# Tensile behaviors of layer-to-layer 2.5D angle-interlock woven composites with/without a center hole at various temperatures

Sep 18, 2020

### Effect of temperature on mechanical responses

Figure 2 shows the typical stress vs. strain curves of 2.5DWC without a center hole loaded at 20 °C, 180 °C and 240 °C, and the related strengths as well as hole factor are listed in Table 1. Here, the strength is calculated by dividing the maximum load by the effective load-bearing cross-section. It can be seen from Fig. 2 that the relationship between stress and strain is almost linear for each curve, followed by a slightly nonlinear at the end. Finally, a sudden failure happens, resulting in the loss of load-bearing capacity. It is very similar to that of unidirectional composites loaded along the fiber direction34,35.

The tensile stress–strain curves of open-hole 2.5DWC at different temperatures are illustrated in Fig. 3, and the strength as well as hole factor are listed in Table 1. As seen in Fig. 3, all the stress–strain curves of open-hole samples are similar, including an initial linear segment, nonlinear segment and sudden failure segment. Compared to the related curve of 2.5DWC without a center hole at the same temperature, the nonlinearity for open-hole 2.5DWC is more obvious, especially at 240 °C, demonstrating the influence of hole on the mechanical responses. It is also interesting to note that the open-hole samples at 180 °C and 240 °C can remain a certain degree of load-carrying and a brittle fracture mode in stress–strain curves also exists. It indicates that the failure mode of open-hole 2.5DWC is still dominated by fiber yarns. For the open-hole samples, a hole factor, Ch, is defined to assess the hole sensitivity of strength, as also listed in Table 1. The hole factor can be obtained as (C_{h} = left( {sigma_{u}^{notched} /sigma_{u}^{unnotched} } right)_{T}), where (sigma_{u}^{notched}) and (sigma_{u}^{unnotched}) mean the strengths of 2.5DWC with and without a center hole tested at the same temperatures. As shown in Table 1, the hole factors of samples at 20 °C and 180 °C are higher than those at 240 °C, resulting from the softening of resin at high temperature, as later confirmed in Fig. 5. This reason was also found in Ref.20, where the residual stress at elevated temperature was enhanced compared to that at room temperature. Additionally, when the temperature remains constant, the load-carrying capacity of open-hole 2.5DWC gradually decreases with the increase of hole size, which is consistent with the open-hole laminated composites29,37. Nevertheless, the extension of reduction induced by the enlargement of hole size is not significant at 240 °C, indirectly suggesting that the temperature plays a major role in the mechanical response at 240 °C.

The tensile strength of 2.5DWC with/without a center hole as a function of temperature is illustrated in Fig. 4. It can be noted that the strength decreases with the increase of temperature. The strength retention rates of open-hole samples with different hole sizes are obviously higher than that of 2.5DWC without a center hole when the temperature is lower than 180 °C, which can be ascribed to the principal load-bearing weft yarns and stress concentration accommodation. When the temperature rises to 240 °C, it is worth noting that the strength retention rates of samples without a center hole have a slight decline trend, indicating that the 2.5DWC without a center hole remains a high load-carrying capacity at 240 °C. The main reason is that although the mechanical performance of resin is significantly affected by the high temperature (closed to the Tg), the 3D net-shaped carbon fiber yarns (Fig. 1b–d) still maintain a certain load-carrying capacity. For the open-hole samples, the retention rates of open-hole samples at 240 °C experience a relatively large reduction, but the sample with a hole size of 8.5 mm still maintains a 65.79% retention rate, which is higher than the CF/Epoxy angle-ply woven composites studied in Ref.20. As mentioned above, the high tensile retention of CF/QY8911-IV 2.5D woven composites can be ascribed to the temperature-resistant characteristic of QY8911-IV resin, load-bearing warp yarns and stress concentration accommodation. The coupling effect of temperature and hole size can be quantitively expressed using Eq. (R1), SI.

### Effect of temperature on resin matrix

It is well known that resin matrix plays a significant role in the tensile responses and failure modes of fiber reinforcement composites at different temperature15,38. As shown in Fig. 1b,c, resin exists in the resin-rich domain and inside of fiber yarns. Figure 5 illustrates the dynamic mechanical analysis (DMA) curve of QY8911-IV resin and tensile stress–strain curves of CF/QY8911-IV [90]8 at 20 °C, 160 °C and 200 °C. The dimensions of samples for DMA and tensile tests are 15 mm × 15 mm × 5 mm and 300 mm × 25 mm × 1.59 mm, respectively. The aforesaid dimensions have been added in the revised manuscript It can be seen from Fig. 5a that the storage modulus, E’, declines with the increase of temperature from 20 °C to 200 °C, but it experiences a rapid drop, indicating that the mechanical properties of resin will obviously reduce after 200 °C. Furthermore, the glass transition (Tg) of ~ 256 °C is seen as a peak value in the energy dissipation, tan δ, vs. temperature curve. In addition, referring to the stress vs. strain curves (Fig. 5b), the tensile modulus and strength retention of QY8911-IV resin decline significantly at 160 °C. When the temperature rises to 200 °C, the modulus and strength retention experience a more severe declination. As listed in Table 2, the modulus and strength retention rates at 160 °C are 69.23% and 57.21%, and those at 200 °C only remain 48.80% and 34.05%. Compared to the corresponding data given in Table 1, the higher retention rates can be associated to the 3D net-shaped architecture joined together by fiber yarns.

Apart from the effect of temperature on the mechanical properties of resin, the bonding performance between fiber and matrix should be considered at elevated temperatures. The fracture morphologies as demonstrated in Fig. 6 suggest the bonding states of CF/QY8911-IV [90]8 unidirectional composites at various temperatures. As illustrated in Fig. 6a,b, large-scale matrix hackles are found along the fibers at 20 °C and 160 °C, and substantial fibers are completely covered by the thick matrix layer. It is suggested that the resin matrix remains a brittle characteristic at 20 °C and 160 °C, and the fiber and matrix is well bonded. However, at 200 °C, few resin adheres to fibers, and plastic hackles are noted, indicating that the softening of matrix at a large degree and possesses a weak interface bonding. It can be concluded from the viewpoint of resin matrix that for the 2.5DWC with/without a center hole, the softening of matrix at 180 °C in the resin-rich domain results in the rotation of load-carrying fiber yarns towards the loading direction, enhancing the load-bearing capacity to some degree. But, at 200 °C, excessive high temperature leads to massive failure of matrix and interface, and therefore the strength retention rates obvious drop (Fig. 4).

### Effect of temperature on fracture morphologies of 2.5DWC without a center hole

To be the best of our knowledge, the tensile behavior and failure modes greatly depend upon the experimental temperature for the samples with a center hole. Figure 7 presents the corresponding fracture morphologies at different temperatures.

At 20 °C, the macroscopic fracture morphology is considerably even, and the crack propagation direction is totally perpendicular to the loading direction, viz. weft direction. It should be noted that there is nearly no delamination and necking phenomenon near the fracture site, which is similar with the fracture morphology of unidirectional composites15, indicating that the 2.5DWC has a good structural integration. In addition, from the microscopic SEM images, inclined warp yarns in the failed cross-section are all broken, suggesting that they are the principal load-bearing elements. The fiber-dominated failure mode was also noted for other 3D woven composites39,40. Moreover, the transversal damaged weft yarns and brittle failed resin in the resin-rich domain are found, but fibers inside of fiber yarns are still covered by lots of resin.

Figure 7b illustrates that the fracture morphology at 180 °C is basically similar with that at 20 °C. First, there is no obvious delamination damage phenomenon near the fracture surface for each case. Second, both the fracture surfaces are quite even, and no necking phenomenon is noted. Third, the corresponding fracture direction is perpendicular to loading direction, giving an indication of a uniform load-bearing capacity. It indicates that the samples still possess a high load-bearing capacity at 180 °C, which is well agreement with the experimental data shown in Table 1. From the microscopic view, substantial broken inclined warp yarns in the cross-section are still found, alongside of transversal damaged weft yarns and brittle failed resin in the resin-rich domain. But, particularly, filament bundles in fractured carbon tow are bare, meaning that the interfaces among fibers are to some degree affected by temperature. At 240 °C, the color of matrix in the remote domain has turned from dark grey to mahogany (Fig. 7c), suggesting that resin becomes soften seriously. The fracture surface is no longer uniform, and delamination as well as pull-out damaged fiber yarns are clearly observed. But, there is still no necking phenomenon and the fracture site is totally restricted within the warp direction length of smallest repeating unit (Fig. 1c). From the SEM images, the inclined warp yarns in the cross-section are also broken, and massive broom-like broken fibers are found inside of fiber yarns. Apart from the failed resin, the transversal damaged weft yarns still exist in the cross-section. Moreover, the fibers inside of broken warp yarns are bare, verifying that the temperature plays an important role in the mechanical properties of resin and interfaces.

### Effect of temperature on fracture morphologies of open-hole 2.5DWC

Figures 8, 9, 10 illustrate the fracture morphologies of the open-hole 2.5DWC at 20 °C, 180 °C and 240 °C. Previous researches in terms of open-hole laminated/2D woven composites8,18,41,42,43,44 have shown that the large-area delamination fracture morphologies can be examined near the hole-edge domain, particularly at elevated temperatures. In this work, for the samples with a hole size of 4.1 mm, due to the occurrence of z-direction interlinkage (Fig. 1b,c), no obvious large-area delamination and necking phenomena are observed in the cross-section at 20 °C, 180 °C and 240 °C (Fig. 8), which also confirms that the fracture mechanism is dominated by the 3D net-shaped carbon fiber yarns. More specifically, from the macroscopic images, the failed cross-section at 20 °C shown in Fig. 8a exhibits a considerably even fracture morphology, but pull-out warp yarns are observed near the hole edge, which could be ascribed to the effect of hole-edge stress concentration. When increasing the temperature, more broom-like damage features are inspected, indicating that the interfaces of fiber yarns/yarns and fiber/matrix are remarkably affected by temperature. However, near the hole edge, there are still pull-out damaged warp yarns, and this phenomenon was also found by the Ref.32,45. The fracture sites at 180 °C and 240 °C are also restricted within a warp direction length. From the microscopic view, longitudinal warp yarn damage mode, transversal weft yarn damage mode and resin failure in the resin-rich domain are found in the cross-section. Moreover, all the inclined warp yarns are broken on the root. From the enlarged SEM images, plastic strips are observed among the fiber yarns, giving a suggesting of the brittle failure for resin in the resin-rich domain. And more and more bared fibers inside of fiber yarns are noted, revealing the resin inside of fiber yarns fails, giving rise to the weakness of interface performance.

For the samples with a hole size of 6.3 mm and 8.5 mm, shown in Figs. 9 and 10, the pull-out warp yarns near the hole edge and broom-like damage features without necking phenomenon are also observed at 20 °C, 180 °C and 240 °C, which are very similar with that of samples with a hole size of 4.1 mm. From the microscopic images, all the inclined warp yarns are broken on the root, and transversal damaged weft yarns as well as failed resin are also found around the broken warp yarns. Additionally, the number of bared fibers is also significantly increased as the rise of temperature. Thus, based on the aforesaid inspections, the influence of temperature on the fracture morphologies of open-hole 2.5DWC is more obvious than that of 2.5DWC without a center hole.

Compared to the fracture morphologies of samples without a center hole at the same temperatures, the fracture morphologies of open-hole ones are similar, apart from the hole-edge domain, which suggests that the effect of hole on the tensile behaviors is restricted within the hole-edge domain due to the 3D net-shaped architecture. It is very meaningful for the material to apply in the composite structures. However, the pull-out damaged warp yarns and later discussed strain behaviors (“Initial equivalent stress distribution”) are verified the hole-edge stress concentration.

### Initial equivalent stress distribution

Figure 11 exhibits the initial equivalent stress distribution of open-hole 2.5D woven composites subjected to warp direction loading at 20 °C and 180 °C (“X” axis in Fig. 11, initial pressure loaded on the end surface of samples is 0.1 MPa). It can be clearly seen that the stress distribution is not uniform. To be more exactly, the hole-edge regions emerge obvious stress concentration in comparison with the remote regions, especially the hole-edge warps component, the maximum stress of which is approximately 5 times greater than that of the remote regions (Fig. 11c,f). Moreover, the stress in warps (not to be considered the cut-off yarns) is also greater than that in the others, indicating that the warps are the primary load-carrying objects when it comes to warp loading, regardless of temperature condition (Fig. 11b,e). Likewise, due to the obstruction of consecutive warp yarns, stress concentration phenomenon is effectively restricted in a small region (about one warp’s width along the “Y” axis direction) rather than large-area stress concentration likely in notched laminated/2D composites 43.

In addition to stress distribution, the higher stress state at 180 °C is adequately suggestive of the influence of thermal stress induced by considerable discrepancy between fiber and resin (Fig. 11a,d). The warps’ stress distribution along the “Y” axis appears a certain degree of gradient distribution characterization, which could be attributed to the effect of thermal stress, but the warps still undertake the main loading.

### Localized strain responses at various temperatures

Figure 12 presents the localized strain vs. time curves obtained from the three strain gauges adhered to three signed sites (Fig. 1a). Here, the data onto the curves present a certain degree of discreteness and premature failure of some strain gauges, especially at elevated temperatures (Fig. 12b,c,e,g,h). But, in general, the real-time strains measured by the strain gauges “1” and “2” are quite similar (Fig. 12a,c,d,f), which are basically higher than those measured by the strain gauge “3” (adhered to the remote domain). It indicates that the stress concentration phenomenon still exists near the hole-edge domain for the open-hole 2.5DWC. When temperature remains constant, there is a linear relationship between the strain and time at 20 °C, which also confirms that the open-hole 2.5DWC possesses a great structural integration, thereby resulting in a brittle fracture characteristic. However, a nonlinear behavior in the strain- time curves is found at 180 °C or 240 °C, indicating that the effect of temperature on the mechanical behaviors. Interestingly, in most cases, even though the discrepancies indeed exist at various temperatures, the failure time points are basically same for the strain gauges measured adhered to the hole-edge sites and the remote site (Fig. 12). Additionally, it can be seen that the failure time gradually reduces with the increase of temperature for the samples with the same hole size. It indicates that the load-carrying capacity of open-hole 2.5DWC gradually weakens with the rise of temperature. Thus, it gives an indication of hole-edge stress concentration, regardless of temperature.