High velocity impact properties of composites reinforced by stitched and unstitched glass woven fabrics

Abstract

The present research aimed to investigate the effect of stitching angle and stacking sequence of stitched layers on high velocity impact behavior of composites reinforced by glass woven fabrics. To study the effect of stitching angle on ballistic impact behavior, six different angles of (0°), (90°), (45°), (0°,90°), (±45°) and (0°,90°,±45°) were chosen as stitching angles. These stitching angles were applied on eight layers of glass woven fabric. To study the effect of stacking sequence of stitched layers, a different number of layers were stitched together with the angle of 0°. Unstitched and stitched composites were exposed to high velocity impact with 180 m/s using a spherical projectile. The residual velocity of projectile and dimensions of damage area on the composites’ front and back sides were measured. It was found that the sample with the 45° stitching angle had the best behavior against ballistic impact and its energy absorption was significantly higher than the other samples. Stitching also reduces damage area in front and back sides of the composites and inhibits delamination.

Keywords

Stitching high velocity impact delamination glass fibers

Introduction

Composites are increasingly used in different industries due to their low weight and high mechanical properties such as high strength, corrosion resistance and high fatigue resistance. Despite the widespread use of composites in various structural applications, their usage has been restricted because of some issues in the manufacturing process or mechanical properties. The costly process of composite manufacturing, poor drape of prepregs, low through-thickness mechanical properties, delamination and low impact resistance are some of the most important limiting factors in composites [1,2]. The vulnerability of composites to impacts and their poor interlaminar resistance is a serious problem, especially when used in aircraft, military, or marine industries since they are always at risk of high or low velocity impacts. Dropped objects and tools during service and maintenance, bullets, bombs and grenade fragments, hails, birds and rocks cause penetration and delamination in composites, which reduce their after-impact properties dramatically [1 –6]. Different approaches have been proposed to improve the impact properties of composites. The methods that have been the subject of more interest are using toughened matrices, modifying adhesion of reinforcement and matrix and reinforcing composites through the thickness [7]. Reinforcing composites along their thickness is a more desirable method because the other two methods are highly expensive and require extensive research [1,7 –10]. Several methods have been proposed to reinforce composites in the thickness direction. Using 3 D fabrics (woven and knitted), braiding fabrics, stitching and z-pinning are the most important procedures [1,7–9,11]. Stitching is an easy, low cost and highly productive method [1,7,9]. In addition, manufacturing large dimension composites and complex shapes is easier by stitching. It is also a suitable alternative for mechanical fasteners, e.g. bolts and rivets. Another advantage of this method is that stitched yarns can keep a composite’s integrity due to their high tensile strength even after the composite is fractured. This method will prevent fatal disasters when using stitched composites in fan blades for aero engines because, in such applications, fragments of a damaged blade could be sucked back into the engine [3]. Stitching also facilitates the handling of preform plies by preventing them from slipping during composite production [1,3,7,9]. In addition to fasteners, stitched composites are also used in applications where they are subjected to high or low velocity impacts, such as aircrafts’ fuselage and wing panel, marine industries, and wind turbine blades [1,2]. The effect of through-the-thickness reinforcement on the mechanical properties of composites has been investigated in some previous research. The results of the researches are widely varied from each other. In some researches, reinforcing composites through their thickness by stitching or Z-pinning improved in-plane mechanical properties, while in some others, it left composites without significant changes. In some cases, through-the thickness reinforcement even deteriorated composites’ performance. This degradation in in-plane properties is due to the damage which occurs during the stitching process. This damage includes fiber breakage and fiber misalignment caused by needle and stitch yarn, leading to resin-rich regions around stitches in composites [1,3,7,9,11 –15]. However, delamination resistance was always improved in the studies in which impact behavior was investigated [1,2,4,7 –9,12,16,17].

Several studies investigated the impact behavior of trough-the-thickness reinforced composites under low and high velocity impacts. It was found that the stitch yarns or Z-pins were not able to prevent the initiation of cracks. However, they could effectively suppress crack propagation, which resulted in the restriction of delamination. Moreover, stitching yarns cause a proper distribution of energy in the composite structure. As a result, the damage area is smaller, and the composite structure’s integrity will be preserved [7 –9,18 –25]. The post-impact strength and mechanical properties of stitched composites are higher because of the smaller damage area and suppression of delamination [14]. In one research, stitching could increase the ballistic limit of composites up to 10%. The ballistic limit was defined as a projectile’s velocity which has a 50% probability of penetrating the specimens [9].

Although the effect of stitching on ballistic impact behavior of fiber-reinforced composites has been investigated, little attention was given to stitching parameters such as stitch angle or stitching a different number of layers for thick composites that stitching all layers might be impossible. This paper presents an experimental study to investigate the effect of stitching angle and stacking sequence of stitched layers of glass-reinforced composites subjected to ballistic impact.

Materials and methods

Materials and manufacturing

In each composite, eight layers of woven E-glass fabrics were used as reinforcement and epoxy resin (ML-506, Mokarrar Ind. Group) as the matrix. To study the effect of stitching angle, eight layers of fabrics were stitched together using filament polyester 30 tex as stitch yarn with different stitch angles 0°, 45°, 90° and a combination of them. The sewing machine was Durkopp-Adler model 271. Stitch angle was described as the angle between stitch lines and the warp direction of the fabric. Pitch length and space between stitch lines were constant of 4 mm, and the modified-lock stitched pattern was used. Figure 1 shows the schematic diagram of stitching arrangement in composites, and the descriptions of specimens are shown in Table 1. In this table, every G represents one layer of glass fabrics; those layers that are stitched together are given in parentheses. The layers that are not stitched together are separated by a slash mark. To investigate the effect of the stacking sequence of layers, fabrics were stitched with different orders as described in Table 1. For specimens B-0, C-0, D-0, E-0 and F-0, the stitch angle of 0° was chosen. It is noteworthy that the difference in thickness of composites with different stitch angles is due to the difference in the samples’ stitch density. In composites with different stacking sequences of layers, the thickness variation is because of the difference in stitched layers and their sequence in samples.

Figure 1.

Schematic of composite structure: (a) stitching arrangement; (b) modified-lock stitch.

Table 1.

Properties of composites.

Specimen	Stacking sequence of layers	Description	Thickness (mm)	Fiber volume fraction (%)	Void content (%)
A-U	[G/G/G/G/G/G/G/G]	8 unstitched layers	1.81	60.47	4.41
A-0	[(GGGGGGGG)]	8 layers stitched with 0°	2.07	61.18	5.07
A-90	[(GGGGGGGG)]	8 layers stitched with 90°	1.95	53.83	5.46
A-45	[(GGGGGGGG)]	8 layers stitched with 45°	2.16	54.75	4.98
A-±45	[(GGGGGGGG)]	8 layers stitched with ±45°	2.29	56.88	6.12
A-0,90	[(GGGGGGGG)]	8 layers stitched with 0° and 90°	2.22	56.15	6.33
A-0,90,±45	[(GGGGGGGG)]	8 layers stitched with 0°, 90° and ±45°	2.51	51.26	9.56
B-0	[(GG)/(GG)/(GG)/(GG)]		2.58	59.26	6.99
C-0	[G/G/(GGGG)/G/G]		2.41	67.81	6.15
D-0	[(GG)/(GGGG)/(GG)]		2.82	64.59	7.61
E-0	[G/(GGGGGG)/G)]		1.67	46.43	6.42
F-0	[(GGGG)/(GGGG)]		2.42	63.24	7.02

For specimens with different stitching angles, eight layers of dry glass fabrics were placed on each other and taped together on the edges in order to prevent slipperiness during the stitching. Then, the layers were stitched together, according to Table 1. For specimens with different stacking sequences of layers, the numbers of dry glass fabrics were placed on each other, according to Table 1, and then were taped on the edges and stitched together. The final specimens were prepared according to Table 1. For example, for specimen B-0, two layers of dry glass fabrics were stitched together. Then four sets of such stitched layers were placed on each other and taped together on the edges to prevent slipperiness during the vacuum infusion process. After preparing all specimens, they were placed in a sealed vacuum bag. The resin was then vacuumed into the bag using a vacuum pump. The applied pressure was 0.8 bar. Composites were cured for 72 hours at 25 ± 1°C, as recommended by the resin manufacturer. The samples were prepared according to ASTM D8101/D8101M - 17 [26] with dimensions of 10 × 10 cm². Figure 2 shows the manufacturing and preparation of the specimens.

Figure 2.

Manufacturing specimen by vacuum infusion process.

High velocity impact

The high velocity impact test was carried out according to ASTM D801/D8101M – 17 [26]. Three samples were tested for each specimen, and they were subjected to a high velocity impact using a gas gun (Figure 3(a)). The gas gun consisted of a pressure tank, a firing valve, a 5-meter barrel and a capture chamber in which the samples were placed. For all samples, a spherical stainless steel bullet with 8.7 mm diameter and weight of 2.71 g was used as the projectile (Figure 3(b)). As projectile speed is directly related to gas pressure, a pressure of 20 bars was chosen, which made a 180 m/s speed. Optical sensors were sued for measuring the residual velocity of the projectile. Then, the mean of residual velocities of the three tests of every specimen was calculated, and the energy absorption for each specimen was obtained using the conservation of energy equation

E = \frac{1}{2} m (V_{1}^{2} - V_{2}^{2})

(1)

where E is the absorbed energy by specimens, m is projectile weight and V₁ and V₂ are projectile’s incident and residual velocity, respectively. Frictional resistance between projectile and air was neglected.

Figure 3.

(a) gas gun; (b) projectile.

Investigation of damage

After the test was completed, all the samples were scanned using a scanner. Then the damage areas on the front and back sides of composites were calculated by image processing using Adobe Photoshop CC 2017. The front and back sides of composites are the surface that projectile strikes and the surface that projectile exits, respectively. However, the composite’s damage area is not adequate to investigate the composites’ impact behavior because they absorb energy through different mechanisms such as friction, shear, crack in the matrix, delamination, fracture and rupture of fibers, fiber separation from matrix, deformation and perforation. Therefore, for a more detailed investigation into composites’ destruction, scanning electron microscope (SEM) pictures of the damage surface and cross-section of damage area of composites were also prepared. For SEM pictures, the scanning electron microscope (Seron, model AIS-2100) was used.

Results and discussions

Effect of stitch angles

Calculated damage areas in the front and back sides of samples are reported in Figure 4. The residual velocity of projectiles and composites’ energy absorption are indicated in Figures 5 and 6, respectively. These figures were statistically analyzed using IBM SPSS Statistics for Windows, version 22.0 (Anova analysis). Scanned pictures of composites are shown in Figure 7, and SEM pictures of the surface and cross-section of samples are given in Figure 8.

Figure 4.

The damage area of specimens.

Figure 5.

The residual velocity of specimens.

Figure 6.

The absorbed energy of specimens.

Figure 7.

Front side (FS) and back side (BS) of composite specimens after impact.

Figure 8.

SEM pictures of cross-section (C) and surface (S) of composite specimens.

Complete perforation happened in all composites. As shown in Figure 4, stitching successfully reduced damage area through restricting delamination. Stitching has also reduced the projectile’s residual velocity in most of the specimens (Figure 5); therefore, improved their energy absorption (Figure 6). It is also obvious in Figure 7 that the delamination of A-U on both sides of the composite is larger than other composites. This result is in agreement with the findings of previous studies [2,7,8]. In pictures of composites’ surface (Figure 8, A-0(S), A-0,90(S), A-45(S)), it is obvious that stitching yarns prevented the growth of damage and fiber fracture. Sample A-0,90,±45 had the smallest damage area, which is due to its high stitch density. This finding is incompatible with a previous study’s results where the sample with higher stitch density suffered from a smaller damage area [8]. This specimen’s damage area is 76% and 81% smaller than that of A-U on the front and back side, respectively.

When a woven fabric is used as reinforcement, composites are strengthened along 0° and 90° directions, but no reinforcement is present along the 45° direction. In Figure 8 of A-45’s surface, it is obvious that the presence of stitch yarns in this direction has restricted the damage more sufficiently as stitch yarns inhibited fiber fracture. Moreover, when the composite is subjected to ballistic impact, shear stress is applied to the specimen, and it can grow at different angles. In composites with stitch angle of 0° and 90°, the stitch yarns are in alignment with the fibers’ directions; therefore, the stitch yarn are unable to reinforce the composite in other directions. In the composite with 45°, by contrast, the specimen can control shear stress in an additional direction because stitch yarns are reinforcing the composite at an additional angle. These are probably the reasons for a smaller damage area of sample A-45 compared to samples A-0, A-90 and A-0,90.

Another improvement of composites’ behavior by stitch yarns is on damage area of back side of specimens. When projectile impacts composite, applied stress is compressive on the composite’s front side and tensile on its back side. This tensile stress causes the outer layers to detach from the main layers, which leads to severe delamination. As can be seen in Figures 4 and 7, damage area in the back side of the A-U specimen is almost 5% larger than the front side. However, in stitched composites with different stitch angles, the damage area in the back side of the composites is smaller by an average of 9%, which indicates the importance of stitch lines to control the tensile stress and, subsequently, inhibition of the delamination.

In agreement with what was found in previous studies [7,8], it is clear from Figure 5 that residual velocity in most of the stitched specimens is slightly lower than unstitched composite, which demonstrates that stitch lines improved energy absorption (Figure 6). However, only the energy absorption of specimen A-45 was statically significant compared to A-U. This specimen had the highest energy absorption, which is 18% higher than unstitched composites. This specimen’s damage area was smaller than A-U in both the front and back sides of the composite by 71% and 73%, respectively. This small damage area and the severe disordering of the fibers in cross-section compared to that of A-U (Figure 8, A-45(C)) indicates that composite absorbed energy through fibers rupture. The reason for the better performance of A-45 compared to samples A-0,90, A-±45 and A-0,90,±45 is probably the lower stitch density because, as explained before, stitching may cause fiber damage and resin-rich regions in composites [3]. The lower void content of specimen A-45 in Table 1 supports this assumption. In regards to specimen A-0 and A-90, the reason for the better performance of A-45 might be the presence of reinforcement in the 45° direction. In samples A-0 and A-90, stitching is along the fibers’ direction of woven fabrics, and stitch lines only add reinforcement through composite’s thickness. In sample A-45, on the other hand, stitching not only reinforces composite through its thickness but also reinforces it on its surface in a direction where no reinforcement exists.

As shown in Figure 5, residual velocity in samples A-±45 and A-0,90,±45 is slightly higher than A-U; therefore, composites failed to absorb projectile’s energy. The damage area in these samples is smaller compared to A-U, so they successfully suppressed delamination. However, by comparing their cross-section pictures (Figure 8, A-U(C), A-±45(C), A-0,90,±45(C)) with that of A-U, it is evident that these composites failed to absorb energy through other mechanisms, such as fiber fracture, which resulted in lower absorbed energy compared to A-U. As mentioned earlier, delamination was higher in the unstitched composite. Moreover, the less fiber disordering of A-U than that of A-±45 (Figure 8) indicates that the main mechanism of absorbing energy in A-U was delamination. Although A-±45 suppressed delamination, other mechanisms such as fiber fracture were not adequate to absorb the projectile’s energy. For A-0,90,±45, fiber disordering is even less than A-U and A-±45, which resulted in lower energy absorption. This poor delamination strength is probably due to the higher stitch density of these samples and reduction of flexibility or destruction of reinforcement structure during the stitching process and creation of resin-rich regions [3].

Effect of stacking sequence of stitched layers

In composites with different stacking sequence of stitched layers, except for D-0, the damage area on the front side is smaller compared to A-U. However, dissimilar to composites with different stitching angles, the damage area in the back side of these composites is larger than their front side, demonstrating that composites failed to suppress delamination. It is also shown in Figure 4 that damage area in the back side of C-0 and D-0 was even larger than the A-U composite. In these composites, four layers were stitched together and placed between two layers of fabrics on top and bottom. A possible explanation is that because the two outer layers were not attached to the main four layers, stronger tensile stress was imposed on them, and the layers’ separation occurred. In the D-0 specimen, the two last layers were stitched together. Therefore, the tensile stress on the eighth layer affects the seventh layer more, and the delamination was severe. This is probably the reason for a larger damage area in D-0 in comparison with the C-0 specimen. Moreover, the damage area on the front side of D-0 is larger than A-U, which is probably because of an unbalance in the distribution of energy through the layers.

As shown in Figure 5, the residual velocity of projectile in composites with different stacking sequence is lower than the unstitched specimen, which means composites absorbed more energy compared to A-U. However, as explained before, composites could not inhibit delamination sufficiently; therefore, the main energy absorption mechanism was delamination. The severe delamination is easily noticeable in cross-section pictures of specimens, indicated by circles in Figure 8. It is also obvious from these pictures that slight fiber disordering happened in specimens compared to A-U. In addition, the surface pictures of these composites (Figure 8) show that fibers were fractured in clusters; therefore, composites could not absorb the projectile’s energy through fiber fracture. For instance, specimen D-0 had the highest absorbed energy, but the large damage area on its back side, its severe delamination (Figure 8, D-0(C)) and rupture of fibers in clusters (Figure 8, D-0(S)) demonstrate that this composite’s delamination strength under high velocity impact was poorer than the unstitched specimen.

Specimen B-0 had the best performance among composites with different stacking sequence of stitched layers. Energy absorption of this composite was around 10% higher than unstitched composite, and the damage area on its back side was about 39% smaller. Therefore, for thick composites in which stitching of all layers is not possible, it is suggested that few and equal numbers of layers be stitched together.

Conclusions

The effect of stitch angle and stacking sequence of stitched layers on the impact behavior of glass/epoxy composites were investigated. Stitched and unstitched composites were tested under high velocity impact, and their energy absorption and damage area on their front and back sides were calculated.

The main findings of the study are summarized herein:

In general, stitching reduced the residual velocity of the projectile and increased the energy absorption of composites. However, the composite with the stitch angle of 45° was the only specimen which its energy absorption was statically significant compared to unstitched and other stitched specimens.

Stitching also reduced the damaged area and inhibited delamination. The composite with the stitching angle of A-0,90,±45 had the smallest damage area; however, its energy absorption was lower than the unstitched composite. This is probably due to high stitch density, which caused more damage during stitching, reduced composites flexibility and created resin-rich regions in the composite’s structures.

Composite B-0 had the best performance among composites with different stacking sequence of stitched layers; therefore, it is suggested to few and equal numbers of layers be stitched together in situations where stitching all layers is not possible.

Delamination was higher in C-0 and D-0 compared to the unstitched specimen. This is presumably because of the separation of the two last layers, which bear the tensile stress. Delamination in D-0 is more severe than in C-0. In D-0, the last two layers are stitched together and tensile stress caused deeper delamination compared to C-0.

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) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Zeynab Behroozi

Hooshang Nosraty

Majid Tehrani

References

Mouritz

Bannister

Falzon

, et al. Review of applications for advanced three-dimensional fibre textile composites. Compos Part Appl Sci Manuf 1999; 30: 1445–1461.

Hosur

Vaidya

Ulven

, et al. Performance of stitched/unstitched woven carbon/epoxy composites under high velocity impact loading. Compos Struct 2004; 64: 455–466.

Mouritz

Leong

Herszberg

A review of the effect of stitching on the in-plane mechanical properties of fibre-reinforced polymer composites. Compos Part Appl Sci Manuf 1997; 28: 979–991.

Hosur

Karim

Jeelani

Experimental investigations on the response of stitched/unstitched woven S2-glass/SC15 epoxy composites under single and repeated low velocity impact loading. Compos Struct 2003; 61: 89–102.

Atas

Sayman

An overall view on impact response of woven fabric composite plates. Compos Struct 2008; 82: 336–345.

Tehrani Dehkordi

Nosraty

Shokrieh

, et al. The influence of hybridization on impact damage behavior and residual compression strength of intraply basalt/nylon hybrid composites. Mater Des 2013; 43: 283–290.

Zhao

Rodel

Herzberg

, et al. Stitched glass/PP composite. Part I: tensile and impact properties. Compos Part Appl Sci Manuf 2009; 40: 635–643.

Hosur

Karim

Jeelani

Studies on stitched woven S2 glass/epoxy laminates under low velocity and ballistic impact loading. J Reinf Plast Compos 2004; 23: 1313–1323.

Kang

Lee

SH.

Effect of stitching on the mechanical and impact properties of woven laminate composite. J Compos Mater 1994; 28: 1574–1587.

10.

Sadeghi

Esfandiari

Study on morphological, rheological and physico/mechanical behavior of SEBS/CaCO3 nanocomposite. Fibers Polym 2013; 14: 556–565.

11.

Mouritz

Cox

BN.

A mechanistic interpretation of the comparative in-plane mechanical properties of 3D woven, stitched and pinned composites. Compos Part Appl Sci Manuf 2010; 41: 709–728.

12.

Najafloo

Rezadoust

Latifi

Effect of through-the-thickness areal density and yarn fineness on the mechanical performance of three-dimensional carbon-phenolic composites. J Reinf Plast Compos 2016; 35: 1447–1459.

13.

Bilisik

Kaya

Tensile/shear behavior of multi-stitched/nano composites. Journal of Elec Materi 2017; 46: 3987–3994.

14.

Francesconi

Aymerich

Effect of stitching on the flexure after impact behavior of thin laminated composites. J Mech Eng Sci 2018; 232: 1374–1388.

15.

Yudhanto

Watanabe

Iwahori

, et al. Effect of stitch density on tensile properties and damage mechanisms of stitched carbon/epoxy composites. Compos Part B Eng 2013; 46: 151–165.

16.

Tan

Yoshimura

Watanabe

, et al. Effect of stitch density and stitch thread thickness on damage progression and failure characteristics of stitched composites under out-of-plane loading. Compos Sci Technol 2013; 74: 194–204.

17.

Ladani

Ravindran

, et al. Multi-scale toughening of fibre composites using carbon nanofibres and z-pins. Compos Sci Technol 2016; 131: 98–109.

18.

Francesconi

Aymerich

Effect of Z-pinning on the impact resistance of composite laminates with different layups. Compos Part Appl Sci Manuf 2018; 114: 136–148.

19.

Bilisik

Yolacan

Experimental characterization of multistitched two-dimensional (2D) woven E-glass/polyester composites under low-velocity impact load. J Compos Mater 2014; 48: 2145–2162.

20.

Erdogan

Bilisik

Compression after low velocity impact (CAI) properties of multi-stitched composites. Mech Adv Mater Struct 2018; 25: 623–636.

21.

Bilisik

Yolacan

Low-velocity impact characterization of multistitched woven E-glass/polyester nano/micro composites. Text Res J 2014; 84: 1411–1427.

22.

Bilisik

Characterization of multi-stitched woven nano composites under compression after low velocity impact (CALVI) load. Polym Compos 2018; 39: 3750–3764.

23.

Tan

Watanabe

Iwahori

X-ray radiography and micro-computed tomography examination of damage characteristics in stitched composites subjected to impact loading. Compos Part B Eng 2011; 42: 874–884.

24.

Tan

Watanabe

Iwahori

Effect of stitch density and stitch thread thickness on low-velocity impact damage of stitched composites. Compos Part Appl Sci Manuf 2010; 41: 1857–1868.

25.

Aymerich

Pani

Priolo

Eﬀect of stitching on the low-velocity impact response of [0₃/90₃]_s graphite/epoxy laminates. Compos Part Appl Sci Manuf 2007; 38: 1174–1182.

26.

ASTM D8101/D8101M 2017. Standard test method for measuring the penetration of composites materials to impact by a blunt projectile. West Conshohocken, PA: ASTM.