GMX composite sys­ • Incorporating a GCL with a woven geotextile can result in a low
tems were evaluated via isoloating a single parameter to the extent critical shear strength in a GCL/GMX composite system. Superficial
possible. Considering the results in Fig. 16, a fundamental question interlocking between the GMX and woven geotextile limits the
would be what is the combined influence of certain GCL and GMX developed interface shear stress. Thus, GCLs with woven geotextiles
properties on shear behavior of GCL/GMX composite systems, or which are not recommended for applications with high shear stress.
GCL and GMX properties have the most pronounced influence on the • Peel strength of a GCL can influence the critical shear strength of a
critical shear strength of GCL/GMX composite systems? Based on the GCL/GMX composite system only if the majority of internal shear
experimental program of this studies and the materials tested, data resistance is mobilized. Considering that complete internal failure
presented in Fig. 16 highlights the pronounced impact of geotextile and usually occurs at high normal stress (e.g., ≥1000 kPa), an increase in
GMX characteristics on the shear strength of GCL/GMX composite sys­ peel strength more profoundly impacts critical shear strength of a
tems. However, additional research is required to prioritize the influ­ GCL/GMX composite system at high normal stress.
ence of different GCL and GMX properties on the shear behavior of GCL/
GMX composite systems. Declaration of competing interest

8. Conclusions The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
The objective of this study was to evaluate the effect of geotextile the work reported in this paper.
type, geotextile mass per area, GMX spike density, and GCL peel strength
on the shear behavior of GCL/GMX composite systems. These parame­ Acknowledgements
ters were evaluated via conducting direct shear tests to assess the in­
ternal shear strength of three GCLs and the interface shear strength of Support for this study was provided by Colorado State University
eight unique GCL/GMX composite systems consisting of the three GCLs (CSU), the Colloid Environmental Technologies Company (now Minerals
paired with two GMXs. The following are the main conclusions obtained Technology, Inc.) and the Geosynthetic Research Institute (GRI). Addi­
based on the results of this study: tional thanks are extended to Agru America, Inc. and Atarfil, Inc. for
providing material for this study. The opinions, findings, conclusions, or
recommendations expressed herein are those of the authors and do not

802
S. Ghazizadeh and C.A. Bareither Geotextiles and Geomembranes 49 (2021) 789–803

necessarily represent the views of CSU, Minerals Technology, Inc., GRI, Gilbert, R.B., Fernandez, F., Horsfield, D.W., 1996. Shear strength of reinforced GCLs.
Journal of Geotechnical Engineering 122 (4), 259–266.
Agru America, Inc, or Atarfil, Inc.
Guyonnet, D., Touze-Foltz, N., Norotte, V., Pothier, C., Didier, G., Gailhanou, H.,
Warmont, F., 2009. Performance-based indicators for controlling geosynthetic clay
Appendix A. Supplementary data liners in landfill applications. Geotext. Geomembranes 27 (5), 321–331.
Hanson, J.L., Chrysovergis, T.S., Yesiller, N., Manheim, D., 2015. Temperature and
moisture effects on GCL and textured geomembrane interface shear strength.
Supplementary data to this article can be found online at https://doi. Geosynth. Int. 22 (1), 110–124.
org/10.1016/j.geotexmem.2020.12.009. Hebeler, G.L., Frost, J.D., Myers, A.T., 2005. Quantifying hook and loop interaction in
textured geomembrane–geotextile systems. Geotext. Geomembranes 23 (1), 77–105.
Hewitt, R.D., Soydemir, C., Stulgis, R.P., Coombs, M.T., 1997. Effect of normal stress
References during hydration and shear on the shear strength of GCL/textured geomembrane
interfaces. In: Testing and Acceptance Criteria for Geosynthetic Clay Liners. ASTM
Allen, J.M., Fox, P.J., 2007. Pyramid-tooth Gripping Surface for GCL Shear Testing. International.
Geosynthetics ‘07. North American Geosynthetics Society, Washington, DC, (CD- Hillman, R.P., Stark, T.D., 2001. Shear strength characteristics of PVC geomembrane-
ROM). geosynthetic interfaces. Geosynth. Int. 8 (2), 135–162.
Anson, R.W.W., Hawkins, A.B., 1998. The effect of calcium ions in pore water on the Jones, D.R.V., Dixon, N., 1998. Shear strength properties of geomembrane/geotextile
residual shear strength of kaolinite and sodium montmorillonite. Geotechnique 48 interfaces. Geotext. Geomembranes 16 (1), 45–71.
(6), 787–800. Khilnani, K., Stark, T.D., Bahadori, T.M., 2017. Comparison of single and multi-layer
ASTM, 2015. Standard Test Method for Determining Average Bonding Peel Strength interface strengths for geosynthetic/geosynthetic and soil/geosynthetic interfaces.
Between Top and Bottom Layers of Needle-Punched Geosynthetic Clay Liners. In: Geotechnical Frontiers, pp. 42–51.
D6496/6496M. ASTM International, West Conshohocken, PA, USA. Koerner, R.M., Soong, T.-Y., Koerner, G.R., Gontar, A., 2001. Creep testing and data
ASTM D5261, 2010. Standard Test Method for Measuring Mass Per Unit Area of extrapolation of reinforced GCLs. Geotext. Geomembranes 19 (7), 413–425.
Geotextiles. D5261. ASTM International, West Conshohocken, PA, USA. Li, M.H., Gilbert, R.B., 2006. Mechanism of post-peak strength reduction for textured
ASTM D5993, 2014. Standard Test Method for Measuring Mass Per Unit Area of geomembrane–nonwoven geotextile interfaces. Geosynth. Int. 13 (5), 206–209.
Geosynthetic Clay Liners. D5993. ASTM International, West Conshohocken, PA, McCartney, J.S., Zornberg, J.G., Swan Jr., R.H., 2009. Analysis of a large database of
USA. GCL-geomembrane interface shear strength results. J. Geotech. Geoenviron. Eng.
ASTM D6072, 2015. Standard Practice for Obtaining Samples of Geosynthetic Clay 135 (2), 209–223.
Liners. D6072. ASTM International, West Conshohocken, PA, USA. Müller-Vonmoos, M., Løken, T., 1989. The shearing behaviour of clays. Appl. Clay Sci. 4
ASTM D6243-09. Standard Test Method for Determining the Internal and Interface Shear (2), 125–141.
Resistance of Geosynthetic Clay Liner by the Direct Shear Method, D6243, ASTM Mesri, G., Olson, R.E., 1970. Shear strength of montmorillonite. Geotechnique 20 (3),
International, West Conshohocken, Pennsylvania, USA. 261–270.
Athanassopoulos, C., Yuan, Z., 2011. Correlation between needle punched-reinforced Müller, W., Jakob, I., Seeger, S., Tatzky-Gerth, R., 2008. Long-term shear strength of
geosynthetic clay liner peel strength and internal shear strength. In: Geo-Frontiers geosynthetic clay liners. Geotext. Geomembranes 26 (2), 130–144.
2011: Advances in Geotechnical Engineering, pp. 1922–1930. Olsta, J.T., Swan, R.H., 2001. Internal Shear Strength of a Geosynthetic Clay Liner at
Bacas, B.M., Konietzky, H., Berini, J.C., Sagaseta, C., 2011. A new constitutive model for High Normal Loads, Tailings and Mine Waste. Colorado State University, Fort Collins,
textured geomembrane/geotextile interfaces. Geotext. Geomembranes 29 (2), CO, USA, pp. 197–200.
137–148. Robbins, M.C., Lu, H., Swift Jr., A.H.P., 1995. Investigation of the Suitability of a
Bacas, B.M., Cañizal, J., Konietzky, H., 2015. Shear strength behavior of geotextile/ Geosynthetic Clay Liner System for the El Paso Solar Pond (No. CONF-950725-).
geomembrane interfaces. Journal of Rock Mechanics and Geotechnical Engineering American Solar Energy Society, Boulder, CO (United States).
7 (6), 638–645. Ross, J.D., 2009. Static and Dynamic Shear Strength of a Geomembrane/geosynthetic
Bareither, C.A., Soleimanian, M., Ghazizadeh, S., 2018. Direct shear testing of GCLs at Clay Liner Interface. Doctoral dissertation, The Ohio State University.
elevated temperature and in a non-standard solution. Geosynth. Int. 25 (3), Ross, J.D., Fox, P.J., 2015. Dynamic shear strength of GCL/GMX composite liner for
350–368. monotonic loading. J. Geotech. Geoenviron. Eng. 141 (7), 04015026.
Chen, Y.M., Lin, W.A., Zhan, T.L., 2010. Investigation of mechanisms of bentonite Rowe, R.K., Garcia, J.D.D., Brachman, R.W.I., Hosney, M.S., 2019. Chemical interaction
extrusion from GCL and related effects on the shear strength of GCL/GM interfaces. and hydraulic performance of geosynthetic clay liners isothermally hydrated from
Geotext. Geomembranes 28 (1), 63–71. silty sand subgrade. Geotext. Geomembranes 47 (6), 740–754.
Chiu, P., Fox, P.J., 2004. Internal and interface shear strengths of unreinforced and Rowe, R.K., 2020. Geosynthetic clay liners: perceptions and misconceptions. Geotext.
needle-punched geosynthetic clay liners. Geosynth. Int. 11 (3), 176–199. Geomembranes 48 (2), 137–156.
Dellisanti, F., Calafato, A., Pini, G.A., Moro, D., Ulian, G., Valdrè, G., 2018. Effects of Sawada, Y., Nakazawa, H., Take, W.A., Kawabata, T., 2019. Full scale investigation of
dehydration and grinding on the mechanical shear behaviour of Ca-rich GCL damage mechanisms in small earth dam retrofit applications under earthquake
montmorillonite. Appl. Clay Sci. 152, 239–248. loading. Geotext. Geomembranes 47 (4), 502–513.
Eid, H.T., 2011. Shear strength of geosynthetic composite systems for design of landfill Siebken, J.R., Swan Jr., R.H., Yuan, Z., 1997. In: Well, L.W. (Ed.), Short-term and Creep
liner and cover slopes. Geotext. Geomembranes 29 (3), 335–344. Shear Characteristics of a Needle Punched Thermally Locked Geosynthetic Clay
Eid, H.T., Stark, T.D., Doerfler, C.K., 1999. Effect of shear displacement rate on internal Liner, STP 1308 Testing and Acceptance Criteria for Geosynthetic Clay Liners. ASTM
shear strength of a reinforced geosynthetic clay liner. Geosynth. Int. 6 (3), 219–239. International, West Conshohocken, PA, USA, pp. 89–102.
Feng, S.J., Shi, J.L., Shen, Y., Chen, H.X., Chang, J.Y., Wang, H.T., 2020. Experimental Sivakumar Babu, G.L., Sporer, H., Zanzinger, H., Gartung, E., 2001. Self-healing
study on the monotonic shear strength of GM/CCL composite liner interface. properties of geosynthetic clay liners. Geosynth. Int. 8 (5), 461–470.
Environmental Geotechnics 1–13. Stark, T.D., Choi, H., Akhtarshad, R., 2004. Occurrence and effect of bentonite migration
Fox, P.J., Rowland, M.G., Scheithe, J.R., 1998. Internal shear strength of three in geosynthetic clay liners. Geosynth. Int. 11 (4), 296–310.
geosynthetic clay liners. J. Geotech. Geoenviron. Eng. 124 (10), 933–944. Stark, T.D., Niazi, F.S., Keuscher, T.C., 2015. Strength envelopes from single and multi
Fox, P.J., Kim, R.H., 2008. Effect of progressive failure on measured shear strength of geosynthetic interface tests. Geotech. Geol. Eng. 33 (5), 1351–1367.
geomembrane/GCL interface. J. Geotech. Geoenviron. Eng. 134 (4), 459–469. Thielmann, S.S., Fox, P.J., Athanassopoulos, C., 2016. Shear strength of GCL/GMX
Fox, P.J., Ross, J.D., 2011. Relationship between GCL internal and GCL/GMX interface composite liner under high normal stress. J. Geotech. Geoenviron. Eng. 142 (5),
shear strength. J. Geotech. Geoenviron. Eng. 137 (8), 743–753. 1–11.
Fox, P.J., Stark, T.D., 2015. State-of-the-art report: GCL shear strength and its Trauger, R.J., Swan, R.H., Yuan, Z., 1997. In: Well, L.W. (Ed.), Long-term Shear Strength
measurement–ten-year update. Geosynth. Int. 22 (1), 3–47. Behavior of a Needle Punched Geosynthetic Clay Liner, STP 1308 Testing and
Fox, P.J., Nye, C.J., Morrison, T.C., Hunter, J.G., Olsta, J.T., 2006. Large dynamic direct Acceptance Criteria for Geosynthetic Clay Liners. ASTM International, West
shear machine for geosynthetic clay liners. Geotech. Test J. 29 (5), 392–400. Conshohocken, PA, USA, pp. 103–120.
Frost, J.D., Lee, S.W., 2001. Microscale study of geomembrane-geotextile interactions. Triplett, E.J., Fox, P.J., 2001. Shear strength of HDPE geomembrane/geosynthetic clay
Geosynth. Int. 8 (6), 577–597. liner interfaces. J. Geotech. Geoenviron. Eng. 127 (6), 543–552.
Ghazizadeh, S., Bareither, C.A., 2018a. Stress-controlled direct shear testing of Vukelić, A., Szavits-Nossan, A., Kvasnička, P., 2008. The influence of bentonite extrusion
geosynthetic clay liners II: assessment of shear behavior. Geotext. Geomembranes 46 on shear strength of GCL/geomembrane interface. Geotext. Geomembranes 26 (1),
(5), 667–677. 82–90.
Ghazizadeh, S., Bareither, C.A., 2018b. Stress-controlled direct shear testing of Zornberg, J.G., McCartney, J.S., Swan Jr., R.H., 2005. Analysis of a large database of GCL
geosynthetic clay liners I: apparatus development. Geotext. Geomembranes 46 (5), internal shear strength results. J. Geotech. Geoenviron. Eng. 131 (3), 367–380.
656–666. Zornberg, J.G., McCartney, J.S., 2009. Internal and interface shear strength of
Ghazizadeh, S., Bareither, C.A., 2020. Temperature effects on internal shear behavior in geosynthetic clay liners. In: Geosynthetic Clay Liners for Waste Containment
reinforced GCLs. J. Geotech. Geoenviron. Eng. 146 (1), 04019124. Facilities. CRC Press, pp. 149–174.

803