Abstract
In this study, the thicknesswise fibers of the low-melting polyethylene terephthalate (LPET) nonwoven fabrics are needle punched and intertwined with the intra-laminar basalt fibers (BF) of basalt plain woven fabric in order to strengthen the stab-resistant property of LPET/BF sandwich composites as well as to fabricate armor that is composed of less lamination layers. Two LPET nonwoven fabrics and a BF plain woven fabric as an interlayer are laminated and combined using a needle-punch reinforcing method. The response surface analysis based on the Box–Behnken design is used to examine the influences of structure parameters of low-melting PET nonwoven fabrics including areal density (AD) and manufacture parameters including needle punching density (ND), and depth of needle punch (DP) on the spike stab resistance, knife stab resistance, bursting resistance, and tensile property. An empirical regression model of AD, ND, and DP is thereby established. The test results show that the bursting strength and quasi-static stab resistance of sandwich composites are highly dependent on AD and ND. Likewise, DP has a significant influence on the knife stab resistance and bursting strength, while the tensile strength is solely dependent on ND. According to the empirical regress model, the acquired optimal needle punching parameters of sandwich composites are an AD of 400 g/m2, ND of 143.77 needles/cm2, and DP of 6.41 mm. The 95% confidence interval yielded by the empirical regression model is in conformity with the test results. The empirical regression model of the stab resistance is proven to provide effective prediction of the number of lamination layers required by armor in the future.
Keywords
Introduction
Stab-resistant composites have broad application fields as in civil engineering (e.g., geotechnical cloth, filter material), constructional engineering (e.g., sound-absorbing fabrics, transportation engineering, packing material), and armor (e.g., stab-resistant armor, protective working clothing, and ballistic clothing) [1,2]. Especially in recent years, the occurrence of unexpected events in China continues soaring, which increase the duty load on police officers on special missions for the people and the nation. In China, the government has a severe ban standard against gun possession. Therefore, when encountering an incident, police officers are commonly found lethally hurt by knives or spikes. On the other hand, the individual protection field is limited to the cost. As a result, high-performance textiles are rarely used as stab-resistant armor, which makes metallic plates, ceramic plates, and PC boards the mainstream of the stab-resistant material [3,4]. However, these materials are cumbersome and have a low flexibility, the disadvantage that put the wearers at a higher risk. In response to the ever-complex protective occasions and a great diversity of stab apparatus, producing flexible stab-resistant composites that features low cost and high performance is critical and becomes a popular research topic regarding safe equipment for police officer.
High performance fibers, such as aramid fiber, ultra-high molecular weight polyethylene, and polyamide, has high strength, high modulus, high shear resistance and high impact resistance, which gives these materials the priority to be used in stab-resistant armors [5–10]. They are commonly made into woven fabrics, knitted fabrics, nonwoven fabrics, and braided fabrics in order to provide the stab-resistant composite fabrics with flexibility. The intrinsic properties, structure, elasticity, and strengths of the materials thus determine the level of stab resistance of the composite fabrics. Weaving fibers with a high count and a high density increases the friction between fibers, but the reinforcing effect of stab-resistant property appears limited [11,12].
Some researchers devoted to the improvement in the stab resistance of multi-layer woven fabrics. Reiners et al. placing one-layer woven wool structure on the top and bottom of the panels increased the stab resistance of resulting body armor [13]. Suhaimi et al. investigated influence of different stitching patterns on increase in the stab-resistant properties of multi-layer woven fabrics [14]. Messiry et al. designed a Vectran triaxial weave fabric to improve the impact performance in developing lightweight soft body with flexibility and comfort [15]. Besides, soft/layered structures such as 3D fully interlaced woven fabric, multiaxis 3D woven fabric and 3D braided fabric as well as 3D nonwoven can be also used for designing stab body armors [16]. This study proposes a woven/nonwoven hybrid stab-resistant structure.
In addition, the combination of nonwoven manufacture and woven manufacture combines the advantages of fiber compositions and structures. The complementary benefits of differing fibers simultaneously decrease the production cost without affecting the stab resistance of the hybrid composites. Our previous studies investigate how process parameters are in relation to the spike stab resistance [11,17,18], but the relationship between manufacture parameters and spike stab property is not established. Therefore, in this study, the hybrid structure and needle punching parameters are optimized using the Box–Behnken design, and a regression model is established based on manufacture parameters and stab resistance in order to provide a theoretical basis that indicates the number of lamination layers of armor.
Experimental programs
Establishing the experimental programs
The areal density (AD), needle punching density (ND), and depth of needle punch (DP) of low-melting polyethylene terephthalate (LPET) nonwoven fabrics have the corresponding level of −1, 0, and 1 as specified in the Box–Behnken design [19]; −1 is the lowest value of the parameter, 0 is the medium value of the parameter, and 1 is the highest value of the parameter. The experimental data of AD, ND, and DP are provided with corresponding levels in Table 1. Table 2 shows a total of 17 experimental programs as related to maximum spike stab resistance, maximum knife stab resistance, maximum bursting strength, and maximum tensile strength based on the Box–Behnken design.
The corresponding level to each parameter of LPET nonwoven fabrics.
DP: depth of needle punch; ND: needle punching density; AD: areal density.
Experimental groups with corresponding test results.
DP: depth of needle punch; ND: needle punching density; AD: areal density.
Materials
LPET nonwoven fabric (Hsinnjy Nonwoven Limited Liability Company, Taiwan) has an AD of 200, 300, and 400 g/m2, and a fiber fineness of 4 Denier (D). Basalt fiber (BF) plain fabric (Chin Carbon Fiber Technology Co., Ltd, China) has a warp density of 50 counts/10 cm, a weft density of 49 counts/10 cm, and AD of 251 g/m2.
Preparation of LPET/BF composite fabrics
There are 17 combinations of the samples. LPET nonwoven fabrics and BF plain fabrics have different ADs. Two surface layers of LPET nonwoven fabrics and one interlayer of BF plain woven fabric – are laminated and needle punched using a RSZ-80 needle punching machine (Rom Seen, China). The needle type is 15 × 16 × 25 × 3 1/2 M332 G 53017 (Groz-Beckert, Germany). During the needle punching process, ND and DP are changed based on different sample designs. Figure 1 shows the stab-resistant LPET/BF sandwich composites.

Diagram of the structure of an LPET/BF sandwich composite. LPET: low-melting PET.
Tests
The stab-resistant LPET/BF sandwich composites are tested in terms of the spike stab resistance, knife stab resistance, bursting strength and tensile strength.
An Instron 5969 (Instron, USA) is used to measure the spike stab resistance and knife stab resistance of samples as specified in ASTM F1342–91. The puncture rate is 508 mm/min, and the sample size is 100 mm × 100 mm. Figure 2(a) shows the spike probe and Figure 2(b) shows the knife probe. Ten samples for each specification are used for the tests.

Images of the (a) spike, (b) knife, and (c) burst probes.
An Instron 5969 (Instron, USA) is used to measure the bursting resistance of samples as specified in ASTM F2054–07. The bursting rate is 100 mm/min, and samples have a size of 150 mm × 150 mm. The probe has a spherical head with a diameter of 25 mm, as seen in Figure 2(c). Six samples for each specification are used for the test.
An Instron 5969 (Instron, USA) is used to measure the tensile strength of samples as specified in ASTM D5035–11. The tensile rate is 305 mm/min, and the sample size is 180 mm × 25.4 mm. The distance between gauges is 76 mm. Six samples for each specification are tested.
Results and discussion
Regression model establishment and experimental verification of LPET/BF sandwich composites
Table 2 summarizes the spike stab resistance, knife stab resistance, bursting strength-, and tensile strength of the 17 LPET/BF sandwich composites (i.e., experimental groups). A quadratic regression model is used to examine how structural parameters of LPET nonwoven fabrics (i.e., AD (x1)), and processing parameters of sandwich composites (i.e., ND (x2) and DP (x3)) are in relation to the spike stab resistance, knife stab resistance, bursting strength, and tensile strength in terms of 17 groups using the quadratic nonlinear equation (1). Each empirical regression model revealed in equation (1) includes independent, interaction, and quadratic effects, respectively.
Coefficient of the regression model of LPET/BF sandwich composites as related to the spike stab resistance, knife stab resistance, bursting strength, and tensile strength.
Based on the optimal structure parameters of LPET nonwoven fabrics and manufacture parameters of LPET/BF sandwich composites predicted by the regression model, samples are prepared as follows: the LPET/BF sandwich composites consist of optimal LPET nonwoven fabrics that are composed of an AD of 400 g/m2, ND of 143.77 needles/cm2, and DP of 6.41 mm. The sandwich composites are evaluated in terms of the spike stab resistance, knife stab resistance, bursting strength, and tensile strength tests. The test results of the samples are compared to the prediction of the empirical regression model as seen in Figure 3.

The comparison of the prediction of the empirical regression model and experimental results in terms of mechanical properties of LPET/BF composite fabrics. CV: coefficient of variation.
Comparing the test results to the prediction of the regression model, the coefficient of variation (CV) is 1.18% for the spike stab resistance, 2.45% for the knife stab resistance, 0.25% for the bursting strength, and 1.59% for the tensile strength. All CVs are within 5%, suggesting that the empirical regression model is consistent with the experimental results. This empirical regression model will provide the reliable prediction for the knife and spike stab-resistant performance of body armor in the future.
Effects of structure and manufacture parameters on the spike stab resistance of LPET/BF composite fabrics
Figure 4(a) shows that by increasing the AD of LPET nonwoven fabrics, the spike stab strength of LPET/BF sandwich composites increases linearly, while Figure 4(b) shows that by increasing the ND of LPET/BF sandwich composites, the spike stab resistance first increases and then decreases. The spike stab mechanism is primarily that the spike probe pushes fibers from contacting place to neighboring place while breaking some basalt fibers [20]. The images of fractured samples are shown in Figure 5(a) and (b).

The three-dimensional diagram of the spike stab resistance (N) of LPET/BF sandwich composites as related to (a) the interactive effect between AD and ND when the DP is 6.41 mm, (b) the interactive effect between ND and DP when the AD is 400 g/m2, and (c) the interactive effect between AD and DP when the ND is 143.77 needles/cm2. DP: depth of needle punch; ND: needle punching density; AD: areal density.

Images of fractured LPET/BF sandwich composites caused by a spike stab. (a) shows the fractured surface of BF layer (b) shows the fractured surface of nonwoven layer.
During the needle punching of the stab-resistant composite fabric, the fibers of the surface of nonwoven fabrics are pushed into inter-laminar to entangle with basalt yarns, forming vertical fibrous bundles. Increasing AD per unit area results in an increase in the amount of fibers as well as fiber compactness. A greater resistance the sandwich composites against the spike probe means a greater spike stab resistance [21]. A small ND causes a less amount of fibers that are intertwined vertically, which in turn weakens the inter-laminar bonding. The low friction thus leads to a low spike stab resistance. Conversely, an excessive ND makes the inter-laminar sandwich composites firmly bonded. Meanwhile, an increase in the amount of needle points over the surface of sandwich composites renders the yarns of BF plain fabrics with breakage, which has a negative influence on the spike stab resistance. When the DP increases, the spike stab resistance of sandwich composites first increases and then decreases. The maximum spike stab resistance occurs when the DP is 7.2 mm. DP and ND have a similar failure mechanism. A small DP loosens the compact level of composite fabrics, and thus a low spike stab resistance. Similarly, an excessive DP causes the damage to the fibers and structure of the composite fabrics, which result in a low spike stab resistance [11].
Effects of structure and manufacture parameters on the knife stab resistance of LPET/BF sandwich composites
Figure 6(a) shows that the knife stab strength of LPET/BF sandwich composites is linearly proportional to the AD of LPET nonwoven fabrics. Figure 6(b) shows that the knife stab strength of LPET/BF sandwich composites first increases and then decreases when the ND of LPET nonwoven fabrics is increased. In particular, the maximum knife stab resistance occurs when the ND is 200 needles/cm2. Similarly, Figure 6(c) shows that the knife stab resistance of sandwich composites first decreases and then increases when the DP of LPET nonwoven fabrics is increased and reaches the minimum when the DP is 7 mm. The failure mechanism of knife stab probe depends on the force of friction and shear force between the knife and the LPET fibers and basalt yarns, which in turn leads to the breakage of LPET fibers and basalt yarns [22,23]. Unlike the spike stab, the knife stab causes the breakage of the fibers as well as bundles of fibers of the sandwich composites [13,24]. The fractured patterns are shown in Figure 7(a) and (b). In particular, the AD of LPET nonwoven fabrics has a significant influence on the knife stab resistance of composite fabrics. Increasing the AD of the nonwoven fabrics means increasing the amount of fibers per unit area. The friction between the single fiber and the knife is constant, and the knife stab resistance of sandwich composites is thus in proportion to the amount of fibers strength [25].

The three-dimensional diagram of the knife stab resistance (N) of LPET/BF sandwich composites as related to (a) the interactive effect between AD and ND when the DP is 6.41 mm, (b) the interactive effect between ND and DP when the AD is 400 g/m2, and (c) the interactive effect between AD and DP when the ND is 143.77 needles/cm2. DP: depth of needle punch; ND: needle punching density; AD: areal density.

Images of fractured LPET/BF sandwich composites caused by a knife probe.
A low ND of LPET nonwoven fabrics results in a less firmly bonded intra-laminar structure. The LPET/BF sandwich composites cannot resist the stab of a knife probe. An excessive ND leads to the breakage of fibers of the basalt plain fabric. Basalt yarns have a greater anti-shearing efficacy than LPET fibers. As a result, the knife stab resistance of sandwich composites first increase and then decrease [26].
A small DP of nonwoven fabrics triggers slippage of fibers when they are damaged by a knife probe. At the same time, the sum of the force of friction and the anti-shearing force provides the LPET/BF composite fabrics with a greater knife stab resistance. A high DP creates a large amount of vertical fibrous bundles, decreasing the fiber slippage and strengthening the compact structure. Subsequently, the anti-shearing becomes higher. A high friction and a high anti-shearing efficacy thus provide the LPET/BF sandwich composites with a high knife stab resistance.
Effects of structure and manufacture parameters on the bursting strength of LPET/BF sandwich composites
Figure 8(a) shows that the bursting strength of LPET/BF sandwich composites increases when the AD of LPET nonwoven fabrics increases. Figure 8(b) shows that the bursting strength of sandwich composites diminished as a result of increasing ND. Figure 8(c) shows that the bursting strength of sandwich composites is inversely proportional to the DP of LPET nonwoven fabrics. The failure mechanism of the bursting strength is that the spherical probe extrudes the basalt yarns, which leads to the pull out of fibers, breakage of fibers, and breakage of upper and lower surfaces of LPET nonwoven fabrics. The fractured sample is shown in Figure 9. A small ND or a small DP of LPET nonwoven fabrics creates fewer and shorter vertical fibrous bundles. When encountering the round-headed probe, the basalt yarns slide and render the vertical fibrous bundled with a shear force. Then, the probe overcomes the friction to break the basalt yarns, resulting in a greater bursting strength of composite fabrics. In contrast, a large ND and a large DP of LPET nonwoven fabrics can firmly bond the sandwich composites using the vertical fibrous bundles. Therefore, the shear force can only exert a lower bursting strength over the sandwich composites [27] and the bursting strength of sandwich composites is in proportion to the AD of LPET nonwoven fabrics.

The three-dimensional diagram of the bursting strength (N) of LPET/BF sandwich composites as related to (a) the interactive effect between AD and ND when the DP is 6.41 mm, (b) the interactive effect between ND and DP when the AD is 400 g/m2, and (c) the interactive effect between AD and DP when the ND is 143.77 needles/cm2. DP: depth of needle punch; ND: needle punching density; AD: areal density.

Image of fractured sample caused by a bursting force.
Effects of structure and manufacture parameters on the tensile strength of LPET/BF composite fabrics
The tensile strength is an index to the stab resistance. Figure 10 shows that the ND of LPET nonwoven fabrics has a significant influence on the tensile strength of LPET/BF composites, and the tensile strength decreases with the increase of ND. The failure mechanism of the tensile strength of sandwich composites is because of the slippage of basalt yarns as well as the breakage of fibers. The BF plain fabric withstands the primary tensile force during the tensile test. Needle punch damages the basalt yarns of LPET/BF sandwich composites, and the greater the ND, the higher the damage. This finding is similar to the carbon fabric-reinforced composites [28]. As a result, the tensile strength of LPET/BF sandwich composites is inversely proportional to the ND of LPET nonwoven fabrics.

The three-dimensional diagram of the tensile strength (N) of LPET/BF sandwich composites as related to (a) the interactive effect between AD and ND when the DP is 6.41 mm, (b) the interactive effect between ND and DP when the AD is 400 g/m2, and (c) the interactive effect between AD and DP when the ND is 143.77 needles/cm2. DP: depth of needle punch; ND: needle punching density; AD: areal density.
Conclusions
The empirical regression model based on the Box–Behnken design is useful to predict and acquire the optimal structure and manufacture parameters of the LPET/BF sandwich composites consisting of two surface layers of LPET nonwoven fabrics and an interlayer of a basalt woven fabric. The empirical regression model predicts the optimum AD, ND, and DP that mutually interact, and the optimal parameters are in line with the experimental results. In contrast, the spike stab resistance first increases and then decreases with the increasing ND. Moreover, the knife stab resistance first decreases and then increases when DP increases. In addition, the AD, ND, and DP of LPET nonwoven fabrics are correlated with the bursting strength of the LPET/BF composite fabrics. The bursting strength increases when AD increases. The bursting strength increases and then decreases when ND increases, but decreases when DP increases. For the tensile strength of LPET/BF composite fabrics, only ND has the more significant influence. The tensile strength of LPET/BF sandwich composites is inversely proportional to the ND of LPET nonwoven fabrics. The stab-resistant empirical regression model proposed in this study is a powerful tool to predict and provide the valid number of lamination layers of armor in future. This study only presents the optimum structure and manufacture parameters of LPET/BF sandwich composites layers, which provides the research foundation for the lightweight stab-resistant materials in the future. In the following study, we will focus on the higher performance of stab-resistant composites by combinations of LPET/BF sandwich composites as the interlayer and Kevlar/Nylon nonwovens as the surficial layers reinforcing with multiple needle-bonding and thermal-bonding processes.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors gratefully acknowledge the financial support from the Open Project Program of Fujian Key Laboratory of Novel Functional Fibers and Materials (Minjiang University), China [No. FKLTFM 1704 and FKLTFM1722]; National Natural Science Foundation of China [grant numbers 51503145, 11502163, 11702187]; Natural Science Foundation of Tianjin [grant number 18JCQNJC03400]. This work is also supported by the Natural Science Foundation of Fujian [grant numbers 2018J01504, 2018J01505]; the Opening Project of Green Dyeing and Finishing Engineering Research Center of Fujian University [grant numbers 2017001B, 2017002B and 2017001A]; the Program for Innovative Research Team in University of Tianjin [grant number TD13-5043].
