The repeated low-velocity impact response and damage accumulation of shape memory alloy hybrid composite laminates

2.1 Specimen preparation

For the traditional glass fiber-reinforced epoxy (GF/epoxy) composite, the vinyl ester resin purchased from Tianqi Chemical Factory, Dongguan, China, and the unidirectional glass fiber cloth provided by Nongchao Composites Co., Ltd. Yancheng, China, were employed as the matrix and reinforcement, respectively. With a weight ratio of 100:1, the matrix resin was fully cured with the addition of methyl ethyl ketone peroxide as a hardening agent at room temperature. The single layer thickness of glass fiber cloth with the 450 g·m⁻² surface density was 0.2 mm. The mechanical properties of glass fiber were 1.68 GP in tensile strength and 2.15% in failure strain, respectively. The superelastic 56.09 wt% Ni balance Ti wires with 0.2 mm in diameter were provided by PeierTech, Jiangyin, China. The finishing temperature of austenite phase transformation (A _f) was 10°C; thus, the SMA wires were in the austenite phase before suffering from impact loading. The mechanical performance of SMA is presented in Table 1. The stress-induced martensitic transformation occurred when the stress reached the upper plateau stress of 647 MPa, this phase transformation process would keep over a large strain as the impact loading increased. The austenite phase transformation appeared as the stress decreased to the lower plateau stress of 270 MPa during the unloading stage. Therefore, a hysteresis region in the stress–strain response was produced, which rendered the superior energy absorption capability.

The vacuum-assisted resin injection (VARI) method (29) was used to fabricate the traditional and SMAHC laminates, as shown in Figure 1(a). First, the release cloth, glass fiber layers, SMA wires, release cloth, and diversion net were placed on the workbench from bottom to up. Second, the full structure was covered with a vacuum bag with the aid of sealant. Then, a vacuum level of 600 mbar was created by a vacuum pump, and the mixed resin was injected into the whole structure. The resin was cured and the system was maintained for 20 h at room temperature. Finally, the testing specimens with dimensions of 100 × 100 × 3 mm³ were cut from the fabricated panels applying the water-jet cutting method. To improve the SMA/matrix interface performance, the impurities attached to the alloy surface were cleaned with acetone, and then using 400-grit paper to polish the surface. A typical cross-ply [0°₂/90°₂]_s stacking sequence was used for the traditional laminate. As presented in Figure 1(b), SMA wires with two perpendicular layers were embedded into the lower central region with the stacking sequence of [0°/0°/90°/90°/90°/SMA/90°/0°/SMA/0°]. A total of seven SMA wires with 0.3 mm in the gap were oriented to the adjacent glass fiber direction in each lay.

Figure 1

Schematic diagram of (a) specimen fabricated by VARI method and (b) positions of SMA wires.

2.2 Repeated low-velocity impacts

The Instron Dynatup CEAST 9340 drop-weight impact testing machine was applied to carry out the repeated low-velocity impacts based on the standard of ASTM D7136 (40). The testing system mainly contains the drop weight device, the data acquisition system, and the clamping fixture parts, as shown in Figure 2. The hemispherical projectile with a radius of 8 mm was 7.33 kg in mass. Two steel plates with a 76 mm circular cut-out at the center were used to fix the specimens firmly. During an impact event, the projectile was released at the setting height. Once contacting the specimen, the contact force was measured in real time by a piezo-electric load cell above the projectile head and was recorded by the data acquisition system. The continuous strikes were prevented by an anti-secondary impact device after the contact process. The 20, 25, and 30 J impact energies were investigated in this work, corresponding to the impact velocities of 2.34, 2.61, and 2.86 m·s⁻¹, respectively. After the previous impact, the impactor was consecutively raised at the same initial height to achieve repeated impacts until the occurrence of penetration. To ensure the reliability and accuracy of the experimental results, at least three repetitive tests were conducted for the traditional and SMA hybrid laminates at the same impact energy.

Figure 2

Set-up of low-velocity impact testing: (a) impact device, (b) drop weight tower and data acquisition unit, and (c) fixture.

3.1 Damage morphology

The representative damage morphology of the traditional and SMAHC laminates at the three different impact energies are shown in Figure 3(a–c). On the whole, the damage modes and damage propagation were similar for the two types of laminates regardless of the impact energies. Matrix cracking could be distinguished around the impact position, and delamination with a peanut shape was also found, especially in the last impacts in a front view. Compared to the front side, a larger delamination damage region was evidently seen from the rear side. The peanut shape delamination was formed in the first few impacts and then gradually developed into an elliptical shape until the perforation of the laminates. As indicated in previous studies (29,41), the delamination was produced at interfaces between two different fiber orientation plies and propagated along the orientation of the lower ply. Thus, the front and rear delamination areas are oriented in the 90° and 0° directions, respectively. The images clearly showed the generation of delamination damage during the first impact event for both laminates. Due to the appearance of fiber breakage, a small pit on the impact surface was induced beneath the impactor at the second impact for the traditional laminates at 20 J, the second impact at 25 J, and the first impact at 30 J, respectively. The pit caused by fiber breakage was generated at the fourth impact for the SMAHC laminates at 20 J, the second impact at 25 J, and the first impact at 30 J, separately. Moreover, the pits grew larger with the impact number increasing, as more fiber breakage was induced. As an example, the microscopic damage morphology of both laminates detected from the 25 J impact events is also shown in Figure 3(d). The damage modes including matrix cracking, fiber/matrix debonding, delamination, and fiber breakage were induced in the traditional laminate. Compared to the traditional laminate, the damage of SMA/matrix debonding was also found in the SMAHC laminate. Moreover, the bottom SMA wires did not fail and the damage of fiber breakage was caused around the SMA wires.

Figure 3

Damage morphology of the traditional and SMAHC laminates: (a) 20 J, (b) 25 J, (c) 30 J, and (d) microscopic damage morphology.

3.2 Dynamic mechanical behaviors

The repeated impact behaviors, contact force–time/displacement, and energy–time histories of the traditional and SMAHC laminates at 20 J impact energy are presented in Figure 4. On the whole, both laminates underwent the rebounding stage and the penetrating stage until the perforation during the repeated impacts. For both laminates during the first impact event, a local fluctuation was observed in the initial rising phase of the force–time/displacement histories. As indicated in studies (42,43), this unstable phenomenon was mainly induced by the rapid development of matrix cracking and the occurrence of delamination, which could be observed visually from the damage morphology presented above. Drops and fluctuations were also seen before the contact force reached the maximum value, which resulted from the extension of delamination (30,42). As the impact number increased, a distinguished impact force drop immediately after the peak force was found at the second impact, and a sharp drop at the fourth impact for the traditional laminate. The first evident drop of impact force was discovered at the fourth impact for the SMAHC laminate. The occurrence of fiber breakage resulted in the significant degradation of the load-bearing capacity of composite laminates, which was in agreement with the results in the literature (43,44). Meanwhile, the absorbed energy had a large increase due to the high energy dissipation capability of fiber breakage damage. Finally, an approximately constant force plateau representing the perforation appeared at the 9th and 14th impacts for the traditional and SMACH laminates, respectively. As described by Papa et al. (45), the friction between the projectile and specimen led to a roughly stable contact force. Furthermore, as indicated in energy–time curves, the impact energy was completely absorbed in the perforation impact.

Figure 4

Impact behaviors for (a) traditional and (b) SMAHC laminates at 20 J.

The dynamic mechanical responses of the two types of laminates under repeated impacts at 25 and 30 J impact energies are also plotted in Figure 5. For the traditional laminates, the appearance of fiber breakage was at the second and first impact under 25 and 30 J, respectively. In addition, the perforation occurred at the fifth and fourth impacts under the two impact energies, separately. For the SMAHC laminates, the fiber breakage was caused by the second impact for both energies. The failure occurred at the seventh and fifth impacts under 25 and 30 J, respectively. On the whole, compared to the impact behaviors at 20 J, the impact number to cause the fiber breakage damage and perforation reduced for higher impact energy levels. This is because severe damage was caused at the same impact number, and the damage accumulation was faster for higher impact energy. Moreover, the fluctuations resulted from different damage modes were more intense for laminates at higher impact energy, due to the more unstable propagation of damage. It is evident that more impact number was needed to cause the failure for SMAHC laminates than traditional laminates for all impact energies, which suggested that the damage tolerance of SMAHC laminates was better. The impact resistance of SMAHC laminates was improved attributed to the hybrid SMA wires; thus, more impact energy could be absorbed and the value of repeated impact number increased.

Figure 5

Repeated impact responses for (a) traditional and (b) SMAHC laminates at 25 J and (c) traditional and (d) SMAHC laminates at 30 J, respectively.

4.1 Analysis of impact responses

The variations of peak force and maximum displacement for the two studied laminates are plotted in Figure 6. On the whole, the peak force first had a little rise at the initial few impacts. As explained in research (43,46), the compaction effect produced by the enhanced central region of the laminate contributed to the rise of peak force. For the traditional laminate at 30 J, the peak force of the second impact decreased compared with the first impact, resulting from the appearance of fiber breakage. The peak force then reduced slowly with the development of dominant damage of delamination, followed by decreasing rapidly with the intensified propagation of fiber breakage. Finally, the peak force had a disastrous reduction at the perforation impact due to the failure of the laminates. Results showed that the peak force of the two studied laminates increased with the impact energy rising at the first impact, which had been demonstrated in works (23,37). Moreover, compared with the SMAHC laminates, the peak force of the traditional laminates was lower at the same impact number regardless of impact energy. The trend of the maximum displacement shows that it first increased slowly with the dominant expansion of delamination, and then grew rapidly with the development of fiber breakage. At the same impact number, the maximum displacement of SMACH laminate was smaller in comparison with the traditional laminate. Furthermore, the maximum displacement of traditional laminate increased much faster than that of hybrid laminate, which suggested that the damage accumulation was relatively faster for the traditional laminate. As indicated in the literature (16,17), the impact properties were improved for SMAHC laminate thanks to the hybridization, and part of the impact loading was undertaken by the embedded SMA wires; thus, the load-carrying capacity of the SMAHC laminate was strengthened and the resistance to deformation was enhanced.

Figure 6

Variations of peak force and maximum displacement at (a) 20 J, (b) 25 J, and (c) 30 J.

The change in absorbed energy and the total absorbed energy of the two laminates are presented in Figure 7. The absorbed energy decreased at the initial few impacts for both laminates at low-impact energies. As reported in the literature (43,47), the first strike produced major damage to the laminate as the induced damage was dominated by the delamination. As discussed above, the impact load was principally supported by the intact fibers instead of the failed matrix at the laminate center owing to the compaction effect, and the delamination extension was also restrained. Meanwhile, a little more energy was stored in the laminate with a larger deflection, which would be released at the rebounding phase. The dissipated energy increased at the second impact for both laminates due to the serious damage of fiber breakage. Then, the dissipated energy increased with the rise in impact number. Finally, the absorbed energy reached the value of impact energy, which indicated the occurrence of perforation. The total absorbed energies of traditional laminates in Figure 7(b) were 169.81, 119.74, and 112.83 J under 9, 5, and 4 repeated impacts at the 20, 25, and 30 J impact energies, respectively. The values were 241.65, 166.67, and 137.96 J for SMAHC laminates with a total of 14, 7, and 5 impacts at the 3 different impact energies separately. As a result, an increase of 42.31%, 39.19%, and 22.27% was acquired for the energy absorption of hybrid laminates over traditional laminates, respectively. Hence, higher damage tolerance was achieved for SMAHC laminates. Furthermore, at the same impact number, the energy dissipation of SMAHC laminate was lower than that of traditional laminate, which demonstrated less damage was induced and the impact performance of hybrid laminate was better. As revealed in works (14,18), the hysteresis mechanical behavior of SMA wires caused by the stress-induced martensitic transformation could dissipate large impact energy as the embedded SMA wires did not fail during the impact event. So less impact energy was dissipated by the failure of resin matrix and glass fiber. Thus, less damage was induced in hybrid laminate. Moreover, the debonding of SMA/matrix interfaces and the friction between them also dissipated some impact energy. All this proved that the repeated impact resistance and damage tolerance of SMAHC laminate were enhanced due to the hybridization.

Figure 7

(a) Change of absorbed energy and (b) total absorbed energy for both laminates with different repeated impacts.

4.2 Damage evolution and accumulation

The damage evolution of composites suffered from repeated impacts was extremely complicated attributed to the damage accumulation. The damage evolution was closely in connection with the extension of different damage modes. As suggested in previous studies (29,30), the delamination projected area as an important damage characteristic could be used to assess the damage degree of composites. Figure 8 presents the damage areas of the two laminates at the three different energies. The damage area gradually increased first and then approximately reached a constant value until the perforation. After the full development of delamination, the increasing rate of the damage area was slowed, as more impact loading was supported by fiber. As indicated in previous studies (48,49), the delamination saturation would reach when no new large delamination developed. The final delamination projected area of traditional laminates was about 5,306, 6,529, and 6,869 mm² at 20, 25, and 30 J, respectively. The damage area was about 3,768, 5,078, and 5,613 mm² for the SMAHC laminates at the same impact number under the three impact energies, which had a much reduction of 28.99%, 22.22%, and 18.29% compared to the traditional laminates, respectively. Furthermore, the ultimate damage areas of SMAHC laminates were about 4,438, 5,532, and 5,871 mm² at the studied impact energies, which were also lower than those of the traditional laminates. Therefore, the impact resistance of SMAHC laminate was better as less damage was induced in comparison to the traditional laminate. As discussed above, some impact energy was dissipated by the excellent superelastic mechanical performance of SMA wires. Thus, the impact performance of SMAHC laminate was improved and the damage degree was reduced attributed to the enhancement mechanism of embedding SMA wires.

Figure 8

Damage areas of the two studied laminates.

The damage accumulation of composites under repeated low-velocity impacts could well be evaluated by the damage index DI-B developed by Liao et al. (30). The index DI-B not only revealed the occurrence of perforation clearly but also characterized the damage accumulation accompanied by different damage modes. Thus, the damage index DI-B was applied to evaluate the damage accumulation of the two studies composites in this work. The index DI-B was defined as

where d _max stands for the maximum displacement of each impact and d _p represents the perforation displacement. The R _s is the impact bending stiffness reduction rate, expressed as

where k is the bending stiffness, which was determined from the slope of the ascending section in force–displacement curves. k _i is the impact bending stiffness of ith impact, and k ₀ and k _f are the initial and final bending stiffness under repeated impacts, respectively.

Figure 9 shows the variation of impact bending stiffness and the trend of damage index DI-B in terms of the impact number. As revealed in the literature (29,30), the variation of bending stiffness was strongly related to the induced damage. The bending stiffness first gradually decreased due to the dominated damage of the delamination at the lower 20 and 25 J impact energies. Then, there was a large reduction of the bending stiffness immediately after the occurrence of severe fiber breakage. Finally, with intensified propagation of the fiber breakage, the bending stiffness reduced rapidly until the perforation. The bending stiffness decreased rapidly until the perforation resulted from the premature occurrence of fiber breakage at the higher 30 J impact energy. It is evident that the bending stiffness of SMAHC laminates was higher at the same impact number and the rate of decline was much slower, compared to the traditional laminates. This is because that less damage was caused for hybrid laminate in each impact event, and thus, the degradation of mechanical performance was relatively weaker.

Figure 9

(a) Variation of bending stiffness and (b) tendency of index DI-B under repeated impacts.

As indicated in Figure 9(b), the damage index DI-B was monotonically increased from zero of initial no damage to one corresponding to the failure. In agreement with the tendency of bending stiffness, the damage accumulation dominated by the delamination and matrix cracking, fiber breakage, and the occurrence of perforation could be described by the index DI-B. For the same impact energy, SMAHC laminates showed lower DI-B values at the same impact number, indicating less damage generated in each impact and thus better impact resistance. It can also be observed that the value of DI-B was higher for both laminates with the rise of impact energy at the same impact number. The reason is that severe damage was induced as the impact energy increased. Furthermore, the rate of increase in the DI-B was faster with less repeated impact numbers at higher impact energy. With impact energy increasing, the damage of SMA/matrix debonding was easier to produce, and the synergistic action of the embedded SMA wires and matrix was suppressed due to inhomogeneous deformations in hybrid laminate. As a consequence, the excellent energy dissipation capability of SMA was restrained, and more impact energy was absorbed by the damage of delamination and fiber breakage. Therefore, the damage accumulation was more rapid with the impact energy rising, and the positive effect of SMA hybridization on the impact resistance of composites was less significant.