CC BY
87
УДК 669.018
Влияние графеноксидных нанопластин на отклик слоистого алюмостеклопластика при высокоскоростном ударном нагружении
1 112 A. Heydari , H. Khoramishad , H. Alikhani , F. Berto
1 Научно-технологический университет Ирана, Тегеран, 16846, Иран 2 Норвежский университет естественных и технических наук, Тронхейм, 7491, Норвегия
В статье исследовано влияние графеноксидных нанопластин на отклик слоистых алюмостеклопластиков класса СИАЛ при высокоскоростном ударном нагружении. Для упрочнения СИАЛов применяли нанонаполните-ли с различной массовой долей: 0.25, 0.50 и 1.00 % графеноксидных нанопластин; 0.25 % гибридных графеноксидных нанопластин; 0.25 % многостенных углеродных нанотрубок. Показано, что при добавлении 0.5 % графеноксидных нанопластин остаточная скорость ударника снижается на 33 %, а энергопоглощающая способность СИАЛа возрастает на 22.9 %. Исследование методами электронной и оптической микроскопии выявило увеличение адгезии между волокнами наполнителя и матрицей, а также улучшение когезии в матрице упрочненного СИАЛа. В результате значительно уменьшилась область вторичного повреждения в композитном слое. С другой стороны, графеноксидные нанопластины оказывают негативное воздействие на подавление расслоения композитного слоя, а также на прочность на границе раздела между композитным и алюминиевым слоями. Это приводит к интенсивной пластической деформации алюминия, которая является важным механизмом поглощения энергии в СИАЛах. Результаты испытаний на растяжение композитных слоев неупрочненных и упрочненных СИАЛов показали, что введение 0.5 % графеноксидных нанопластин в матрицу повышает жесткость на растяжение, прочность и ударную вязкость на 33.4, 45.0 и 25.6 % соответственно.
Ключевые слова: слоистый металлопластик, слоистый алюмостеклопластик класса СИАЛ, графеноксидные нанопластины, высокоскоростное воздействие, механизмы разрушения
DOI 10.24411/1683-805X-2020-13004
The effect of graphene-oxide nanoplatelets on the high-velocity impact response of glass laminate aluminum reinforced epoxy
A. Heydari1, H. Khoramishad1, H. Alikhani1, and F. Berto2
1 School of Mechanical Engineering, Iran University of Science and Technology, Narmak, Tehran, 16846, Iran 2 Department of Mechanical and Industrial Engineering, Norwegian University of Science and Technology,
Trondheim, 7491, Norway
In this paper, the influence of graphene-oxide nanoplatelets (GONPs) on the response of the glass laminate aluminum reinforced epoxy (GLARE) under high-velocity impact loading was investigated. Different weight percentages of nano-fillers including 0.25, 0.50 and 1.00 wt % GONPs and a hybrid of 0.25 wt % GONPs and 0.25 wt % multiwalled carbon nanotubes (MWCNTs) were used for reinforcing GLAREs. Addition of 0.5 wt % GONPs reduced the residual impactor velocity by 33% and increased the energy absorption capability of GLARE by 22.9%. The SEM micrographs showed reinforced adhesion between the fibers and matrix and the optical microscope images showed improved cohesion in the matrix of the reinforced GLARE. This caused the secondary damage region in the composite layer to decrease considerably. However, GONPs imposed negative effect on the suppression of composite layer delamination and also interfacial strength between the composite and aluminum layers that provided more extensive plastic deformation of aluminum as an important energy absorption mechanism of GLAREs. Furthermore, the composite layers of the unreinforced and reinforced GLAREs were subjected to tensile test and the results showed improvements in the tensile stiffness, strength and toughness by 33.4, 45.0 and 25.6%, respectively, due to adding of 0.5 wt % GONPs to the matrix.
Keywords: fiber metal laminates, glass laminate aluminum reinforced epoxy, graphene-oxide nanoplatelets, highvelocity impact, damage mechanisms
© Heydari A., Khoramishad H., Alikhani H., Berto F., 2020
1. Introduction
In recent years, considerable attention has been devoted to developing innovative material design and efficient manufacturing techniques in various industries. Integrating different materials into one structure has been utilized as an efficient technique for achieving superior and tailored mechanical properties [1]. Polymeric composites are an important class of materials that have attracted a large number of researches focusing on their design and development to achieve multifunctional materials [2-4]. Although polymeric composites are widely employed in different industries, they suffer from some drawbacks such as low impact resistance because of their low deformability.
A technique of improving the impact resistance of composites is combining them with ductile metallic layers resulting in a class of structures named fiber metal laminates. This type of sandwich structures has received a wide range of applications in aerospace and other industrial sectors due to their high resistance against impact and fatigue loading [5]. Each of metallic and composite layers have some advantages and disadvantages. The low fatigue resistance of metals and brittle behavior of polymeric composites can be overcome by integrating them into a single structure. The high mechanical strength, fatigue and impact resistance, formability and fire resistance are some advantages of fiber metal laminates [6].
Another technique for improving the low impact resistance of polymeric composites and sandwich structures is introducing nanofillers into such brittle materials. The nanotechnology development provided considerable improvements in properties of polymeric composites. Many researchers have studied the effect of nanofillers on the mechanical behavior of composites and sandwich structures. Hosur et al. [7] studied the effect of adding nanoclays to a sandwich panel made of glass fiber-reinforced composite shells and a polyurethane core. 0.5 and 1.0 wt % nanoclays were dispersed in the polymeric core. The low-velocity impact tests with the impact energies of 15, 30 and 40 J were carried out on the specimens. They found out that added nanoclays restricted the damaged area and consequently improved the response of the sandwich panel by increasing the structural toughness. Gupta et al. [8] added 1.25 and 2.50 wt % cloisite 30B nanoclay and graphite nanoplatelet to the polyester resin, in order to modify the impact behavior, damping and curing properties of polymeric composites. The low velocity impact tests with strain rate of 15 s-1 were performed on the notched and unnotched specimens. The addition of 2.5 wt % graphite nanoplatelets improved
the energy absorption capability of unnotched composites by 100%, while reduced this parameter in the notched specimens by 70%. Naghizadeh et al. [9] investigated the high-velocity impact response of a sandwich panel containing plywood core and composite face-sheets reinforced with 0.3, 0.5 and 1.0 wt % mul-tiwalled carbon nanotubes (MWCNT). They reported that adding 1 wt % MWCNT, as the optimal weight percentage, to the epoxy and nylon 6 matrices of composite face-sheets enhanced the ballistic limit of the sandwich panel by 9.95 and 5.5%, respectively. Moreover, the presence of MWCNTs in the matrix restricted the damaged area caused by high-velocity impact loading.
Among nanofillers, graphene nanoplatelets have attracted considerable attention. Because of the special geometry and high strength of graphene nanoplate-lets, they can considerably improve the load-bearing capacity of polymeric composites. Bulut [10] worked on the influence of graphene nanoplatelets on the mechanical properties and impact resistance of ba-salt/epoxy composites. He showed that the presence of graphene nanoplatelets in the resin can considerably improve bonding between the matrix and fibers.
The impact response of fiber metal laminates has been attracted considerable attention in the field of impact mechanics in composites. Abdullah and Cant-well [11] studied the high-velocity impact response of fiber metal laminates made of woven glass fiber-reinforced polypropylene resin and aluminum sheets. Aluminum plastic deformation, aluminum rupture, composite delamination, fiber rupture and matrix cracking were reported as the main damage mechanisms. Fatt et al. [12] studied the ballistic limit velocity of fiber metal laminates experimentally and analytically and found out that 84-94% of the impact energy was dissipated by the flexural and cortical deformations of the aluminum layer and 2-9% of energy was dissipated through the composite layer delamination. Eslami-Farsani and Khazaie [13] investigated the effect of embedding shape memory alloys on high-velocity impact response of fiber metal laminates made of basalt fibers. The volume fraction and prestrain of wires were the considered variables. Two, four and eight shape memory alloy wires with different prestrain levels of 0, 2 and 4% were embedded in the fiber metal laminates. The eight shape memory alloy wires improved the energy absorption of fiber metal laminates by 40%. Moriniere et al. [14] concluded that the GLARE type of fiber metal laminate had the highest impact resistance compared to the other fiber metal laminate types. They explained that aluminum 2024-T3 could
be considered as the best metal layer in fiber metal laminates due to its high flexibility. Khoramishad et al. [15] worked on the influence of MWCNTs on the impact response and damage mechanisms of GLAREs. 0.25, 0.50 and 1.00 wt % MWCNTs were dispersed in the matrix of the glass fiber-reinforced composite layer. They reported an optimum MWCNT weight percentage of 0.5% resulting in an improvement of 19% in the energy absorption capability of fiber metal laminate. Moreover, the presence of MWCNTs in the matrix had a significant positive effect on the tensile strength of the composite layer and the adhesion between the fibers and matrix causing reduced damaged area and damage height in the composite layer of fiber metal laminate.
The objective of this paper was investigating the influence of graphene-oxide nanoplatelets on the impact and tensile behaviors of GLAREs paying considerable attention to the recognition of the nanoscale and microscale damage mechanisms caused by the na-nofillers. To ensure the efficiency of the dispersion process, the reinforced matrix was investigated using SEM micrographs. The fracture surfaces were assessed by SEM and optical microscopy techniques to determine the effect of GONPs on the nanoscale and microscale damage mechanisms of GLAREs, interfacial strength and damaged area.
2. Materials and methods
2.1. Materials
The fiber metal laminates considered in this research were GLARE type made of two 2024-T3 aluminum layers with a thickness of 0.8 mm and a composite layer in the middle with a thickness of 1.6 mm composed of six layers of E-glass woven fibers with a density of 200 g/m2 and the Epon 828 epoxy resin as the matrix of composite layer. The composite contained the fiber total volume fraction of 65 v/v%. A hardener named TETA was mixed with the epoxy resin with a ratio of 11 : 100. To reinforce GLAREs, gra-phene-oxide nanoplatelets with a thickness of 3.47.0 nm and a diameter of 4-20 (im were used in this study. Furthermore, the combination of GONPs and MWCNTs with the length of 30 (m, the outer diameter of 10-20 nm and the inner diameter of 5-10 nm were used. To achieve better dispersion and adhesion between the nanofillers and matrix, MWCNTs were functionalized with COOH (carboxyl) groups.
2.2. Manufacturing process
In order to obtain better adhesion between the metal and composite layers, a surface treatment process
was performed on the aluminum layers. For this, the aluminum surfaces were washed with distilled water and soap and then dried and cleaned with acetone. Afterward, sandpaper with a grain size of 600 was used to create tiny scratches on the surfaces followed by cleaning with acetone and cotton. To disperse the na-nofillers, a method consisting of mechanical stirring and ultrasonication were carried out. First, the epoxy resin was heated up to 40°C to reduce the resin viscosity and facilitate the dispersion process. Then, the mixture of epoxy resin and nanofillers was mechanically stirred with a velocity of 200 rpm. Afterward, ultrasonication was carried out at a power of 70 W for 50 min. After the sonication process, the curing agent was added to the mixture with a mixing ratio of 11 : 100 for the hardener to resin ratio followed by mechanical stirring for 5 min. After preparation of the matrix, GLAREs were fabricated by manual hand lay-up process. The specimens were cured under a temperature of 110°C for 3 hours while they were subjected to a uniform pressure of 4 bar. Due to the size of the gas gun fixture, the impact test specimens were manufactured with the dimensions of 100 * 100 * 3.2 mm3.
2.3. Test method
The high-velocity impact tests were conducted using a gas gun test machine. The gas gun test machine used N2 gas stored in the main tank with a pressure capacity of 150 bar. This main gas tank was connected to a capsule with a volume of 1 L and the same pressure capacity as the main tank. The impactor initial velocity was controlled as 235 m/s for all tests by setting a gas pressure of 50 bar for the gas capsule. The target was firmly clamped using a fixture in front of the gas gun barrel. To measure the residual velocity, the time of impactor passing through a chronograph with two optical sensors having a fixed and known distance was measured. A spherical impactor made of steel 1015 with a diameter of 8.7 mm and the weight of 2.71 g was utilized in the impact tests.
At the second step of experimental tests, the composite layer of GLAREs reinforced with 0, 0.25, 0.50 and 1.00 wt % GONPs were subjected to tensile loading, to assess the effect of adding GONPs on the tensile properties. The composite specimens were fabricated using identically the same method as used for the composite layer of GLAREs with the dimensions of 250 * 25 * 1.6 mm3. The tensile tests were performed according to standard ISO 3039 at a loading rate of 2 mm/min [16].
Fig. 1. Residual velocities (a) and absorbed energies of GLAREs (b)
3. Results and analysis
3.1. High-velocity impact test results
The residual velocity and absorbed energy were considered as the main outputs of the impact tests for evaluating the response of GLAREs against high-velocity impact loading. The impactor collision velocity of 235 m/s was considered for all specimens. The residual velocity was measured using a chronograph and the energy absorption was determined using the difference between the initial and residual kinetic energies of the impactor. The residual velocities and absorbed energies of the neat and reinforced GLAREs are compared in Fig. 1.
Adding 0.5 wt % GONPs reduced the residual velocity and enhanced the absorbed energy by 33 and 22.9%, respectively. The reinforced GLARE with 0.5 wt % GONPs had the minimum residual velocity and consequently experienced the maximum absorbed energy level. The improving effect of 0.25 wt % GONPs was considerably lower than that of 0.5 wt % GONPs. The presence of 0.25 wt % GONPs caused the energy absorption of GLARE to increase by 9.2% and the residual velocity to decrease by 11% relative to the neat specimen. Further increasing the amount of added GONPs to 1 wt % resulted in a negative effect on the impact response of GLAREs, such that the energy absorption decreased by 3.3% and the residual velocity
increased by 3.6% relative to the unreinforced specimen.
Furthermore, the obtained results were compared with the results of Ref. [15] in which the influence of MWCNTs on the impact response of GLARE was studied. It can be seen that the optimum weight percentages of GONPs and MWCNTs were both 0.5 wt %. It was reported in Ref. [15] that adding 0.5 wt % MWCNTs changed the residual velocity and the absorbed energy of the reinforced GLARE by -29.8 and 18.9%, respectively. Therefore, the reinforcing effect of GONPs at optimum weight percentage was about 4% higher than that of MWCNTs. Moreover, the hybrid reinforced specimens containing a combination of GONPs and MWCNTs, with a total weight percentage of 0.5 and with a mixing ratio of 1 : 1 were fabricated and subjected to impact and tensile loadings in order to study the reinforcing effect of the mixed GONPs and MWCNTs. The hybrid GONP/MWCNT-reinforc-ed GLARE did not experience considerable changes in their high-velocity impact responses compared to the corresponding single type nanofiller-reinforced GLAREs.
3.2. Tensile test results
The elastic deformation, delamination and fiber rupture are the main mechanisms of absorbing highvelocity impact energy in composite materials [17]. Moreover, toughness is a material parameter that indicates the ability of material in absorbing energy in the course of deformation. The tensile tests were carried out on the composite layer of GLARE to study the effect of nanofillers on the mechanisms involved in energy absorption. The influence of nanofillers on the tensile strength, modulus, strain at failure and toughness was experimentally studied. Figure 2 compares the behavior of the unreinforced and reinforced composite specimens against tensile loading.
It was found out that the trend of the results obtained from the tensile tests was similar to the trend of the high-velocity impact test results. By increasing the weight percentage of nanofillers from zero to 0.5 wt %, the tensile strength, stiffness and toughness were increased and afterward further raising GONP content to 1 wt %, the tensile quasi-static properties of the composite laminate decreased. However, the strain at failure was first decreased by adding up to 0.5 wt % GONPs and then increased by further raising nano-filler content to 1 wt % GONPs. The hybrid GONP/ MWCNT-reinforced composite layers had slightly lower tensile strength, stiffness and toughness and slightly higher strain to failure value compared to the
Fig. 2. The tensile properties of the unreinforced and reinforced composite layers: tensile strength (a), stiffness (b), toughness (c), strain at failure (d)
GONP-reinforced composite containing 0.5 wt % na-nofillers. The presence of 0.25 and 0.50 wt % GONPs in the composite layer improved the stiffness by 24.3 and 45%, respectively. While the 0.25 and 0.50 wt % GONPs-reinforced composites experienced 17.6 and 33.4% improvements in the tensile strength, respectively. The enhancements in the toughness due to the addition of 0.25 and 0.50 wt % GONPs were 21.4 and 25.6%, respectively, compared to the unreinforced composites. Moreover, the improvements in the tensile strength, stiffness, toughness of the hybrid GONP/ MWCNT-reinforced composites were 29.6, 41.4 and 23.3%, respectively. Increasing the weight percentage of GONPs to 1 wt % degraded the tensile properties of nanocomposites relative to the unreinforced specimens. The comparison between the results of tensile tests of the MWCNT-reinforced composites [15] and the GONP-reinforced composites showed more dominant reinforcing effect of GONPs than the MWCNTs.
3.3. SEM and XRD analyses
SEM micrographs and XRD analyses were used to assess the dispersion and nanostructure of nanofillers
and the improvement mechanisms. Figure 3 shows the dispersion of GONPs in the matrix. As can be seen in Fig. 3, d, GONPs were agglomerated in the samples containing 1 wt % GONPs. The nanofiller agglomerations that were observed more considerably in the specimens with a high weight percentage of GONPs was responsible for the limited improving effect of GONPs. Because the agglomerated nanofillers act as the local stress risers giving rise to material properties degradation as seen in the high-velocity impact and tensile behaviors of GLAREs and composite samples in this study [18].
Figure 4, a shows the reinforcing mechanisms introduced by MWCNTs leading to the higher energy absorption. Figure 4, b depicts the agglomeration of MWCNTs in the hybrid specimen that was a reason for the reduction in mechanical properties relative to the GLARE specimens containing 0.5 wt % GONPs. The presence of well-dispersed GONPs along with agglomerated MWCNTs is visible in Fig. 4, b.
To evaluate the dispersion quality and study the gra-phene-oxide nanostructure, the XRD analysis was performed on the graphene-oxide nanoplatelets, MWCNTs,
Fig. 3. The GONP dispersion in the matrix of composite layers: neat (a), 0.25 (b), 0.50 (c), 1.00 wt % GONPs (d)
Fig. 4. Improving mechanisms caused by MWCNTs (a); agglomeration of MWCNTs in the hybrid specimen (b)
10° 20° 30° 40° 28
Fig. 5. XRD patterns of the reinforced matrix and epoxy resin
the reinforced matrix and the neat epoxy resin. The results of XRD analyses are shown in Fig. 5.
The XRD pattern of graphene-oxide nanoplatelets indicates an intensive peak at 29 = 12.6°. The inter-layer spacing of graphene-oxide layers was obtained about 0.703 nm calculated by Bragg's law that was more than the distance of nanographene layers due to the embedment of oxygen functional group in the interlay er of graphite [19]. A weak diffraction peak was seen at 9 = 42.5° that can be attributed to incomplete oxidation. A broad peak at 29 = 26° and a weak peak at 29 = 43° were observed in the XRD pattern of MWCNTs indicating the interwall spacing in MWCNTs. The EPON 828 epoxy resin demonstrated a broad diffraction peak at 29 = 18° due to the amorphous nature of epoxy. The XRD patterns of the reinforced matrices with 0.25, 0.50 wt % GONPs and hybrid MWCNTs and GONPs were approximately similar to the epoxy resin XRD pattern indicating the homogeneous dispersion of nanofillers in the matrix, while a minor peak at 29 = 12.6° was observed in the XRD pattern of the sample with 1 wt % GONPs representing the agglomeration of GONPs in the matrix.
The effect of adding GONPs on the adhesion of glass fibers to matrix was investigated using the SEM micrographs. Figure 6 shows the SEM micrographs of the fracture surfaces of the unreinforced and reinforced composite layers with different weight percentages of GONPs. The cavities seen in Fig. 6, a are the locations of fiber pull-out from the matrix in the neat composite caused by the low interfacial strength between the glass fibers and matrix. However, in the specimens reinforced with GONPs, the fibers were fully embedded in the matrix signifying the positive
effect of nanofillers on the fiber-matrix bonding, even in the reinforced composite with 1 wt % GONPs.
Graphene-oxide nanoplatelets have an effective role in crack deflection because of their planer shape, high surface area and aspect ratio. Figure 7 depicts the deflection of crack growth due to the presence of GONPs. Forcing cracks to follow a longer path improves the material toughness by dissipating more energy. As there is a weak Van der Waals bond between the graphene layers and strong covalent bond between the carbon atoms in each layer, the crack growth path changed from being perpendicular to gra-phene layer to parallel with it, for easier and faster growing. The crack path change led to higher energy absorption of the composite layer.
Multiwalled carbon nanotubes can be considered as the rolled graphite sheets. There are some mechanisms induced by MWCNTs to enhance the strength of composite materials such as CNT pull-out, crack deflection, crack bridging, nanotube rupture and improvement of the fiber-matrix interfacial strength [15].
3.4. Damage mechanisms and patterns
The neat and reinforced impacted specimens were sectioned and investigated to assess the effect of GONPs on the damage mechanisms. Figure 8 shows the photos taken from the cross sections of the unreinforced and reinforced GLAREs by an optical microscope.
Matrix cracking and delamination between the composite laminas in an unreinforced GLARE are visible in Fig. 8, a. However, the addition of GONPs did not suppress the laminate delamination, yet more extensive delamination was observed in the composite
Fig. 6. SEM micrographs obtained from the fracture surfaces of the composite layers: neat (a), 0.25 (b), 0.50 (c), 1.00 wt % GONPs (d)
Fig. 7. FESEM image indicating crack deflection caused by GONPs
layer of the reinforced GLAREs (Fig. 8, b). This phenomenon was also previously reported in other studies [20]. This was in contrast with the effect of MWCNTs on improving the bonding between the laminas of the MWCNT-reinforced composite laminates [15]. This can be considered as the main difference between the improving mechanisms caused by GONPs and MWCNTs. The planar shape of GONPs can be accounted as the reason for this negative effect on the interfacial strength between the laminas. Whereas, MWCNTs improved the bonding between the composite laminas by pinning the laminas because of their high length-to-diameter ratio. This prevented the composite laminas from separation as shown in Fig. 8, c.
Fig. 8. Optical microscope photographs of the cross sections of neat GLARE (a), 0.5 wt % GONPs-reinforced GLARE (b), hybrid GONP/MWCNT-reinforced GLARE (c) and the adhesion of the composite lamina to aluminum layer in the unreinforced GLARE (d) (color online)
However, GONPs caused the matrix cracking in the an unreinforced GLARE. Figure 8, d shows that a lam-composite layer to be restricted. Figure 8, d provides a ina of glass fiber was completely bonded to the alumi-closer look at the lamina next to the aluminum layer in num surface that can be attributed to the higher com-
Fig. 9. Comparison between the diameters of the aluminum-composite deboned region: neat (a), 0.25 (b), 0.50 (c), 1.00 wt % GONPs (d), hybrid GONPs/MWCNTs (e) (color online)
Fig. 10. The primary and secondary damage regions in the composite layers: neat (a), 0.25 (b), 0.50 (c), 1.00 wt % GONPs (d) (color online)
posite-aluminum interfacial strength in the unreinforced GLARE than the reinforced GLARE while this was not seen in the reinforced GLAREs.
An important damage mechanism in GLAREs is the interfacial failure between the metal and composite. The sectioned views of the unreinforced and reinforced GLAREs are shown in Fig. 9. The addition of nanofillers weakened the composite-aluminum interfacial strength in the reinforced GLAREs. This allowed the aluminum layer of the reinforced GLAREs to deform more extensively relative to the unreinforced GLARE. Wider plastic deformation of the aluminum layer led to higher energy dissipation in GLARE.
The collision of impactor to a structure generates stress wave propagating through the whole structure until the full absorption of energy. The extent of da-
maged area is important in impact mechanics. The extension of damaged area can reduce the resistance of structure against the consequent impact loadings. The damage regions of the composite layers of the unrein-forced and reinforced GLAREs are shown in Fig. 10 when the aluminum layers were removed. As can be seen in Fig. 10, a change in color is visible in the damaged area indicating two different damage regions namely the primary and secondary damage regions. Damage in polymeric composites is a combination of matrix cracking, delamination, fiber breakage and resin-fiber separation [21]. The primary damage region outside the perforated area was more dominantly influenced by delamination taking place in the area closer to the impact location. However, by getting away from the primary damage region, the intensity of im-
Fig. 11. Comparison of the hole diameters of the composite layers: neat (a), 0.25 (b), 0.50 (c), 1.00 wt % GONPs (d), hybrid GONPs/MWCNTs (e) (color online)
pact energy was declined being unable to separate the composite plies. The dominant damage mechanisms observed in the secondary damage region were matrix cracking and separation of resin from fibers. Figure 10 indicates that the reinforced GLAREs containing GONPs had considerably smaller secondary damage region than the unreinforced GLARE. This was because the added GONPs to the matrix of composite reinforced the fiber-matrix interfacial adhesion and that limited the separation of resin from fibers. This was seen from the SEM micrographs (see Fig. 6) that bonding between the resin and fibers was improved in microscale when GONPs were introduced into the resin. Unlike the secondary damage region, the addi-
tion of GONPs in the matrix of composite layer extended the primary damage region to some extend compared to the neat specimen by reducing the interfacial strength between the laminas of composite signifying larger delamination in the reinforced composite layers as shown in Fig. 8, b.
During the high-velocity impact loading, the portion of laminate exactly underneath the impactor or very close to the impact location experiences fiber rupture and tensile flexural deformation because of the brittle behavior of polymeric composites giving rise to a circular hole with a specific dimension. The hole dimensions induced in the composite layer caused by impactor penetration are compared for the neat and re-
Fig. 12. Comparison between the damage heights for the composite layers reinforced with different nanofiller weight percentages of 0 (a), 0.25 (b), 0.50 (c), 1.00 wt % GONPs (d), hybrid GONPs/MWCNTs (e) (color online)
inforced composites in Fig. 11. As is illustrated in Fig. 6, adding GONPs to the composite layer and subsequently strengthening the interfacial bond between the matrix and fibers caused the load transfer in the fiber-matrix interfaces to be improved and consequently the matrix failure and fiber rupture were confined. Therefore, more energy should be spent on overcoming the bonding between the resin and fibers causing smaller holes in the reinforced specimens. The minimum hole diameter was obtained for the 0.5 wt % GONP-reinforced specimen. The performance of the specimen reinforced with a mixture of nanofillers was approximately similar to that of the 0.5 wt % GONP-reinforced specimen.
In high-velocity impact loading, in the course of projectile perforation, after fiber rupture and elastic and plastic deformation of composite layer, the fibers were pulled out from the laminate and formed an impact parameter called damage height which was measured for the unreinforced and reinforced specimens and shown in Fig. 12. Similar to the damaged area, the
addition of nanofillers diminished the damage height due to the improved adhesion between the fibers and matrix signifying stronger composite layer.
4. Conclusions
The effect of adding graphene-oxide nanoplatelets to the epoxy resin of fiber metal laminates was experimentally studied on the high-velocity impact response of GLAREs. The specimens were made of 2024-T3 aluminum, woven glass fiber and epoxy resin and reinforced with 0.25, 0.50 and 1.00 wt % GONPs and a hybrid of GONPs and MWCNTs. The residual velocity and absorbed energy of GLAREs were investigated. A decrease of 33% in the residual impactor velocity and an increase of 22.9% in the energy absorption capability were obtained by the addition of 0.5 wt % GONPs. Further increasing the added GONPs to 1 wt % did not cause higher reinforcement in the highvelocity impact behavior of GLAREs. Moreover, hybrid GONPs and MWCNTs resulted slightly lower improvements in the impact response of GLAREs. Vi-
sual and optical microscopy inspections and SEM micrographs were employed for assessing the damage mechanisms in the specimens and determining the underlying reasons of the improvements caused by the addition of nanofillers. The SEM micrographs showed improved adhesion between the matrix and fibers of the composite layer due to the addition of GONPs. Furthermore, the optical microscopy images showed the improved cohesion of the resin leading to limited matrix cracking in the presence of GONPs. However, GONPs imposed a negative effect on the delamination of composite laminates and the composite-aluminum interfacial strength. This caused the aluminum layers to experience considerably more extensive plastic deformation compared to the unreinforced specimen. Moreover, the addition of GONPs could considerably reduce the secondary damage region dimensions. Moreover, the tensile tests were performed on the composite layers of the unreinforced and reinforced GLAREs and a similar trend was observed for the improvements in the tensile stiffness, strength and toughness due to the presence of GONPs.
References
1. Jones R.M. Mechanics of Composite Materials. - CRC Press, 2014.
2. Feng J., Fu J., Shang C., Lin Z., Li B. Sandwich panel design and performance optimization based on triply periodic minimal surfaces // Comput. Aided. Design. - 2019. - V. 115. -P. 307-322.
3. Bourennane H., Gueribiz D., Freour S., Jacquemin F. Modeling the effect of damage on diffusive behavior in a polymeric matrix composite material // J. Reinf. Plast. Comp. - 2019. -V. 38. - P. 717-733.
4. Fiore V., Valenza A., Di Bella G. Mechanical behavior of carbon/flax hybrid composites for structural applications // J. Compos. Mater. - 2012. - V. 46(17). - P. 2089-2096.
5. Tsamasphyros G.J., Bikakis G.S. Analytical modeling to predict the low velocity impact response of circular GLARE fiber-metal laminates // Aerosp. Sci. Technol. - 2013. -V. 29(1). - P. 28-36.
6. Vlot A. Impact properties of fibre metal laminates // Compos. Eng. - 1993. - V. 3(10). - P. 911-927.
7. Hosur M.V., Mohammed A.A., Zainuddin S., Jeelani S. Impact performance of nanophased foam core sandwich composites // Mater. Sci. Eng. A. - 2008. - V. 498(1-2). - P. 100109.
8. Gupta S., Raju Mantena P., Al-Ostaz A. Dynamic mechanical and impact property correlation of nanoclay and graphite platelet reinforced vinyl ester nanocomposites // J. Reinf. Plast. Comp. - 2010. - V. 29. - P. 2037-2047.
9. Naghizadeh Z., Faezipour M., Hossein Pol M., Liaghat G., Abdolkhani A. High velocity impact response of carbon nano-tubes-reinforced composite sandwich panels // J. Sandw. Struct. Mater. - 2017. - 1099636217740816.
10. Bulut M. Mechanical characterization of basalt/epoxy composite laminates containing graphene nanopellets // Compos. B. Eng. - 2017. - V. 122. - P. 71-78.
11. Abdullah M.R., Cantwell W.J. The impact resistance of polypropylene-based fibre-metal laminates // Compos. Sci. Tech-nol. - 2006. - V. 66(11-12). - P. 1682-1693.
12. Fatt M.S., Lin C., Revilock D.M., Jr., Hopkins D.A. Ballistic impact of GLARE™ fiber-metal laminates // Compos. Struct. - 2003. - V. 61(1-2). - P. 73-88.
13. Eslami-Farsani R., Khazaie M. Effect of shape memory alloy wires on high-velocity impact response of basalt fiber metal laminates // J. Reinf. Plast. Comp. - 2018. - V. 37. - P. 300309.
14. Moriniere F.D., Alderliesten R.C., Sadighi M., Benedictus R. An integrated study on the low-velocity impact response of the GLARE fibre-metal laminate // Compos. Struct. - 2013. -V. 100. - P. 89-103.
15. Khoramishad H., Alikhani H., Dariushi S. An experimental study on the effect of adding multi-walled carbon nanotubes on high-velocity impact behavior of fiber metal laminates // Compos. Struct. - 2018. - V. 201. - P. 561-569.
16. ASTM Committee D-30 on Composite Materials, Standard Test Method for Tensile Properties of Polymer Matrix Composite Materials. - ASTM Int., 2008.
17. Koloor S.S.R., Ayatollahi M.R., Tamin M.N. Elastic-damage deformation response of fiber-reinforced polymer composite laminates with lamina interfaces // J. Reinf. Plast. Comp. -2017. - V. 36. - P. 832-849.
18. Yanovsky Yu.G., Nikitina E.A., Karnet Yu.N., Valiev Kh.Kh., Yumashev O.B., Semenov N.A. Quantum mechanics simulation and experimental study of adhesive interaction and aggregation of carbon-silicate nanoparticles—reinforcing fillers of polymer composites // Phys. Mesomech. - 2014. - V. 17. -No. 1. - P. 39-49.
19. Vu T.H., Tran T.T., Le H.N., Nguyen P.H., Bui N.Q., Essay-em N. A new green approach for the reduction of graphene oxide nanosheets using caffeine // Mater. Sci. B. - 2015. -V. 38(3). - P. 667-671.
20. Avila A.F., Neto A.S., Junior HN. Hybrid nanocomposites for mid-range ballistic protection // Int. J. Impact Eng. - 2011. -V. 38(8-9). - P. 669-676.
21. Chang P.Y., Yeh P.C., Yang J.M. Fatigue crack initiation in hybrid boron/glass/aluminum fiber metal laminates // Mater. Sci. Eng. A. - 2008. - V. 496(1-2). - P. 273-280.
Received 02.06.2020, revised 02.06.2020, accepted 05.06.2020
Сведения об авторах
Amir Heydari, MSc, Iran University of Science and Technology, Iran, amir2360heidary@gmail.com Hadi Khoramishad, PhD, Assoc. Prof., Iran University of Science and Technology, Iran, Khoramishad@iust.ac.ir Hossein Alikhani, MSc, Iran University of Science and Technology, Iran, hosseinalikhani1372@yahoo.com Filippo Berto, Prof., Norwegian University of Science and Technology, Norway, filippo.berto@ntnu.no
