Composites Science and Technology 177 (2019) 57–65

Contents lists available at ScienceDirect

Composites Science and Technology

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

Direct observation of micro delamination in graphene oxide incorporated T

carbon fiber/epoxy composite via in-situ tensile test
Nitai Chandra Adaka,b, Suman Chhetria,b, Satvik Sabaradc, Himadri Roya,d,
Naresh Chandra Murmua,b, Pranab Samantaa,b,∗∗, Tapas Kuilaa,b,∗
a
Surface Engineering and Tribology Division, Council of Scientific and Industrial Research-Central Mechanical Engineering Research Institute, Durgapur, 713209, India
b
Academy of Scientific and Innovative Research (AcSIR), Ghaziabad, 201002, India
c
Department of Mechanical Engineering, National Institute of Technology, Raipur, 492010, India
d
NDT and Metallurgy Group, Council of Scientific and Industrial Research-Central Mechanical Engineering Research Institute, Durgapur, 713209, India

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

Keywords: The microdelamination analysis is a critical issue in current failure analysis of delamination phenomenon in
Carbon fibers carbon fiber (CF) reinforced epoxy composites. Both, the improvement of delamination resistance of the CF/
Matrix cracking epoxy composites using filler materials and the microdelamination studies of these hybrid composites through
Scanning electron microscopy (SEM) nondestructive technique are great concern to the material researchers. Herein, 0.2 wt% graphene oxide (GO)
Finite element analysis (FEA)
nanofiller was incorporated in CF/epoxy composite to resist the micro crack formation during delamination and
Failure criterion
in-situ tensile test was conducted to investigate the microdelamination behavior of the hybrid composite using a
microtester in the specimen chamber of a field emission scanning electron microscope. It was observed that the
GO modified epoxy matrix showed better matrix cracking resistance than the pure epoxy resin. The improve-
ment in matrix cracking resistance of the GO (0.2 wt%)/CF/epoxy composite is due to the good adhesion be-
tween GO modified matrix and CF. Besides the experimental study, Finite Element Analysis of the hybrid
composite was performed to predict the generated stress in each ply as it is not possible through experimental
study. Tsai-Wu failure criteria was also applied to analyze the damage in each ply of the composite laminate.

1. Introduction acid treatment, plasma treatment, gamma irradiation, rare earth

treatment and electrophoretic deposition etc to enhance the mechanical
In recent times, carbon fibers (CFs) reinforced polymer composites properties of the CF/polymer composites [5–9]. The polymer matrices
are primarily used in automotive, aerospace and marine industries due were modified with different inorganic and carbonaceous nanofillers
to their high strength and stiffness, corrosion and fatigue resistance and i.e. Al2O3, SiO2 and TiO2, graphite, carbon nanotubes, graphene etc
easy handling for production [1–3]. However, the CF/polymer com- [10–14]. In this regard, graphene and its derivatives are the most
posites are highly sensitive in thickness direction. The CF/polymer prominent nanofiller for surface treatment of CFs as well as for mod-
composites show matrix cracking, fiber breakage at micro level re- ification of the polymer matrixes. For example, Zhang et al. studied the
sulting delamination at component level due to their weak strength in mechanical properties of graphene oxide (GO) hybridized CF/epoxy
the thickness direction [4]. The failure of the composites at micro level composite and reported ∼12.7% improvement in interlaminar shear
is ascribed to the low adhesion between the binder and reinforcing strength (ILSS) compared to the base CF/epoxy composite [15]. Lee
materials. Hence, over the last few years, extensive works have been et al. reported ∼14% improvement in ILSS of the CF/epoxy composites
carried out on the enhancement of fiber-matrix adhesion. Two different after the surface treatment of CF by using partially reduced GO (RGO)
techniques have been followed to enhance the fiber-matrix adhesion i.e. via electrophoretic deposition process [16]. Adak et al. modified epoxy
surface treatment of the CFs and modification of the polymer matrix by matrix by incorporating of 0.2 wt% GO and showed ∼34% and 55%
using different types of nanofillers. The CF surfaces were modified via improvement in the tensile and flexural strength of the prepared hybrid

∗
Corresponding author. Surface Engineering and Tribology Division, Council of Scientific and Industrial Research-Central Mechanical Engineering Research
Institute, Durgapur, 713209, India.
∗∗
Corresponding author. Surface Engineering and Tribology Division, Council of Scientific and Industrial Research-Central Mechanical Engineering Research
Institute, Durgapur, 713209, India.
E-mail addresses: [email protected] (P. Samanta), [email protected] (T. Kuila).

https://doi.org/10.1016/j.compscitech.2019.04.006
Received 12 July 2018; Received in revised form 13 February 2019; Accepted 6 April 2019
Available online 17 April 2019
0266-3538/ © 2019 Elsevier Ltd. All rights reserved.
N.C. Adak, et al. Composites Science and Technology 177 (2019) 57–65

CF/epoxy composite [17]. The authors also reported ∼33 and ∼62% Hummers' method as per our previous report [28]. The spectroscopic
enhancement in plane strain fracture toughness of the chemically re- and microscopic studies of the synthesized GO were described in our
duced GO (rGO) and thermally reduced GO (TRGO) loaded CF/epoxy previous report [29]. For GO/epoxy composite preparation, the desired
composites respectively [18,19]. Menbari et al. recommended that amount of GO was dispersed in THF via water bath sonication for 2 h in
functionalized GO could be more appropriate filler for increasing the a bath sonicator. The GO dispersion (0.2 wt% with respect to epoxy and
mechanical strength of the fiber reinforced polymer composites as it hardener) was then mixed with epoxy resin and sonicated further for
holds different reactive groups, which assist to disperse the graphene 1 h. The excess THF was removed from the GO/epoxy mixture via
nano-sheets in the matrix material and facilitates interaction with the evaporation at 85 °C under reduced pressure for an hour. The appro-
polymer matrix [20]. In addition to the experimental study, a number priate amount of hardener was then added into the GO-epoxy mixture
of multiscale modeling and numerical analysis of the graphene-based and stirred at 500 rpm for 2 min for homogenization. Finally, the GO-
CF/epoxy composites are available in literature. Aluko et al. built a incorporated epoxy matrix was obtained to prepare the hybrid com-
multiscale model of GNPs/CF/epoxy composite and studied the me- posite after degassing the mixture inside a vacuum chamber for 5 min at
chanical properties of the composite through the molecular dynamic room temperature. The composites were prepared using the vacuum
(MD) simulation. The simulated results report that the modulus of the assisted resin transfer molding process. At first, the fabrics were cut into
composite increases with increasing the GNP loading in the epoxy resin dimension of 15 cm × 20 cm and were placed on a molding tool that
[21]. was previously coated with a release agent for easy removal of the
As stated previously, most of the studies focused on the develop- prepared composites. The stacking sequence of the lamina was (0/90/
ment of mechanically stable CF/epoxy composites by incorporating the 0/90) for the tensile test. A peel ply was placed on the top of the lamina
graphene-based nanofiller. To the best of our knowledge, there is no for the separation of the bagging film from the composites. A nylon
report on the in-situ micro test of the developed composites. In-situ mesh was also placed on the top of the peel ply to ensure the continuous
micro test of the developed composites is very crucible to understand resin transfer in the lateral direction and the inlet of the resin supply
the damage mechanisms of the hybrid composites at micron level. was fixed with a spiral pipe around the fabrics. The vacuum pressure
According to Tong et al., transverse cracks initiate in the composites at was 0.2 Pa and the resin was cured with the shape of the performance
first and then propagate until the cracks expand to ply boundaries, at fabrics. The prepared composite panels were post-cured at 70 °C for 2 h
which point they create micro-delaminations [22]. Once the micro- and 120 °C for 1 h after curing at room temperature for 24 h. Finally,
delamination ceases after saturation of the crack density, then dela- the desired samples were cut from the panels to conduct mechanical
mination becomes the leading failure mechanism [23,24]. A few in-situ tests.
micromechanical studies are available in literature, which are focused
on the types of fibers, forms of fiber, stacking sequences of the fiber, 2.2. Setup for in-situ tensile test in FE-SEM
types of loads etc [25–27]. In this regard, effect of graphene on the
micromechanical properties of the CF/epoxy composites has not been A schematic of the experimental arrangement is shown in Fig. 1,
studied previously by using in-situ micro test. Therefore, it is an obvious including the technical details of the specimens. The micro tests under
gap to analyze the micromechanical properties of graphene hybridized tensile load were performed inside FE-SEM (Ʃigma HD, Carl Zeiss,
CF/epoxy composites experimentally as well as numerically at macro Germany) to examine failure mechanism of the prepared composites.
level. The FE-SEM as shown in Fig. 1a equipped with loading facility to
Herein, it is intended to study the effect of GO on the growth of generate the uni-axial tensile load. MTEST2000 module tensile stage
micro delamination of CF/epoxy composites. The main objective for GO (Deben™) integrated with 2 kN load cell was placed in the sample
incorporation in the epoxy composite is to enhance the mechanical chamber of the FESEM as shown in Fig. 1b. The samples were mounted
strength of the matrix material and consequently decreasing the micro horizontally, clamped to a pair of jaws and supported on stainless steel
delamination of the prepared composite. The in-situ tensile test of the slide bearings. A dual threaded lead screw drives the jaws symme-
developed composite samples was carried out with a micro tester, trically in opposite directions, keeping the sample centered in the field
which was placed inside the sample chamber of a Field Emission of view. The standard speed range of the jaws was 0.033–0.4 mm/min
Scanning Electron microscope (FE-SEM). Real-time video footage from and was controlled by a microtest controller. An optical encoder fitted
the FE-SEM was recorded during testing to identify various failure motor controls the speed with an accuracy level of ∼1%. The loading
events at the microscale. Besides the experimental studies, Finite direction of the samples in the microtester along with the controlling
Element Analysis (FEA) was carried out to predict the stress generated unit is shown in Fig. 1c. The speed of the displacement of the samples
at each ply and deformation of the composite laminate. FEA was also during the tensile testing was 0.2 mm/min. The overall length of sample
used to predict the damage in each ply for both the composite speci- was 41 ± 1 mm and gauge length 15 ± 0.5 mm with thickness of
mens. 1.2 mm and it has been shown in Fig. 1d. The width of the gauge length
was 3 mm and a “V” notch was made in the middle of the gauge length.
2. Experimental The “V” notch was created by inserting a spanking new razor blade in
the test specimen. The resulting tensile force was measured from the
2.1. Materials and specimen preparation load cell.

A plain weave bi-directional CF cloth with a thickness of 0.3 mm 2.3. Finite element modeling of the prepared composite sample
and surface density of 200 g/m2 was purchased from Flips India
Engineering (Mumbai, India). The liquid modified Bisphenol-A epoxy The finite element analysis was carried out for CF/epoxy tensile
resin (LAPOX*C-51) and cycloaliphatic amine hardener (Lapox AH- specimen using commercially available software ANSYS™ (Version
428) were procured from Atul Limited (Gujarat, India). Silicon spray 16.2), which solves complex engineering problems ranging from linear
mold release agent and adhesive spray were procured from Aerol to non-linear behavior. The properties given to the material are ac-
Formulation Pvt. Ltd. (New Delhi, India). Graphite flakes with 99% cording to the experimental tensile test performed and the details are
carbon content and −325 mesh particle size (≥99%) was bought from provided in the supplementary data (S1). ANSYS™ (ANSYS Composite
Indiamart. Potassium permanganate (KMnO4), sulphuric acid (H2SO4), Prepost (ACP)) was used for modeling the stack up of laminate. The
hydrochloric acid (HCl), hydrogen peroxide (H2O2) and tetra- laminate was made up of CF plies, each 0.3 mm thick, with (0/90/0/90)
hydrofuran (THF) were purchased from Merck, Mumbai, India. stack up sequence. The model was imported into the transient structure
GO was prepared from natural graphite flakes according to modified where the stress generated at each sub step can be determined and the

58
N.C. Adak, et al. Composites Science and Technology 177 (2019) 57–65

Fig. 1. Schematics of the (a) FE-SEM (b) position of microtester with enlarged view in the specimen chamber of FE-SEM (c) loading direction of the specimen with
controller unit of the microtester and (d) specimen dimension.

number of sub steps can be changed accordingly to get the value of Fig. 3c. At this position, the generated stress in the CF/epoxy composite
stress instantaneously. The boundary conditions were applied ac- restricted and micro-delamination initiated at 90° ply. Further in-
cording to the actual forces acting on the tensile test specimen in the creasing of load (∼681 N) showed initiation of numerous transverse
microtester. micro cracks across the span length of the specimen, which is shown in
Fig. 3d. When the transverse cracks joined by a continuous micro-
3. Results and discussion delamination, a catastrophic failure of the specimen occurred leading to
the sudden drop in load carrying capacity of the specimen. The gen-
3.1. Experimental results and discussion erated transverse cracks in the 90° plies at the outer tensile surface grew
inwards as shown in Fig. 3e. The ultimate load carrying capacity of the
The damage progression of pure CF/epoxy (0/90/0/90) specimen hybrid composite was found to be ∼702 N with elongation at break of
under tensile loading has been illustrated in Fig. 3. A video (V1) of the ∼0.30 mm. The corresponding applied load, elongation at break, stress
micro delamination during in-situ tensile testing of the CF/epoxy and strain at the marked region for both the composite samples are
composite specimen has been included in the supplementary informa- summarized in Table 1. It may be noted that the in-situ notched tensile
tion. A close-up view of the indicated representative elements were tests were carried out to compare the two materials resistance to crack
captured from the video (V1) during the in-situ tensile test and are propagation and to understand the mechanism of failure under applied
shown in Fig. 2a-e. Each of the representative elements exhibits the load in a notched tensile specimens of identical geometry. The speci-
failure events as seen during testing. As the load was gradually in- mens being of non-standard geometry are primarily under plane stress
creased, a micro crack was generated at the tip of the notch at around condition. Owing to the presence of pre-existing notch the remaining
the applied load of 130 N, which is depicted in Fig. 2a. The microcrack ligament does not get sufficient material to elongate. Additionally, the
propagated slowly on further extension of the sample up to the applied total elongation in both the samples is very less (0.3 mm). Possibly
load of 350 N as shown in Fig. 2b. Interestingly, it was observed that because of all these factors (notched specimen, plane stress and low
beyond the load of 350 N, the crack proliferation speed increased and ductility of this class of material), the total elongation in the notched
around the load of 464 N the matrix ruptured suddenly as shown in specimens are found to be almost of the same order. The obtained test
Fig. 2c. The matrix cracking was followed by the initiation of fiber results highlight that the load carrying capacity ahead of a notch tip of
cracking. The state of fiber cracking at 570 N, is shown in Fig. 2d. The the GO/CF/epoxy composite is higher compared to the CF/epoxy
maximum load carrying capacity of the specimen was 586 N with composite due to better adhesion between CF and epoxy matrix in
elongation of ∼0.32 mm. The load carrying capacity of the composite presence of GO. Therefore, the total elongation at break for both the
dropped suddenly on further extension of the specimen and at this composite samples is very less and comparable.
condition the CFs fully cracked as shown in Fig. 2e. The incorporation Supplementary video related to this article can be found at https://
of 0.2 wt% of GO induced the change in the nature of load-displacement doi.org/10.1016/j.compscitech.2019.04.006.
diagram of CF/epoxy composite specimen, which is shown in Fig. 3 The post fracture analysis of the composite samples is portrayed in
supported by the snapshots captured from the video (V2). The micro- Fig. 4. The representative pictures have been depicted to demonstrate
cracks were found to generate in the matrix at an applied load of 201 N the existence of GO sheets around the CF surface and pull out failure of
and beyond this load, matrix de-bonding started in the composite. The the fiber across the fracture region. It is clearly observed that, the
microcracks proliferated making features akin to a river branch as fracture surface of the pure CF/epoxy composite sample was quite
shown in micrograph 3a. The further increment of load collapsed the smoother than the 0.2 wt% GO modified CF/epoxy composite specimen
generated microcracks and started to form a large transverse crack at as shown in Fig. 4(a&b). The smooth surface of the CF in the fractured
∼462 N load as observed in Fig. 3b. The transverse microcracks rapidly CF/epoxy sample signifies the low adhesion between CF and epoxy
grew up to the next alternately oriented lamina boundary as shown in resin. The fiber and matrix cracking are more in case of base CF/epoxy

59
N.C. Adak, et al. Composites Science and Technology 177 (2019) 57–65

Fig. 2. Results from in situ testing of the CF/epoxy composite sample. The deformations (a) initiation of microcrack (b) crack propagation (c) matrix rupture (d) fiber
pullout and (d) fiber cracking of the specimen with corollary load are shown on the load vs. elongation diagram.

composite sample as a low adhesion acts between CF and epoxy resin. composites as compared to the CF/epoxy composites [30,36–38]. The
The presence of GO in the epoxy resin enhances the adhesion between toughened CF/epoxy/GO composite obstructs the micro-crack pro-
CF and epoxy resin through chemical and hydrogen bonding interac- liferation along the axis of fibers and reduces the microdelamination
tion. Presence of various oxygen functionalities (–COOH, CeOeC, phenomenon. As a result, the presence of GO impedes the crack pro-
eOH, etc.) of GO facilitates to form the bridging/interlocking structure pagation during fracture leading to the formation of new fracture sur-
between epoxy matrix and CF [30–32]. The bridging/interlocking be- faces and the CF/epoxy/GO composite show rougher texture (Fig. 4)
tween GO and epoxy matrix enhances the cross linking density of the CF than the CF/epoxy composite. Similar observation on fracture surface
composites and make it tougher than the unmodified (without GO morphology of the graphene filled epoxy composite has been reported
filler) CF composites resulting good adhesion between the matrix and earlier confirming our contention on the rougher fracture surface of the
reinforcing materials [33–35]. Therefore, additional strain energy is GO filled composite materials [39–41].
required to rupture the bridging/interlocking of the CF/epoxy/GO

Fig. 3. Results from in situ testing of the

GO/CF/epoxy composite sample. The de-
formations (a) initiation of microcrack with
branch cracks (b) collapse of microcracks
and initiation of transverse crack (c) trans-
verse crack propagation up to the next la-
mina boundary (d) transverse cracks along
the span length and (e) fiber cracking of the
specimen with corollary load are shown on
the load vs. elongation diagram.

60
N.C. Adak, et al. Composites Science and Technology 177 (2019) 57–65

Table 1 the specimens and the contour displayed in red color has the maximum
The corresponding load, elongation, stress and strain (%) of the representative deformation. The deformation of the laminate decreased as the point is
elements of both the composite samples. probed near to the v-notch. The non-uniform deformation at one end is
Force (N) Elongation (mm) Stress (MPa) Strain (%) due to the presence of v-notch in the composite sample. The contour
displayed in red color deformed more as compared to the opposite side
CF/epoxy (a) 130 0.05 86 0.33 of the v-notch displayed in yellowish-orange color. The load vs. dis-
(b) 305 0.13 203 0.86
placement curve for CF/epoxy and GO/CF/epoxy composites is plotted
(c) 464 0.21 309 1.40
(d) 570 0.29 380 1.93 in Fig. 7. The curves depict the comparison between experimental and
(e) 490 0.31 326 2.06 simulation results. The elongation at break in both the cases is almost
GO/CF/epoxy (a) 201 0.05 134 0.33 same and is equal to 0.3 mm. However, the load for GO/CF/epoxy
(b) 462 0.14 308 0.93
composite is ∼27% higher as compared to the CF/epoxy composite.
(c) 616 0.23 410 1.53
(d) 681 0.28 454 1.86
The simulation results for both the specimens are conforming well to
(e) 604 0.30 402 2.00 the experimental results but the variation in GO/CF/epoxy is more as
compared to CF/Epoxy.

3.2. FEA results and discussion

3.2.2. Damage in each of the plies
3.2.1. Stress and deformation in each of the plies and laminate In this section, the Progressive Damage Analysis (PDA) associated
The in-situ tensile testing could not able to determine a few para- with the damage phenomenon that occurs in the above described la-
meters developed in each of the plies. Simulating the test specimen with minate is discussed. Damage modeling can be used to determine the
appropriate boundary conditions and extracting the results according to initiation of damage in the composite material and prediction of the
the need can mitigate this disadvantage. The same was performed in post-damage response of the structure. It can also predict the damage in
this study with the test specimen to determine stress developed in each the laminate which can't be determined by the stress values generated
of the plies. ANSYS™ (Version 16.2) provides an opportunity to de- in ANSYS after computation. In ANSYS, the stress value goes on in-
termine and pictorially understand the stress developed and stress creasing as the load keeps on increasing, giving us no information about
concentration in each of the plies explicitly. Fig. 5(a1–d1) depicts the the moment at which the laminate damages. This section focuses on the
stress generated in Ply 1, Ply 2, Ply 3 and Ply 4 of CF/epoxy respec- prediction of damage in each of the plies, determination of damage due
tively. Similarly, the stress generated in each of the plies in GO/CF/ to matrix or fiber and the validation of simulated results with the ex-
epoxy is shown in Fig. 5(a2-d2). It can be clearly inferred from the perimental results. To perform PDA of composite materials, the user
figures that the plies having the fiber orientation parallel to the appli- needs to provide linear elastic orthotropic material properties and two
cation of load, experience more stress than the plies which are aligned delamination models i.e. damage initiation criteria and damage evo-
perpendicular to the direction of application of load. The stress gener- lution law. Damage initiation criteria, which includes failure mode
ated in Ply 1 and 3 for both the composite specimens is significantly based on composite failure theories, defines the criteria type for de-
higher than the stresses generated in the ply 2 and 4. When the lamina termining the initiation of material damage under loading. The damage
is loaded along the direction of the fiber orientation, the developed evolution law defines the way a material degrades following the in-
stress is more than that of the fiber orientation perpendicular to the itiation of damage. There are several failure modes viz. maximum
applied load. Moreover, the stress experienced by CF/epoxy is less as strain, maximum stress, Tsai-Hill, Tsai-Wu, Puck, Hashin, LaRc03,
compared to GO/CF/epoxy composite. Further, it is clearly visible that LaRc04 etc. which follow the damage initiation criteria. Tsai-Wu failure
the maximum stress on all the plies is concentrated near the v-notch of criterion is appropriate for analyze the (0/90)2 laminate among the
the laminate. Unlike the stress generated in each of the plies which can failure theories [42]. The Tsai-Wu failure criterion is a phenomen-
be depicted explicitly, the deformation is determined only for the la- ological material failure theory which is widely used for anisotropic
minate and not for individual plies. It may be due to the fact that the composite materials. In this case, predictions of the Tsai-Wu interactive
deformation cannot be accounted differently for each of the plies as the theory were in good agreement with the experimental results than other
deformation is same for each ply as well as for the laminate. So, the theories. The Tsai-Wu criterion predicts failure when the failure index
deformation of ply is equal to the deformation encountered in the la- in a laminate reaches to 1. This failure criterion is a specialization of the
minate. Fig. 6 a1-a2 shows the deformation after the tensile test in both general quadratic failure criterion proposed by Gol'denblat and Kopnov
[43] and can be expressed in the form fi σi + fij σi σj ≤ 1 whereas i, j= 1

Fig. 4. Fracture surfaces of (a) unmodified and modified by (b) 0.2 wt% GO of the CF/epoxy composites.

61
N.C. Adak, et al. Composites Science and Technology 177 (2019) 57–65

Fig. 5. (a1), (b1), (c1) and (d1) stress generated in ply1, ply 2, ply 3 and ply 4 of CF/epoxy respectively, (a2), (b2), (c2) and (d2) stress generated in ply1, ply 2, ply 3
and ply 4 of GO/CF/epoxy respectively.

… 6 and repeated indices indicate summation, and fi , fij are experi-

mentally determined material strength parameters. If the failure surface
is to be closed and convex, the interaction terms fij must satisfy
fii f jj − fij2 ≥ 0 .The equation for Tsai-Wu criterion can be written as.
f1 σ1 + f2 σ2 + f3 σ3 + f11 σ12 + f22 σ22 + f33 σ32 + f44 σ42 + f55 σ52 + f66 σ62 .
+ 2f12 σ1 σ2 + 2f23 σ2 σ3 + 2f13 σ1 σ3 ≤ 1
Since the laminate is being subjected to the tensile test, the Tsai-Wu
criterion for plane stress can be reduced into,
f1 σ1 + f3 σ3 + f11 σ12 + f33 σ32 + f55 σ52 + 2f13 σ1 σ3 ≤ 1
1 1 1 1 1
whereas, f1 = F1t
− ,f
F1c 3
= F3t
− ,f
F3c 11
= F1t F1c f33 =
1 1 1
,f
F3t F3c 55
= ,f
F52 13
≅ 2
f11 f33 .
Fij is the ultimate strength where ‘i’ and ‘j’ represents the direction
and the type of loading respectively. The appropriate values were put in
the ANSYS™ workbench and the damage in each of the plies in each
specimen is depicted in Fig. 8(a1-d1) and Fig. 8(a2-d2). According to the
Fig. 7. Variation of elongation with load and comparison of Experimental and
ACP-Post processor, the damage has a range between ‘zero’ and ‘two’.
simulation results.
‘Zero’ represents no damage in the material whereas ‘two’ represents
the complete damage and the values between ‘zero’ and ‘two’ signify
the partial damage. It is evident from Fig. 8 that the most of the damage due to loading. The damage in the fiber and matrix can be determined
is encountered by the Ply 2 and 4 which has the fibers oriented in explicitly for each of the plies using ANSYS™. Fig. 9(a1&a3) represents
perpendicular direction to the application of load. The area depicted in the fiber damage in Ply 1 and 2 for CF/epoxy and Fig. 9(a2&a4) re-
light green color has encountered most of the damage and the blue presents the damage in matrix for the same plies and specimens. Si-
color signifies no damage. milarly, Fig. 9(b1&b3) depict the fiber damage in GO/CF/epoxy and
Fig. 9(b2&b4) shows the matrix damage in GO/CF/epoxy. As discussed
in the earlier section, the damage can be quantified between ‘zero’ and
3.2.3. Damage due to fiber/matrix in each of the plies
‘two’. It can be inferred from Fig. 9(a1,a3,b1&b3) that the damage in
Fiber-reinforced composite materials are known to have various
fiber is zero in each of the plies but the matrix in Ply 2 of both the
failure mechanisms, including fiber and matrix failure that can occur

Fig. 6. (a1) Elongation in CF/epoxy (a2) Elongation

in GO/CF/epoxy.

62
N.C. Adak, et al. Composites Science and Technology 177 (2019) 57–65

Fig. 8. (a1) to (d1) damage in ply1, ply 2, ply 3 and ply 4 of CF/epoxy respectively and (a2) to (d2) damage in ply1, ply 2, ply 3 and ply 4 of GO/CF/epoxy
respectively.

Fig. 9. (a1) and (a3) fiber damage ply 1 and ply 2 respectively in CF/epoxy, (a2) and (a4) matrix damage in ply 1 and 2 respectively in CF/epoxy. (b1) and (b3) fiber
damage ply 1 and ply 2 respectively in GO/CF/epoxy, (b2) and (b4) matrix damage in ply 1 and 2 respectively in GO/CF/epoxy.

specimens (Fig. 9 a4 and b4) experience the maximum damage. The composites are mostly contributed by ply 3 as compared to the other
behavior of ply 3 and 4 is same as ply 1 and 2 because of the same three plies. Therefore, ply 3 is the most critical ply in the CF/epoxy and
orientation. The minimum matrix damage in ply 1 and ply 3 in both the GO/CF/epoxy composite specimens.
specimens proves that the maximum amount of load is carried by the
fibers, not by the matrix due to which the stress generated in fibers of
ply 1 and ply 3 is more. The critical ply failure can be predicted by 4. Conclusions
using ANSYS™ ACP (Post). It can also predict the maximum damage in
the ply within the element considered. The color displayed quantifies The hybridization effects of GO on the interfacial adhesion between
the criticalness and the numeric value in parenthesis displays the ply CF and epoxy resin for restricting the microdelamination phenomenon
which is critical. Fig. 10 shows that the damage elements in the in the prepared composite was studied experimentally via in-situ tensile
test inside a FE-SEM. The GO filled CF/epoxy composite showed the

63
N.C. Adak, et al. Composites Science and Technology 177 (2019) 57–65

Fig. 10. (a1) Critical Ply failure in CF/epoxy (a2) Critical ply failure in GO/CF/epoxy.

matrix and fiber cracking at higher load as compared to the base CF/ remote nitrogen oxygen plasma treatment on carbon fabric and its composites with
epoxy composite. The formation of branch cracking in case of hy- specialty polymers, J. Mater. Sci. 4 (2011) 964–974.
[7] J. Li, Y. Huang, S. Fu, L. Yang, H. Qu, G. Wu, Study on the surface performance of
bridized composite sample also signifies that the required energy for CFs irradiated by Gamma-ray under different irradiation dose, Appl. Surf. Sci. 256
crack propagation is more than the CF/epoxy composite. The FEA was (2010) 2000–2004.
also performed to evaluate the stress generation in each plies and [8] Z. Xu, Y. Huang, C. Zhang, G. Chen, Influence of rare earth treatment on interfacial
properties of CFs/epoxy composites, Mater. Sci. Eng. A 444 (2007) 170–177.
comparison was carried out between experimental and simulation re- [9] A. Qi, R.N. Andrew, T. Erik Thostenson, Electrophoretic deposition of carbon na-
sults. The simulated results showed that the stress generation in ply 1 notubes onto carbon-fiber fabric for production of carbon/epoxy composites with
and ply 3 is comparatively higher than the ply 2 and ply 4 as the ap- improved mechanical properties, Carbon 50 (2012) 4130–4143.
[10] F. Kachold, R. Singer, Mechanical properties of carbon fiber-reinforced aluminum
plication of load was parallel to the direction orientation of plies 1 and manufactured by high-pressure die casting, J. Mater. Eng. Perform. 25 (2016)
ply 3. On the contrary, the progressive damage experienced in ply 2 and 3128–3133.
ply 4 is higher than the ply 1 and ply 3. The experimental load vs. [11] F. Cai, F. Gao, S. Pant, X. Huang, Q. Yang, Solid particle erosion behaviors of
carbon-fiber epoxy composite and pure titanium, J. Mater. Eng. Perform. 25 (2016)
elongation study of the prepared laminates conformed well to the si-
290–296.
mulated results and the value of total elongation was 5–8% lower than [12] R.M. Rocha, C.A.A. Cairo, M.L.A. Graca, Formation of carbon fiber-reinforced
the simulated values. ceramic matrix composites with polysiloxane/silicon derived matrix, Mater. Sci.
Eng. 437 (2006) 268–273.
[13] Y. Arao, S. Yumitori, H. Suzuki, T. Tanaka, K. Tanaka, T. Katayama, Mechanical
Acknowledgement properties of injection-molded carbon fiber/polypropylene composites hybridized
with nanofillers, Compos. Part A. 55 (2013) 19–26.
The authors are grateful to the Director of CSIR-CMERI and the [14] C.Y. Lee, J.H. Bae, T.Y. Kim, S.H. Chang, S.Y. Kim, Using silane-functionalized
graphene oxides for enhancing the interfacial bonding strength of carbon/epoxy
technical assistants Mr. Subhasis Biswas and Abhijit Mondal for as- composites, Compos. Part A. 75 (2015) 11–17.
sisting in situ tensile test. The authors are also thankful to the Council of [15] X. Zhang, X. Fan, C. Yan, H. Li, Y. Zhu, X. Li, L. Yu, Interfacial microstructure and
Scientific and Industrial Research, New Delhi, India, for funding the properties of carbon fiber composites modified with graphene oxide, ACS Appl.
Mater. Interfaces 4 (2012) 1543–1552.
Fast Track Translational Project (MLP218112). [16] W. Lee, J.U. Lee, H.J. Cha, J.H. Byun, Partially reduced graphene oxide as a multi-
functional sizing agent for carbon fiber composites by electrophoretic deposition,
Appendix A. Supplementary data RSC Adv. 3 (2013) 25609–25613.
[17] N.C. Adak, S. Chhetri, N.H. Kim, N.C. Murmu, P. Samanta, T. Kuila, Static and
dynamic mechanical properties of graphene oxide-incorporated woven carbon
Supplementary data to this article can be found online at https:// fiber/epoxy composite, J. Mater. Eng. Perform. 27 (2018) 1138–1147.
doi.org/10.1016/j.compscitech.2019.04.006. [18] N.C. Adak, S. Chhetri, T. Kuila, N.C. Murmu, P. Samanta, J.H. Lee, Effects of hy-
drazine reduced graphene oxide on the inter-laminar fracture toughness of woven
carbon fiber/epoxy composite, Compos. Part B Eng. 149 (2018) 22–30.
References [19] N.C. Adak, S. Chhetri, N.C. Murmu, P. Samanta, T. Kuila, Effect of thermally re-
duced graphene oxide on mechanical properties of woven carbon fiber/epoxy
[1] A. Todoroki, H. Kobayashi, J.G. Lee, Image analysis of delamination cracks in composite, Crystals 8 (2018) 1–14.
carbon-fibre composites by scanning acoustic microscopy, Compos. Sci. Technol. 52 [20] S. Menbari, A. Ashori, H. Rahmani, R. Bahrami, Viscoelastic response and inter-
(1994) 551–559. laminar delamination resistance of epoxy/glass fiber/functionalized graphene oxide
[2] S.H. Yoo, S. Park, Y. Park, D. Lee, H. Joh, I. Shin, S. Lee, Facile method to fabricate multi-scale composites, Polym. Test. 54 (2016) 186–195.
carbon fibers from textile-grade polyacrylonitrile fibers based on electronbeam ir- [21] O. Aluko, S. Gowtham, G.M. Odegard, Multiscale modeling and analysis of gra-
radiation and its effect on the subsequent thermal stabilization process, Carbon 118 phene nanoplatelet/carbon fiber/epoxy hybrid composite, Compos. Part B Eng. 131
(2017) 106–113. (2017) 82–90.
[3] Y.X. Zhou, Y. Wang, Y.M. Xia, S. Jeelani, Tensile behavior of carbon fiber bundles at [22] J. Tong, F.J. Guild, S.L. Ogin, P.A. Smith, On matrix crack growth in quasi-isotropic
different strain rates, Mater. Lett. 64 (2010) 246–248. laminates—I experimental investigation, Compos. Sci. Technol. 57 (1997)
[4] H. Liu, B.G. Falzon, W. Tan, Experimental and numerical studies on the impact 1527–1535.
response of damage-tolerant hybrid unidirectional/woven carbon-fibre reinforced [23] J. Rebière, D. Gamby, Strain energy release rate analyse of matrix micro cracking in
composite laminates, Compos. Part B Eng. 136 (2018) 101–118. composite cross-ply laminates, Mater. Sci. Appl. 2 (2011) 537–545.
[5] F. Su, Z. Zhang, K. Wang, W. Jiang, W. Liu, Tribological and mechanical properties [24] P. Ladèveze, G. Lubineau, An enhanced mesomodel for laminates based on micro-
of the composites made of carbon fabrics modified with various methods, Compos. mechanics, Compos. Sci. Technol. 62 (2002) 533–541.
part A 36 (2005) 1601–1607. [25] D.J. Mortell, D.A. Tanner, C.T. McCarthy, In-situ SEM study of transverse cracking
[6] S. Tiwari, M. Sharma, S. Panier, B. Mutel, P. Mitschang, J. Bijwe, Influence of cold and delamination in laminated composite materials, Compos. Sci. Technol. 105

64
N.C. Adak, et al. Composites Science and Technology 177 (2019) 57–65

(2014) 118–126. graphene nanocomposites for high k dielectric and ultra violet radiation resistant
[26] A. Todoroki, H. Kobayashi, Image analysis of delamination cracks in carbon-fibre application, RSC Adv. 4 (58) (2014) 1–11.
composites by scanning acoustic microscopy, Compos. Sci. Technol. 52 (1994) [35] C.H. Hsu, M.H. Hsu, K.C. Chang, M.C. Lai, P.J. Liu, T.L. Chuang, J.M. Yeh, W.R. Liu,
551–559. Physical study of room-temperature-cured epoxy/thermally reduced graphene
[27] S. Liu, Delamination and matrix cracking of cross-ply laminates due to a spherical oxides with various contents of oxygen-containing groups, Polym. Int. 63 (2014)
indenter, Compos. Struct. 25 (1993) 257–265. 1765–1770.
[28] T. Kuila, S. Bose, P. Khanra, A.K. Mishra, N.H. Kim, J.H. Lee, A green approach for [36] F. Li, Y. Liu, C.B. Qu, H.-M. Xiao, Y. Hua, G.-X. Sui, S.-Y. Fu, Enhanced mechanical
the reduction of graphene oxide by wild carrot root, Carbon 50 (2012) 914–921. properties of short carbon fiber reinforced polyethersulfone composites by gra-
[29] N.C. Adak, S. Chhetri, N.C. Murmu, P. Samanta, T. Kuila, J.H. Lee, Experimental phene oxide coating, Polymer 59 (2015) 155–165.
and numerical investigation on the mechanical characteristics of polyethylenimine [37] H. Mahmood, M. Tripathi, N. Pugno, A. Pegoretti, Enhancement of interfacial ad-
functionalized graphene oxide incorporated woven carbon fibre/epoxy composites, hesion in glass fiber/epoxy composites by electrophoretic deposition of graphene
Compos. Part B Eng. 156 (2019) 240–251. oxide on glass fibers, Compos. Sci. Technol. 126 (2016) 149–157.
[30] A.K. Pathak, M. Borah, A. Gupta, T. Yokozeki, S.R. Dhakate, Improved mechanical [38] B. Park, W. Lee, E. Lee, S.H. Min, B.-S. Kim, Highly tunable interfacial adhesion of
properties of carbon fiber/graphene oxide epoxy hybrid composites, Compos. Sci. glass fiber by hybrid multilayers of graphene oxide and aramid nanofiber, ACS
Technol. 135 (2016) 28–38. Appl. Mater. Interfaces 7 (5) (2015) 3329–3334.
[31] H. Yang, C. Shan, F. Li, Q. Zhang, D. Han, L. Niu, Convenient preparation of tunably [39] D.R. Bortz, E.G. Heras, I.M. Gullon, Impressive fatigue life and fracture toughness
loaded chemically converted graphene oxide/epoxy resin nanocomposites from improvements in graphene oxide/epoxy composites, Macromolecules 45 (2012)
graphene oxide sheets through two-phase extraction, J. Mater. Chem. 19 (46) 238–245.
(2009) 8856–8860. [40] L.C. Tang, Y.J. Wan, D. Yan, Y.B. Pei, L. Zhao, Y.B. Li, L.B. Wu, J.X. Jiang, G.Q. Lai,
[32] T. Ramanathan, A.A. Abdala, S. Stankovich, D.A. Dikin, M. Herrera-Alonso, The effect of graphene dispersion on the mechanical properties of graphene/epoxy
R.D. Piner, D.H. Adamson, H.C. Schniepp, X. Chen, R.S. Ruoff, S.T. Nguyen, composites, Carbon 60 (2013) 16–27.
I.A. Aksay, R.K. Prud'Homme, L.C. Brinson, Functionalized graphene sheets for [41] X. Wang, J. Jin, M. Song, An investigation of the mechanism of graphene tough-
polymer nanocomposites, Nat. Nanotechnol. 3 (2008) 327–331. ening epoxy, Carbon 65 (2013) 324–333.
[33] S. Lu, S. Li, J. Yu, Z. Yuan, B. Qi, Epoxy nanocomposites filled with thermotropic [42] I.M. Daniel, Failure of composite materials, Strain 43 (2007) 4–12.
liquid crystalline epoxy grafted graphene oxide, RSC Adv. 3 (2013) 8915–8923. [43] I. Gol'denblat, V.A. Kopnov, Strength of glass reinforced plastic in the complex
[34] K. Sethuramana, P. Prabunathana, M. Alagar, Flexible capron toughened epoxy/ stress state, Polym. Mech. 1 (1966) 54–60.

65