Composite Structures 201 (2018) 79–85

Contents lists available at ScienceDirect

Composite Structures
journal homepage: www.elsevier.com/locate/compstruct

Interlaminar failure behavior of GLARE laminates under double beam five- T

point-bending load
⁎
Yanyan Lina,1, Cheng Liua,1, Huaguan Lia, Kai Jinb, Jie Taoa,c,
a
College of Material Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, PR China
b
College of Mechanical and Electrical Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, PR China
c
Jiangsu Collaborative Innovation Center for Advanced Inorganic Function Composites, Nanjing 210016, PR China

A R T I C LE I N FO A B S T R A C T

Keywords: GLARE laminates were prepared in this work to characterize their interlaminar shear failure behavior by Double
DBS method Beam Shear (DBS) method. The effect of span-to-thickness ratio (L/h ratio) and lay-up configuration on the
GLARE interlaminar failure feature of the laminates was investigated respectively. The results indicated that the failure
ILSS values mode in DBS method was single interlaminar shear delamination due to the existence of pure shear stress points.
Interlaminar shear delamination
Furthermore, the apparent interlaminar shear strength (ILSS) values measured by DBS method were higher than
those of Short-beam Shear (SBS) method, which were closer to the true value. However, the two methods did not
show exclusiveness.

1. Introduction reliable method to evaluate the interlaminar shear resistance of GLARE,

GLARE, consisting of alternative layers of thin aluminum alloy
layers and glass/epoxy composites layers [1], has found its important
application in aeronautics industry due to excellent fatigue, impact and
damage tolerance. However, debonding and delamination may develop
between fiber/epoxy and metal layers in the case of bending load or
torsion, representing the typical failure modes of Fiber Metal Laminates
(FMLs) [2–5]. Apparently, an appropriate evaluation method for in-
terlaminar shear resistance plays a key role in the quality control of
FMLs [6]. Researchers mainly used Short-beam Shear (SBS) method
(ASTM-D2344) to measure the interlaminar shear behavior of GLARE
laminates [7–9] because a standard test method to estimate the inter-
laminar shear properties of GLARE has not been established [10–12].
However, the span-to-thickness ratio (L/h ratio) of 4 or 5 is usually
chosen according to ASTM-D2344 [13], which is inappropriate for
GLARE laminates. The main reason is that the SBS method is only ap-
plicable to the high-modulus fiber-reinforced composite materials
[14,15]. Thus, we have already investigated the interlaminar failure
behavior of GLARE-3/2-0.3 laminates under short-beam three-point-
bending load, and found that specimens with L/h ratio of 8 yielded a
relatively accurate interlaminar shear strength value [16].
Theoretically speaking, the interlaminar shear strength value of
GLARE obtained by SBS method is smaller than true value due to the
unavoidable influence of bending stress. Aimed at finding a more
the Double Beam Shear (DBS) method could be introduced. G. Zhou
et al. from Loughborough University conducted DBS method to char-
acterize the apparent interlaminar shear properties of resin matrix
composites [17]. They found that the DBS method guaranteed inter-
laminar shear failure between two pure shear points by promoting the
dominance of interlaminar shear strength (ILSS). Unlike the homo-
geneous resin matrix composites, GLARE laminates possess multiple
interfaces, including the interfaces between aluminum alloy sheet and
matrix, fiber and matrix, composite and aluminum alloy sheet [18].
Consequently, the fixture for evaluation of interlaminar shear behavior
should be designed and the key parameters (L/h ratio and lay-up con-
figuration) should be considered.
In this paper, failure mode was evaluated to determine the effective
interlaminar shear failure for GLARE laminates under double beam five-
point-bending load. The influence of the L/h ratio and the lay-up con-
figuration on the interlaminar shear failure behavior and mechanisms
under DBS method were investigated. Finally, the respective failure
characteristics and apparent interlaminar shear strength of GLARE la-
minates under DBS and SBS methods were compared.

⁎
Corresponding author at: College of Material Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, PR China.
E-mail address: taojie@nuaa.edu.cn (J. Tao).
1
These authors contributed equally to this work.

https://doi.org/10.1016/j.compstruct.2018.06.037
Received 14 September 2017; Received in revised form 16 May 2018; Accepted 7 June 2018
Available online 15 June 2018
0263-8223/ © 2018 Elsevier Ltd. All rights reserved.
Y. Lin et al.
Table 1
The fundamental performance of the Aluminum alloy (2024-T3).
Thickness Elasticity Coefficient of Yield
(mm) modulus thermal expansion strength
(GPa) (°C) (MPa)
0.3 73.1 22.68 × 10−6 305
2. Experimental
2.1. Materials preparation
The Aluminum alloy plates was 0.3 mm thick layers (2024-T3)
supported by Southwest Aluminum (Group) Co., Ltd, and the funda-
mental performance of the Aluminum alloy plates is given in Table 1.
Moreover, the surface treatment of the aluminum layers was conducted
to achieve a better bonding with the fiber reinforced composites. The
phosphoric acid anodizing (H3PO4 140 g/L, 25 °C, 10 V, 20 min) was
adopted to construct a rough surface. The high strength glass fiber
(HS4) was provided by Nanjing Glass Fiber Research Institute and the
high temperature epoxy resin (E302-2) was provided by Sichuan Xin-
wanxing Carbon Fiber Composites Co., Ltd, and the fundamental per-
formance of the S4-glass/epoxy composites is listed in Table 2. Alu-
minum alloy plates with a thickness of 0.3 mm, S4-glass/epoxy
composites consisting of two unidirectional prepregs with a nominal
thickness of 0.125 mm were used in constituent materials of alleged
GLARE-3/2-0.3 [19]. Four kinds of laminates were prepared as follows:
GLARE 2A-0/0-0.3, GLARE 3-0/90-0.3, GLARE 4A-0/90/0-0.3 and
GLARE 6- ± 45-0.3. Specimens with dimensions of 40 mm × 10 mm
were cut from the massive FMLs by using wire electrical discharge
machining (WEDM) technique. Additionally, to achieve a better po-
lished surface, all specimens were processed by carbide end milling
cutter [20]. Each group contained five qualified specimens without
interfacial debonding.
2.2. Experiment procedures
Double Beam Shear method is not the standard method for inter-
laminar shear test of composite materials, and the corresponding fixture
cannot be acquired directly. Therefore, the five-point-bending fixture
was designed for DBS method as shown in Fig. 1(a) and the dimensions
of the indenter are illustrated in Fig. 1(b). The material chosen for
fixture is Cr12MoV, which is widely utilized in the production of
standard tools and gauges. Convenient to adjust the indenters and the
holders, the rail type is designed for the precise control of L/h ratio.
Fig. 2(a) shows that the double beam five-point-bending test is
conducted in a beam specimen of rectangular cross section symme-
trically supported by three holders and loaded by two indenters, which
the distance (L) between the two holders is equal to the indenter span.
The five-point-bending test is named as DBS method since two short
beams are formed by three equidistant holders. It also needs to choose
the appropriate span-to-thickness ratio, so that the effective inter-
laminar shear failure occurs when the specimen is subjected to a small
bending disturbance [21]. The existence of two pure shear points in the
beam specimen can assure pure shear stress under double beam five-
point-bending load [22], as indicated in Fig. 2(b). In addition, the
loading rate of the experiment is 1 mm/min. According to DBS method
theory for homogeneous material [23], the interlaminar shear strength
is given by
Table 2
The fundamental performance of the S4-glass/epoxy composites.
Surface Density (g/cm3) Resin content (wt%) Thickness (mm)
170 ± 5 30 ± 2 0.125 ± 0.05
Composite Structures 201 (2018) 79–85
33P
τ=
64bt (1)
Tensile
where P represent the load in response to the occurrence of delami-
strength
(MPa) nation, b and t denote the average measuring width and thickness of the
specimen, respectively.
455 Indeed, the Eq. (1) is based on the assumption that the interlaminar
shear failure occurs at the centreplane when the shear stress reaches the
maximum. However, the centreplane for GLARE laminates is the alu-
minum layer and its shear strength is larger than that between the
aluminum sheets and GF/EP layer. We still used the formula applied for
homogeneous material due to the following reasons.
Firstly, according to our experimental results below, the debonding
is near the centreplane instead of the other interface between aluminum
sheets and GF/EP layer. The error is quite limited in this situation.
Secondly, the values deflection of ILSS also exists in this kind of ma-
terial, including fiber composite material, because the bending stress is
unavoidable. It is the reason that interlaminar shear strength obtained
by these methods is still called the apparent ILSS. Actually, the newly
developed DBS method is less affected by bending stress than the tra-
ditional SBS method. In this way, the relatively more accurate results
could be obtained. Thirdly, the stacking structure of all the GLARE la-
minates analyzed in this work was 3/2 configuration. There are some
additional structures in GLARE, like 4/3, 5/4, etc. We intended to seek
a relatively reasonable evaluation methodology that could be widely
used in fiber mental laminates. For the sake of the unity of the formula,
we referred to the formula for calculating the ILSS of homogeneous
material.
3. Failure behavior and mechanisms
3.1. Failure mode and mechanisms
Compared to the multiple failure mode of GLARE laminates in SBS
method, it is single expression of interlaminar shear delamination in
DBS method. In Fig. 3, the specimens on both sides have no evident
shear failure. Near the neutral layer, between the two indenters and
intermediate holder, effective interlaminar shear failure at the interface
of aluminum alloy and composite layer occurs in four kinds of GLARE.
The reason is that the large residual stress caused by the difference in
the coefficient of expansion exists in the interface, where is prone to
generate failure [24]. However, the existence of pure shear stress points
weakens the influence of bending stress on failure behavior, which
causes the specimens to fail under the small displacement. Namely, the
whole shear failure process is completed near the elastic section, and no
large local compression deformation occurs.
As can be seen clearly from the Fig. 4, the typical failure process of
GLARE-3/2 under DBS method present two-stage failure, which could
be divided into four parts. The first part is an elastic section, and the
GLARE specimen shows an elastic deformation at this stage, that is,
after unloading the applied load, the sample can be completely re-
covered without any failure. Before the peak load occurs, the yield of
aluminum sheets causes the sample enter the plastic deformation stage.
Due to the excellent interfacial bonding of the GLARE laminates, the
bonding effect between glass fiber reinforced composite layer and
aluminum alloy layer continues to maintain good adhesion, and the
specimen does not show any failure. Because of the interfacial de-
bonding, the curve indicates a sharp decline when the applied load
reaches the maximum and at this point, the calculated ILSS is the ap-
parent interlaminar shear strength. Subsequently, the applied load in-
creases at a relatively small slope until the end of the debonding failure
to further overcome the interfacial cohesion. In the last stage, the la-
Width (mm) minates were in bending deformed state after the failure subjected to
the transverse shear stress completion.
1000
80
Y. Lin et al.
Fig. 1. (a) The design sketch of fixture for DBS method, (b) the dimension parameter of indenter for DBS method. Dimensions are in millimeters.
Fig. 2. (a) The apparatus for the double beam five-point-bending test, (b) Stress
analysis in DBS method.
Fig. 3. Typical failure mode of GLARE-3/2 under DBS method.
3.2. Influence of L/h ratio on interlaminar shear failure behavior
The correlation between apparent ILSS and L/h ratio for various lay-
up configuration is proved in Fig. 5. Experimental results illustrate the
fact that ILSS values drop with the increase of L/h ratio. The inter-
laminar shear strength obtained by DBS method is still called the ap-
parent ILSS due to the slight influence of bending stress. Therefore, it is
necessary to adjust an appropriate L/h ratio to characterize the ap-
parent ILSS of GLARE laminates. Owing to unidirectional fiber stacking
in GLARE 2A, the apparent ILSS value decreases linearly with L/h ratio
increase. It can be concluded that the L/h ratio has a significant effect
on the failure behavior of unidirectional laminates under the condition
of double beam loading. Besides, the apparent ILSS value of GLARE 4A
laminates reduces quickly at first and slow down after the L/h ratio is 7.
However, the decline in apparent ILSS value of GLARE 3 and GLARE 6
81
Composite Structures 201 (2018) 79–85
Fig. 4. Typical failure process of GLARE-3/2 under DBS method.
Fig. 5. The correlation between apparent ILSS and L/h ratios for GLARE.
tends to be gentle after the L/h ratio is 6, and when the L/h ratio arrives
at 7, there is no obvious decline in apparent ILSS values. The reason for
such trend in different lay-up configuration can be attributed to the
analogous crisscross fiber stacking, which possesses better bending
stress resistance. Accordingly, the large L/h ratio not only alleviate the
stress concentration, but also lessen the bending influence on the
measurement of apparent ILSS values.
The specimens are subjected to bending and shear coupling effect,
which is dissimilar with the change of L/h ratio. Fig. 6 depicts the load-
deflection curves of various GLARE laminates under four different L/h
ratios. Under the L/h ratio of 5, the load values of GLARE laminates are
much higher than others and the failure processes are complex. Fiber
breakage and matrix cracking took place, as shown in Fig. 7(a) and (b).
Y. Lin et al. Composite Structures 201 (2018) 79–85

Fig. 6. The load-deflection curves under four different L/h ratios: specimens of (a) GLARE 2A, (b) GLARE 3, (c) GLARE 4A, (d) GLARE 6.

Although the interlaminar shear stress accounts for a sizeable propor- desirable. In addition, there are perforative cracks along 90°-fiber layer
tion, it is easy to form stress concentration. Owing to the low fracture in Fig. 7(b) and (d). This is caused by the relatively weak adhesion
toughness of resin matrix and glass fibers, the failure was not domi- between resin matrix and glass fibers under the DBS loading condition.
nated by the interlaminar shear delamination, the L/h ratio of 5 was not When L/h ratio is 8, the load-deflection curves of four GLARE

Fig. 7. The failure microgram of GLARE specimens: (a) and (b) for L/h ratio of 5, (c) and (d) for L/h ratio of 8.

82
Y. Lin et al.
Fig. 8. The load-deflection curves of four different lay-up GLARE specimens: (a) for L/h ratio of 5, (b) for L/h ratio of 8.
variants are of typical two sections, representing the occurrence stable
failure of interlaminar shear delamination, as presented in Fig. 7(c) and
(d). The first peak is initial debonding formed by the effect of inter-
laminar shear stress, and the formation of second peak is delamination
between pure shear points. In practice, the results demonstrate that
effective interlaminar shear failure occurs at the first peak, and the
corresponding load values could be used to calculate the apparent ILSS
values of GLARE laminates.
3.3. Influence of lay-up configuration on interlaminar shear failure
behavior
The mechanical behavior of GLARE laminates is mostly influenced
by the lay-up configuration under the DBS method. Fig. 8 (a) presents
the load-deflection curves of GLARE variants at L/h ratio of 5. Effective
interlaminar shear failure occurs at the first peak. Subsequently, owing
to its low L/h ratio, GLARE laminates are easy to be compressed and
buckled lacking stability, displaying irregular fluctuation of the load-
deflection curves [25]. It is worthy of noting that the mechanical be-
havior of all kinds of specimens with different lay-up configurations
under L/h ratio of 5 is complex. The specimens are subjected to a
variety of stresses, mainly controlled by interlaminar shear stress,
bending stress and shear stress which parallel to the loading direction.
It is shown in Fig. 8(b) that the load-deflection curves of all GLARE
variants at L/h ratio of 8 are regular. The interlaminar shear failure
occurred at the first peak of the curve, presenting the debonding be-
tween the middle aluminum alloy layer and the composite layer.
However, fiber layers are the primary bearing part in GLARE laminates,
and the fiber orientation varies in different lay-up configurations,
which determine the difference in the interlaminar failure feature of
laminates. Based on the analysis of four GLARE variants, GLARE 4A has
the largest load value and shortest failure deflection. This is due to the
highest thickness of GLARE 4A among four GLARE variants, which is
sensitive to the small bending disturbance [26]. Notably, GLARE 2A has
the largest failure deflection since the unidirectional fiber layers. 0°-
Fiber layers play a bridging role in the failure process and effectively
transfer the load [27]. In addition, the failure deflection of GALRE 3 is
relatively small due to the inclusion of 90°-fiber layers, since they are
not bearing part in the loading process and easy to crack under stress
condition. Moreover, the load and deflection values of GLARE 6 are
among four of them. ± 45° oblique crossing fiber layers possess better
stress distribution, where delamination between inner aluminum alloy
layer and composite layer happens.
4. Comparison between SBS and DBS methods
4.1. Failure characteristics in SBS and DBS methods
The stress state of DBS method attributes to the effect of bending
83
Composite Structures 201 (2018) 79–85
and shear coupling, which is similar to that of SBS method [28]. The
difference is that the pure shear points exist between intermediate
holder and two indenters under double beam five-point-bending load,
the effect of interlaminar shear force is stronger than that of bending
force, the shear failure is not difficult to happen under interlaminar
shear stress in that region, ensuring the reliability of the test results.
When the influence of bending stress is reduced, the apparent inter-
laminar shear strength values obtained by DBS method are relatively
larger.
The mechanical behaviors of four GLARE variants between SBS and
DBS method are compared in Fig. 9. It can be pointed out that the load
value of SBS method is lower than that of DBS method when the L/h
ratio is 5 or 8. Although the values in GLARE 2A with two methods are
considerably different, the failure deflections are close to each other, as
presented in Fig. 9(a). This is caused by the fiber direction of uni-
directional layers, which is identical with the direction of normal stress
and interlaminar shear stress, and the 0°-fiber layers are the main
bearing part.
Fig. 9(b) shows the load-deflection curves of GLARE 3 under two
methods. The failure deflections are almost same when L/h ratio of 5 or
8 under double beam five-point-bending load. After the external load
hitting the ultimate bearing of matrix, crack initiated between 90°-fiber
layer and aluminum alloy layer, causing interlaminar delamination and
matrix cracking phenomenon in 90°-fiber layers. Under two L/h ratios
manner of SBS method, the failure deflections are inconsistent. It in-
dicates that the stress concentration at L/h ratio of 5 is more profound,
leading to the earlier failure. In addition, the curve presents a fluc-
tuation state at L/h ratio of 5 and a stable expression of two sections at
L/h ratio of 8.
The configuration of fiber layer in GLARE 4A is 0°/90°/0°, and the
two unidirectional fiber layers are the main bearing part in the loading
process, leading to excellent interlaminar shear resistance. It can be
concluded in Fig. 9(c) that the load and the deflection values of GLARE
4A are largest than other GLARE variants at equal level. Under double
beam five-point-bending load, the failure deflections are almost the
same at different L/h ratios. It can be confirmed that the unidirectional
fiber layers are the principal bearing part during the loading process.
Obvious damage does not occur in 90°-fiber layers until the ILSS values
of GLARE 4A reach the shear stress limit.
Fig. 9(d) contains the load-deflection curves of GLARE 6 under SBS
and DBS methods. Among the consequences, the failure deflection at L/
h ratio of 8 is higher than that of 5. The debonding between the ± 45°
oblique crossing fiber layers and the matrix could be alleviated, and the
effect of shear stress would be delayed. However, interlaminar shear
failure occurs earlier in DBS method due to the dominant position of
interlaminar shear stress.
Y. Lin et al.
Fig. 9. The mechanical behavior comparison between SBS and DBS method: specimens of (a) GLARE 2A, (b) GLARE 3, (c) GLARE 4A, (d) GLARE 6.
4.2. Apparent interlaminar shear strength with SBS and DBS methods
The comparison of apparent interlaminar shear strength at L/h ratio
of 8 with two methods is showed in Fig. 10. It illustrates that the ap-
parent ILSS values measured by DBS method are much larger than those
obtained from SBS method. Particularly, for GLARE 6, the average
apparent ILSS value obtained for DBS method is 50.2% greater. In ad-
dition, it demonstrates that the stability of the consequences in DBS
method is better according to the relatively smaller dispersion degree.
This is attributed to the existence of pure shear points in DBS method,
which effectively alleviates the influence of the bending stress on test
results.
However, although pure shear failure could happen among the
modified SBS method specimens with well stability at L/h ratio of 8, it
is still severely subjected to bending stress. Thereby, the fiber rupture or
matrix cracking caused by the bending stress will affect the measured
Fig. 10. Apparent ILSS of SBS and DBS methods at L/h ratio of 8.
84
Composite Structures 201 (2018) 79–85
ILSS values [29]. SBS method is just simple and convenient in terms of
testing set-up to provide information of quality control and material
screening. In fact, based on the study of GLARE-3/2-0.3 in interlaminar
failure behavior, the ILSS values obtained from DBS method are closer
to the true values since the smaller influence of unavoidable bending
stress. Besides, the materials involved in DBS methods (specimens with
dimensions of 40 mm × 10 mm) are larger than those in SBS methods
(20 mm × 10 mm). When it comes to evaluating the interlaminar
failure behavior of GLARE laminates, appropriate test methods could be
selected according to the actual service conditions of samples.
5. Conclusions
(1) Effective interlaminar shear delamination was the single failure
mode in this study. The L/h ratio of 8 under DBS method that leads
to stable results for the load-deflection curves of four GLARE var-
iants, and the corresponding load values could be used to calculate
the apparent ILSS values of laminates.
(2) The apparent ILSS values measured by DBS method were higher
and more stable than those used in SBS method. The ILSS values
obtained from DBS method were closer to the true values since the
smaller influence of unavoidable bending stress. Appropriate test
methods could be selected according to the actual service condi-
tions of samples.
Acknowledgments
The authors gratefully acknowledge the support from the
Postgraduate Research & Practice Innovation Program of Jiangsu
Province (KYCX17-0259), the financial support of the National Natural
Science Foundation of China (51705235), the Jiangsu Province Natural
Science Foundation (BK20170762), the Fund of China Postdoctoral
Science Foundation funded project (2017M611758), the Key Project of
Key Laboratory of Railway Vehicle Door System in Jiangsu Province
Y. Lin et al.
(KN17-13), the Priority Academic Program Development of Jiangsu
Higher Education Institutions and the Fundamental Research Funds for
the Central Universities.
References
[1] Subbaramaiah R, Prusty BG, Pearce GMK, et al. Crashworthy response of fibre metal
laminate top hat structures. Compos Struct 2017;160:773–81.
[2] Yaghoubi AS, Liaw B. Effect of lay-up orientation on ballistic impact behaviors of
GLARE 5 FML beams. Int J Impact Eng 2013;54:138–48.
[3] Sugiman S, Crocombe AD. The static and fatigue response of metal laminate and
hybrid fibre-metal laminate doublers joints under tension loading. Compos Struct
2012;94:2937–51.
[4] Alderliesten RC. Analytical prediction model for fatigue crack propagation and
delamination growth in GLARE. Int J Fatigue 2007;29:628–46.
[5] Bienias Jaroslaw, Jakubczak Patryk, Dadej Konrad. Low-velocity impact resistance
of aluminium glass laminates-experimental and numerical investigation. Compos
Struct 2016;152:339–48.
[6] Sinmazçelik Tamer, Avcu Egemen, ÖzgürBora Mustafa, et al. A review: Fibre metal
laminates, background, bonding types and applied test methods. Mater Des
2011;32(7):3671–85.
[7] Remmers JJC, de Borst R. Delamination buckling of fiber-metal laminates. Compos
Sci Technol 2001;61(15):2207–13.
[8] Park Sang Yoon, Choi Won Jong, Choi Heung Soap, et al. Effects of surface pre-
treatment and void content on GLARE laminate process characteristics. J Mater
Process Technol 2010;210(8):1008–16.
[9] Hinz S, Heidemann J, Schulte K. Damage evaluation of GLARE®4B under inter-
laminar shear loading at different temperature conditions. Adv Compos Lett
2005;14(2):47–55.
[10] Wu Wangqing, Abliz Dilmurat, Jiang Bingyan, et al. A novel process for cost ef-
fective manufacturing of fiber metal laminate with textile reinforced pCBT com-
posites and aluminum alloy. Compos Struct 2014;108:172–80.
[11] Hinz Stephan, Omoori Toshihiro, Hojo Masaki. Damage characterisation of fibre
metal laminates under interlaminar shear load. Compos A Appl Sci Manuf
2009;40(6–7):925–31.
[12] Mujika F. On the difference between flexural modulus obtained by three-point and
four-point bending tests. Polym Test 2006;25:214–20.
[13] ASTM D2344-84. Standard test method for apparent interlaminar shear strength of
parallel fiber composites by short-beam method. Philadelphia: American Society for
Testing and Materials (ASTM); 1989.
85
Composite Structures 201 (2018) 79–85
[14] Standard A. Standard test method for short-beam strength of polymer-matrix
composite materials and their laminates. Annual book of ASTM standards Vol. 15.
2007. p. 54–60. West Conshohocken.
[15] Sideridis Emilio P, Bikakis George S. Shear properties and load-deflection response
of cross-ply glass-epoxy composite short-beams subjected to three-point-bending
tests, and the effect of moisture absorption. J Appl Polym Sci 2013;61:2244–52.
[16] Liu C, Du DD, Li HG, et al. Interlaminar failure behavior of GLARE laminates under
short-beam three-point-bending load. Compos B Eng 2016;97:361–7.
[17] Zhou G, Nash PH, Whitaker J, et al. Double beam shear (DBS) as a new test method
for determining interlaminar shear properties of composite laminates. In: ECCM16-
16TH European Conference on Composite Materials, Seville, June 22–6; 2014.
[18] Hu YB, Li HG, Cai L, et al. Preparation and properties of Fibre Metal Laminates
based on carbon fibre reinforced PMR polyimide. Compos B Eng 2015;69:587–91.
[19] Seyed Yaghoubi A, Liu Y, Liaw B. Stacking sequence and geometrical effects on low-
velocity impact behaviors of GLARE 5 (3/2) fiber metal laminates. Thermoplast
Compos Mater 2012;25(2):223–47.
[20] Ashrafi SA, Sharif S, Farid AA, et al. Performance evaluation of carbide tools in
drilling CFRP-Al stacks. Compos Mater 2013:1–14.
[21] Mujika F, Asensio M, Vargas G, et al. Analysis of a reversible five-point bending
configuration based on a novel two-sense support. Polym Test 2015;43:108–22.
[22] Lopez Juan Angel, Serna Pedro, Navarro-Gregori Juan. A simplified five-point in-
verse analysis method to determine the tensile properties of UHPFRC from un-
notched four-point bending tests. Compos B Eng 2016;91:189–204.
[23] Kim WC, Dharan CKH. Analysis of five-point bending for the determination of the
interlaminar shear strength of unidirectional composite materials. Compos Struct
1995;30:241–51.
[24] Reyes V G, Cantwell WJ. The mechanical properties of fiber-metal laminates based
on glass fiber reinforced polypropylene. Compos Sci Technol 2000;60:1085–94.
[25] Wu HF, Wu LL, Slagter WJ, et al. Use of rule of mixtures and metal volume fraction
for mechanical property predictions of fiber-reinforced aluminum laminates. J
Mater Sci 1994;29:4589–91.
[26] Lopez JA, Serna P, Navarro-Gregori J, et al. An inverse analysis method based on
deflection to curvature transformation to determine the tensile properties of
UHPFRC. Mater Struct 2014;48:3703–18.
[27] Racz ZS, Vas LM. Relationship between the flexural properties and specimen aspect
ratio in unidirectional composites. Compos Interfaces 2005;12:325–39.
[28] Mujika F. On the effect of shear and local deformation in three point bending tests.
Polym Test 2007;26:869–77.
[29] Abdullah MR, Prawoto Y, Cantwell WJ. Interfacial fracture of the fibre-metal la-
minates based on fibre reinforced thermoplastics. Mater Des 2015;66:446–52.