An experimental investigation of the yarn pull-out behavior of plain weave with leno and knitted insertions

Abstract

Yarn–yarn sliding force plays a vital role in absorbing impact energy for plain fabrics. This paper reports the methods and results of an investigation on the mechanisms that enable higher yarn pull-out force of woven fabrics with the incorporation of lenos and knits. The experimental results suggested that the insertion of leno lines on plain weave gives an approximately 20% increase in junction rupture force over the original plain construction. With knitted structures inserted, the structure-modified fabrics showed a junction rupture force up to about 15 times higher than simple plain weave. It was even found that the yarns failed rather than pulled out in multiple yarn pull-out tests. This is because knitted structures tend to become self-locked and consequently restrict yarn displacement when subjected to external loading. This investigation reports a method to increase the frictional force between the warp and weft yarns based on textile technologies. It is expected that the results obtained could provide some useful information for the engineering design of flexible ballistic protection systems.

Keywords

yarn pull-out junction rupture force leno plain weave knitted structure

It has been widely accepted that the ballistic performance of plain weave (PW) is inferior compared to unidirectional (UD) or other forms of ballistic materials.^1–7 This is because the construction of woven fabric is comparatively loose and therefore yarns tend to displace in the out-of-plane direction when being transversely impacted by projectiles. The out-of-plane displacement of a yarn refers to “pull-out”. Nilakantan et al. suggested that yarn de-crimping during the pull-out process provides extra length for a projectile to push yarns aside and open a large hole at the impact site, which is termed as “windowing.”⁸ This phenomenon is exacerbated if the fabric edges are unclamped, which is the situation in an actual body armour system. The fact that fabric penetration is, more often than not, accompanied by yarn pull-out significantly hinders high-performance fibers from exhibiting their mechanical properties, resulting in lower impact resistance of woven constructions.^8,9 If yarns could be tightly locked or gripped, yarn pull-out and windowing would be less likely to occur. The development of UD fabric is based on this very idea. The tight construction of UD material is achieved by binding all the filaments in a matrix such as a polyurethane matrix.¹⁰ As the main penetration mechanisms of UD fabrics are filament stretching and failure, the use of ultra-high-molecular-weight polyethylene filament in this construction is more feasible than aramid fibers due to its lower inter-fiber coefficient of friction and higher modulus and strength. Nevertheless, it has been reported that UD fabric is less resistant to shear force than the woven fabric, indicating that this type of construction tends to fail at an early stage when placed near the impact face of a panel under certain circumstances.^3,11

Apart from UD fabrics, efforts to design more protective woven fabrics have ever continued. Since the contribution of the friction-sliding mechanism to fabric energy absorption could not be neglected,^12,13 producing an appreciable component in the friction between the crossover-forming-yarns is believed to be beneficial. It is interesting that most of the researchers undertake their work on aramid-fiber-based PWs, and the majority of their publications focus on chemical treatments, among which shear thickening fluid (STF) has attracted considerable attention. An STF is a non-Newtonian fluid whose viscosity increases discontinuously above a critical shear rate.^14,15 This response could be attributed to the formation of particle clusters by shear force, leading to a sudden increase in fluid viscosity.¹⁶ There is no shortage of literature correlating the impact resistance of woven fabrics with STF, and the results show that the performance of the composite material is governed by many factors, such as particle size, configuration and concentration, and pull-out rate.^17–21 In addition to STF, other approaches have also been studied, and some of the most recent works are summarized as follows: Sun and Chen used a low-pressure plasma-enhanced chemical vapor deposition (PCVD) method to apply non-polymerizing reactive plasma N₂ and polymerizing plasma (CH₃)₂Cl₂Si to Kevlar woven fabric.²² As the low-temperature non-polymerizing reactive plasma treatment caused etching and ablation action, the fiber surface would get roughened,²³ leading to a higher adhesive force between fibers and yarns. Yarn pull-out tests showed that (CH₃)₂Cl₂Si treated samples rendered more yarn gripping than N₂ treated samples. Chu et al. simplified the process by using atmospheric-pressure PCVD technology with the application of (CH₃)₂Cl₂Si. It appears that chemicals deposited on the fiber surface do not yield any difference. However, samples with the most extended treatment time give the highest static and kinetic coefficients of yarn–yarn friction, which is around 83.6% and 56.7% higher than neat yarn.²⁴ Chu et al. also investigated the effect of different sized TiO₂/ZnO hydrosols. It has been found that yarns treated by sub-micro-sized hydrosol exhibit a nearly 50% increase in coefficient of friction compared to neat threads, while the rise in nano-micro-sized coated yarns is only 10%.²⁵ This effect could be attributed to the irregular coating and rougher surface of the previous sample. Hwang et al. found that the application of ZnO nanowires to fiber surfaces causes interlocking between the wire arrays, leading to a significant increase in yarn gripping.²⁶ In addition, the force-displacement curve of the yarn pull-out test exhibits a double-peak trend. The second load peak could probably be attributed to the build-up of fractured nanowires during the yarn translation process. In spite of the improved yarn pull-out performance, the treatment on neat fabric makes chemical approaches a comparatively labor intensive and time-consuming choice. Besides, the mass and flexibility of the material are biased by the chemicals, which was excluded from most of the research. For instance, there is evidence of mass increase due to the application of STF on neat woven fabrics.^17,27 It is identifiable that an increase in inter-yarn friction leads to a restriction on fiber displacement. Impeding the relative motion between fibers and yarns increases the bending rigidity and consequently decreases the flexibility of the overall fabric.²⁸

To retain fabric properties, some researchers resort to textile-based technologies. The working principle is to change yarn crimp in a plain fabric so that the gripping force is increased. The simplest method is to increase the thread density on the warp and weft directions and weave tight fabrics. Nevertheless, structures with a highly undulated yarn path suffer from low initial tensile modulus, and therefore, the original high strength of the high-performance fiber turns to a lower one. It seems inevitable that a trade-off must be made between yarn crimp and the response of fabric upon ballistic impact. Some researchers incorporated leno structures into PW so that only a small section of yarn path is undulated. Leno weaves, which are also called cross weaves, are open fabrics with warp and weft threads crossing with two adjacent warp yarns crossing over each other and wrapping around a weft yarn.²⁹ A schematic diagram of a leno weave is shown in Figure 1.

Figure 1.

Schematic diagram of a leno weave.

Fabrics made with leno weaves are mainly intended for fashion requirements. Leno structures are used by designers to decorate fabrics in combination with other patterns. The results exhibit a 30% higher pull-out force than single-phased PW, insufficient to provide a noticeable improvement in ballistic performance.^30–32 In this research, knitted structures were inserted and their performance compared with leno structures due to their more crimped yarn path. Looms were retrofitted to facilitate the manufacturing process of PW with knit insertions. The objective of this research is to improve the resistance of fabric-forming-yarns against pull-out force at the sacrifice of surface evenness. The ballistic performance and finite-element-based numerical predictions of the corresponding constructions will be included in our second paper, aiming to provide an insight into the application of flexible woven fabric in a ballistic protection system.

Fabric design

Already mentioned is the work on PW with leno line insertions. Although an increase in yarn pull-out force is observed, the value is not large enough to provide a significant improvement in the ballistic protection. Nevertheless, the results also indicate that, if yarn crimp could be increased by special weaving techniques, the potential of the woven fabrics in ballistic applications could be further explored. This leads to the combination of knitted structures with PW. Three knits, single jersey [Figure 2(b)], rib [Figure 2(c)], and interlock [Figure 2(d)] structures, were incorporated into the PW in the form of a line so that the yarn crimp of the rest of the fabric was not affected. The manufacturing process for single jersey, rib, and interlock structures was illustrated by Spencer:³³ single jersey was created by drawing the loops away from the technical back and towards the technical face side of the fabric; the formation of rib required two sets of needles operating in between each other, so that face and reverse stitches could be created in an alternate sequence; interlock was developed from rib structure and has two wales of loop shown on each side of the fabric, exactly in line with each other. In addition, two wales of aligned single jersey structure were designed to investigate the influencing of loop geometry on yarn pull-out force, which is shown in Figure 2(e); the region where the weft yarn is highly undulated is termed as the “gripping area.” Figure 2(f) shows a close-up of PW with leno structures. PW with leno insertions are collectively known as PWLs, and PW with knit insertions are known as PWKs. Table 1 lists the abbreviations of each specimen. The yarn in use was Kevlar 29 and was 1500D in linear density. Weave densities in the weft and warp directions were both 7 threads/cm. The PW had an areal density of 250 g/m². The value varies slightly and falls into the range of 250–270 g/m² for other structures. The leno structures were created using leno doups and the knitted structures were made on a retrofitted bifacial fabric sampling loom, which are shown in Figure 3. The loom is equipped with a needle board to ensure the formation of yarn loops after picking. More detailed descriptions of the machine can be found in Zhu et al.’s work.^34–36

Figure 2.

Optical micrographs of (a) plain weave; (b) plain weave with one wale of single jersey knitted structure on the warp direction; (c) plain weave with rib knitted structure on the warp direction; (d) plain weave with one wale of interlock knitted structure on the warp direction; (e) plain weave with two wales of single jersey knitted structure on the warp direction; and (f) plain weave with leno insertion.

Figure 3.

(a) Leno doup and (b) co-woven-knitted fabric sampling machine.

Table 1.

Sample descriptions

Samples	Description
PW	Plain weave
PWL1500D	Plain weave with leno insertion (leno yarn count being 1500D)
PWL3000D	Plain weave with leno insertion on the warp direction (leno yarn count being 3000D)
PWSJ	Plain weave with one wale of single jersey knitted structure on the warp direction
PWI	Plain weave with one wale of interlock knitted structure on the warp direction
PWR	Plain weave with rib knitted structure on the warp direction
PWTSJ	Plain weave with two wales of single jersey knitted structure on the warp direction

Experimental setup

Yarn pull-out test

Yarn pull-out tests were performed to verify the increase in yarn gripping in structured modified fabrics. The setup is schematically illustrated in Figure 4. Samples were cut into 6 × 12 mm, and yarn tails were left on the top and bottom sides. A slot was maintained for the yarn to be pulled out on the bottom while a bottom jaw clamped the rest of the tails. One of the advantages expected from this setup was the elimination of fabric in-plane deformation observed in other publications,^26,37–39 which better captures the response of yarns subjected to lateral forces. The upper jaw grips the selected weft yarn and moves at a constant rate of 250 mm/min.

Figure 4.

A schematic diagram of the method for the yarn pull-out test.

Yarn tensile test

It is reported that the weaving process causes various levels of strength degradation on high-performance yarns.⁴⁰ In our fabrics, the weft yarns are subjected to repetitive abrasion due to the ascending and descending movement of latch needles, and therefore, the constituent filaments are susceptible to damage. A yarn tensile test was performed to identify strength degradation according to the ASTM standard D7269-07.

Results and discussion

Single yarn pull-out on PW

Numerous studies have been undertaken to study the quasi-static response of plain fabric upon yarn pull-out.^37,41–46 As they have already provided a clear understanding of what is, undoubtedly, a most complex process, only a brief description will be given here to give a starting point to the discussion. Figure 5 shows an exemplary load-displacement curve from a single yarn pull-out test performed on Kevlar PW, which encompasses many aspects of fabric response. When a yarn is pulled, yarn tension will reach a value called the junction rupture force (JRF) or peak load force.^41–43 The gradient of the pre-JRF region is dependent on the extent of fabric in-plane deformation and yarn un-crimping. The clamping method illustrated in Figure 4 was supposed to eliminate the fabric deformation that was reported in many research works using side-clamps.^{26,37,43–45} In the real test, it was found that the withdrawn yarn causes its orthogonal warp yarns to displace towards the pull-out direction due to the frictional interaction on crossovers, which contributes to pre-un-crimping deformation of the fabric sample. Additional work is in progress to complete a modification on the side-clamp to narrow the gauge width of the sample, aiming to prevent extensive distortion of the fabric plane. The imposition of greater force beyond the JRF leads to progressive yarn sliding, and the curve becomes oscillatory, which is described as yarn “stick-slip”⁴¹ or “translation.”⁴³ The value of the JRF is a complicated function of many factors such as yarn–yarn friction, weave density, yarn count and clamping condition and pull-out rate; it also determines the amount of projectile kinetic energy dissipated by frictional interaction during a ballistic event. The stick-slip region of the load-displacement curve is found to be a “decreasing oscillation” type.³⁸ A pin-joint model, which derived from the work of Kawabata et al.,⁴⁶ is used to explain the two-region response.

Figure 5.

Load-displacement curve of single-yarn pull-out on plain weave.

Figure 6.

Crossover simulated by pin-joints (a) before deformation; (b) after deformation.⁴⁶

When a fabric is intact, the weft and warp yarns, which are represented by the pin-joints in Figure 2(a), exhibit an identical amplitude of crimp; when a loading force is exerted on the weft yarn, crimp exchange occurs. The relationship between F₁ and F₂ could be readily described according to the Capstan equation

F_{1} = F_{2} e^{α μ}

(1)

where α and μ are the angle of contact and the coefficient of friction between the warp and weft yarns, respectively, e is taken to be 2.718. If the value of F₁ on the first crossover is considered to be the ultimate yarn pull-out force, the value could be stochastically obtained from equation (1)

F_{1} = e^{μ α_{1}} e^{μ α_{2}} e^{μ α_{3}} \dots e^{μ α_{X}} F_{X}

(2)

where α₁, α₂, …… α_X is the contact angle between the weft and warp yarns at the first, second…and Xth crossover away from the top edge. Note that X refers to the penultimate crossover rather than the last crossover on the bottom, and therefore, F_X is F₂ on the corresponding crossover. F_X is dependent on the frictional force due to the compression between the last two warp yarns. Although this value is minuscule, it is extremely important in determining the ultimate pull-out force. Accordingly, the value of F₂ on the first crossover is not difficult to obtain

F_{2} = e^{μ α_{2}} e^{μ α_{3}} \dots e^{μ α_{X}} F_{X}

(3)

Zhu et al.⁴⁷ employed a shear-lag-theory-based model to simulate the decreasing trend of the stick-slip region. In their model, the PW is treated as a continuum matrix and the pull-out force is primarily dominated by the shear stress between the matrix and yarn. As a result, the resistant force against yarn pull-out is dependent on the length of yarn left in the matrix during yarn translation, leading to a linearly decreasing trend during the process. Consequently, it is inappropriate to simplify the woven construction into matrix and reinforcement materials. According to equations (2) and (3), on the first crossover, F₁ is greater than F₂ by a factor of $e^{μ α_{1}}$ . When crossovers are progressively cleared, the ultimate pull-out force drops from F₁ to F₂……F_X by a factor of $e^{μ α_{1}}$ $, e^{μ α_{2}}$ $, e^{μ α_{3}}$ …… in a discrete manner. The angle of contact at each crossover depends on the magnitude of crimp exchange, which is a function of the resultant normal force, F_c, which is a joint force of F₁ and F₂. It follows that the series of angles of contact reduce gradually along the pulled weft yarn and the decreasing trend on pull-out force is best described as “exponential-like”.

The oscillating response in the stick-slip region is more often observed on fabrics made from high-performance fibers than natural fibers, primarily in that the former tend to maintain their crimped profile during yarn pull-out. Zhu et al.⁴⁷ described the process to be a repeated transition between “in-phase” to “out-of-phase,” which is schematically illustrated in Figure 7. When yarn interlacement is transformed from “in-phase” to “out-of-phase,” the pulled yarn needs to overcome the resistance of the inclined surface of the orthogonal warp yarns in what is described as “stick” behavior, corresponding to the ascending sections of the post-peak curve; when the situation is transformed from “out-of-phase” to “in-phase,” the yarn slips from the peak to the trough, corresponding to the descending portion of the load-displacement curve. One of the limitations with their work is that they simplified the flexible nature of yarn into a rigid body, which means the yarn profile does not change during the pull-out process. In this case, the variation of the pull-out force in the stick-slip region does not follow the Capstanian law and the decreasing trend presented in their model is linear for obvious reasons.

Figure 7.

Transition between “in-phase” and “out-of-phase” during yarn pull-out.

If the pulled yarn is treated as soft matter, and the force needed to de-crimp unraveled yarn is considerably less than the force required to pull yarn from a woven fabric, then the transition from “in-phase” to “out-of-phase” is not sufficient to elicit curve oscillation. There could be other possible explanations, which are presented as follows. When the last crossover is cleared, the pull-out force drops suddenly from the JRF to the trough of the first oscillation cycle (Figure 5). In this case, yarn tension is released and the pulled yarn tends to recoil slightly back to its original crimped profile, primarily due to the normal force imposed by its orthogonal warp yarns and partially due to the transition from “out-of-phase” to “in-phase.” While the decreasing trend in the “stick-slip” region is “exponential-like” and the force drops from F₁ to F₂ by a factor of $e^{μ α_{1}}$ on the first cycle, F₂ in effect refers to the peak point second to the JRF rather than the value on the first trough. If the first slip portion in Figure 5 following the JRF corresponded to the reduction from F₁ to F₂, the pull-out force would level off for a short displacement rather than increase immediately thereafter, i.e. the slip portion of the first oscillation cycle implies a certain extent of yarn un-crimping. In summary, the motion observed during the post-JRF region on the load-displacement curve is a combination of repeated yarn crimping and de-crimping in what appears to be a slip and a stick, respectively. In addition, it can be seen in Figure 5 that the amplitude of oscillation decreases with the progressive pull-out of the weft yarn. This trend could be explained by analyzing the response of the pulled yarn upon the clearance of the first and second crossovers from the bottom edge. According to the conclusion obtained from equations (2) and (3), it is evident that the yarn tension release due to the clearance of the first crossover is greater than that due to the clearance of the second crossover, and therefore, the corresponding descending section of the former is longer than the latter. The greater yarn tension release also leads to a greater portion of the yarn path being crimped, and hence a greater amplitude of oscillation at the initial stage of yarn pull-out than the latter stage of pull-out would be expected.

Single yarn pull-out on structure modified fabrics

Comparison between PWLs

According to equation (1), yarn pull-out force is very sensitive to the coefficient of friction and the angle of contact between the warp and weft yarns. Already mentioned in the previous section is that fabrics with a modified coefficient of friction are more susceptible to property degradation. Two types of leno structure were manufactured in combination with a PW to increase the gripping force of the weft yarns while retaining fabric properties. It seems from Figure 8 that the values of the JRF for PWLs differ by only approximately 18%. Also, no significant difference in the JRF is found between PWL1500D and PWL3000D, indicating that the selection of yarn count on the leno ends yields little influence on the ultimate pull-out force. It is also observed in Figure 9 that a reduction in the amplitude of oscillation occurs in the middle of the curve. On the specimen, this point could be regarded as the location where the leno structure is cleared by the pulled yarn.

Figure 8.

Junction rupture force (JRF) of yarn pull-out tests for plain weave (PW) and structure modified fabrics.

Figure 9.

Load-displacement curves for single yarn pull-out on plain weave (PW) and plain weave with leno insertions (PWLs).

Comparison between PWSJ and PWI

Unlike leno structures, the gripping area in PWKs is formed by loop intermeshing. One wale of single jersey, interlock structures, rib, and two wales of the single jersey were investigated in comparison with PW. Due to structural similarities, we will analyze PWSJ and PWI to provide a start point. It can be seen in Figure 8 that the insertion of these two types of structure gives an increase in the JRF by a factor of more than two when compared with PW for single yarn pull-out. It also appears that the knitted structures are more effective in gripping weft yarns than leno structures. An interesting phenomenon is observed in Figure 10: the progressive motion of the pulled yarn leads to noticeable in-plane deformation on the “gripping area” towards the pulling direction on PWSJ and PWI, which is less noticeable on PW and PWL. This sort of deformation could be explained by looking into the close-up of a single jersey loop. In Figure 11, the intermeshing is formed by entangling the heads and feet on consecutive loops, creating four pairs of contact. When the loop-forming yarn is stretched, the normal components of the interactive force between the legs and feet are increased, and hence higher frictional force is generated. Compared with one pair of contact between the warp and weft yarns in the leno structures, the greater increase in pull-out force for knitted insertions could be readily explicable. If four pairs of leno yarns are inserted into PW to simulate the four pairs of contact in the single jersey structure, the resultant JRF would probably still be less than that of PWSJ or PWI. That is, there could be other gripping mechanisms taking effect along with the contacting pairs. It is probable that yarn stretching makes the loops become more deformed and tightened up, and consequently, the top and bottom contact pairs compress against each other more intensely. This greatly increases the interactive force between the head and feet, enabling the yarn to be more able to resist pull-out force. When more than one loop gets involved in yarn gripping, the in-plane deformation would ultimately evolve into yarn self-locking, e.g. multiple yarn pull-out on PWR or PWTSJ. In this case, yarn translation would not occur and the pull-out force would increase up to a value, after which the yarn would break.

Figure 10.

Single yarn pulled-out on (a) PW; (b) PWSJ; (c) PWR; (d) PWI; (e) PWTSJ; and (f) PWL.

Figure 11.

A schematic diagram of single jersey structure.⁴⁸

It can be seen in Figure 12 that the pre-JRF regions of the load-displacement curves exhibit a lower gradient for PWSJ and PWI compared to PW. This could be explained as follows: when the knitted line is strained, its comparatively loose structure would accommodate more in-plane deformation than PW and PWL. When this happens, the loop distorts along the loading direction, and the response of fabric to yarn pull-out is delayed. It appears that the weft yarn is more tightly gripped by the “gripping area” at the expense of quick response to the pull-out force due to greater yarn crimp than PWLs. Considering the dual-influences, it is difficult to determine the effect of knitted structures on the bulletproof performance of the whole fabric without ballistic testing. In the stick-slip region, both of the fabrics exhibit an irregular decreasing trend of oscillation, indicating that the progressive motion of crossover clearance is dominated by the complicated combination of yarn deformations, particularly on the knitted section. The load-displacement curves experience a sharp decrease in load force, the magnitude of which is more dramatic than the oscillation retardation region in PWLs. As such, this phenomenon implies the clearance of single jersey and interlock structures by the weft pulled yarn.

Figure 12.

Load-displacement curves for single yarn pull-out on plain weave (PW) and plain weave with knit insertions (PWKs).

The primary difference between PWSJ and PWI is the loop formation process, i.e. the loops are intermeshed on every other yarn for PWI and on consecutive yarns for PWSJ, implying that the loops of PWI have longer legs than those of the PWSJ. The results from the single yarn pull-out test give similar values for the JRF as shown in Figure 8, which means that length of the loop legs plays a less important role in determining the amount of loading force required to withdraw a yarn.

Comparison between PWR and PWTSJ

PWTSJ and PWR are put into comparison due to the fact that they are both formed by two wales of loops but in different dimensions. PWR gives an inclined appearance because one loop wale tends to move over and in front of the adjacent loop wale, forming a “zig-zag” geometry [Figure 2(c)] and PWTSJ has two wales of single jersey structures aligned in parallel [Figure 2(e)]. For PWR, the width of insertions is smaller than that of the two wales of the single jersey. This could be beneficial in a textile-based bulletproof system, primarily in that the probability of a projectile impacting the very line of the knitted structure is decreased. However, if the impact point happens to be on, or in the vicinity of, the insertion, the penetration resistance is yet to be fully understood. By comparing PWSJ and PWTSJ, it is reasonable to accept that an additional wale of single jersey doubles the peak load force and the distortion is more drastic than PWSJ and PWI (Figure 10). Counterintuitively, Figures 8 and 12 show that the rib structure exhibits a drastic increase in JRF by a factor of about 15 over PW and a factor of about six over PWSJ. It appears that additional wale of reversed single jersey structure on rib contributes to loop tightening remarkably. The significant difference between PWTSJ and PWR might be attributed to the restricted mobility and compactness of loops in the corresponding structures. It can be seen in Figure 2(e) that the construction of PWTSJ is comparatively looser than PWR, indicating that the loops are more stretchable when subjected to loading forces. More important is the role loop geometry plays in loop deformation. When a weft yarn is pulled on PWTSJ, the parallel aligned loops become stretched up consecutively and the pairs of contact compress against each other to increase the interaction force, a process which resembles that of PWI and PWSJ. When the loading force is applied on the inclined loops of PWR, its unique “zig-zag” profile produces a resistant force against yarn pull-out. Coupled with the compact nature of the superimposed loop wales, the JRF is significant. The distortion of the gripping area is also observed on PWTSJ and PWLs in Figure 10(e) and (f). This gives rise to a low-modulus trend prior to the JRF on the load-displacement curve. Due to the complex interactions between the pulled weft yarn and its orthogonal warp yarns, the post-peak regions of PWR and PWTSJ exhibit an “irregular decreasing oscillation” type, which are presented in Figure 8.

Multiple yarn pull-out on structure modified fabrics

Comparison between PW and PWLs

During a ballistic event, the load imposed by a projectile is often distributed on more than one yarn, causing multiple yarns to be pulled out. It has also been found that yarns are more likely to be withdrawn near the impact face of a panel, and the number of displaced yarns reduces with an increase of the number of fabric layers.¹² Investigating the quasi-static response of a fabric subjected to multiple yarn pull-out might provide an indirect prediction of the energy absorption capability of the corresponding woven construction. Figure 13 shows the force-displacement curves of a multiple yarn pull-out process on PW. Similar trends could be observed as for single yarn pull-out, all of which consist of un-crimping and stick-slip zones. The values of the JRF for double, triple, and quadruple yarn pull-out are around 260, 389 and 489% that of single yarn pull-out, which is not linear with the number of yarn loaded. This is primarily due to the fact that crimp exchange between the warp and weft yarns becomes more difficult to achieve when more than one yarn are withdrawn simultaneously, and hence the upper jaw needs to overcome more frictional force from crossovers.

Figure 13.

Load-displacement curves for multiple yarn pull-out on plain weave (PW).

The JRF of multiple yarn pull-out tests for PWLs are revealed in Figure 8. It seems that the influence of yarn linear density is more identifiable for PWLs in multiple yarn pull-out tests. When four yarns are pulled simultaneously, the JRF for PWL3000 is 41.3 and 65.4% higher than PW and PWL1500, respectively. If the total force is divided by the number of yarns pulled, it is found that the force sustained by each individual yarn is 70.6 (quadruple yarn pull-out test), 74.9 (triple yarn pull-out test) and 42.4% (double yarn pull-out test) higher than the JRF observed in the single-yarn pull-out test. The results indicate that, if high-count yarns are used for leno formation, the area of contact pairs is increased, and therefore the weft yarns are more tightly gripped.

Comparison between PW and PWKs

It could also be seen in Figure 8 that the effect of leno insertions is minimal when compared to knitted structures. The values of the JRF for PWSJ and PWI increase significantly in triple and quadruple yarn pull-out tests, giving a similar value of around 43 and 115 N, respectively. The JRFs are higher than those of PW by factors of approximately five and 11. For PWTSJ and PWR, yarns are observed to be stretched to break, and the force values presented in Figure 8 are actually associated with yarn failure rather than yarn pull-out. By comparing the corresponding fabric deformation and load-displacement curves for the quadruple yarn pull-out test, it can be concluded from Figures 14 and 15 that the displacement prior to the value of yarn failure is primarily caused by the large deformation of the knitted line, which also contributes the significant increase in the JRF. Also in the post-peak regions of the load-displacement curves, the progressive crossover clearance does not occur under this condition, indicating the absence of the “decreasing oscillation” type trend in the curves.

Figure 14.

Quadruple yarn pulled-out on (a) PW; (b) PWSJ; (c) PWR; (d) PWI; (e) PWTSJ; and (f) PWL.

Figure 15.

Load-displacement curves for quadruple yarn pull-out on plain weave (PW), plain weave with leno insertions (PWLs), and plain weave with knit insertions (PWKs).

It is conjectured that loop tightening, and the resulting self-locking effect, is the major contributing factor for yarn failure in PWR and PWTSJ: when three or four yarns are pulled simultaneously on PWI or PWSJ, the top and bottom yarns would stretch the middle yarns on the vertical direction, increasing the compression force on the contact pairs of the middle loops. That is, the loops in between the top and bottom ones get strained, and yarn mobility is more restricted, causing yarns in the middle to bear most of the total force. An additional wale of single jersey structure intensifies this type of loop tightening, and the middle yarns is ultimately stretched to failure by the upper jaw. This raises a question as to whether the knitting process alters the yarn properties and whether consequently the weft yarns are more prone to failure upon stretching. Yarn tensile tests were undertaken to study the strength of weft yarn taken from PW, PWSJ, PWI, and PWR. It can be seen from Figure 16 that the knitting process has little influence on yarn strength, at least on tensile strength.

Figure 16.

Tensile strength of weft yarns extracted from PW, PWSJ, PWI, and PWR.

It is also interesting to notice that the strength value of a single yarn is around 38% higher than the failure value of PWR and PWTSJ from the double yarn pull-out test. Since the loading force during yarn pull-out is sustained by two yarns simultaneously, the force distributed on each yarn is definitely lower than the breaking strength of a single yarn. This implies that yarn fails at a value much lower than its original strength. This could probably be attributed to the fact that the load path is not along the filament path in a looped yarn. A simple experiment was performed to elucidate this phenomenon. A yarn sample was wound on a metal rod when being stretched on a tensile test machine. Figure 17 reveals the variation of the rod angle. It can be seen in Figure 18 that yarn strength shows a smooth decreasing trend with the rod angle. It is probable that, when the yarn was wrapped around the rod surface, the pure tensile failure mode turned to a combination of a shear and tensile one, which was responsible for reduced strength value. Nevertheless, this raises another important question as to whether this phenomenon might bias the ballistic protection of a fabric. For most cases, yarns in a woven-fabric-based anti-ballistic system tend to be pulled out or pushed aside rather than damaged. It is speculated that, if yarn mobility could be reduced at the expense of acceptable yarn strength reduction, an improvement in energy absorption is theoretically expected. However, this hypothesis needs experimental verification.

Figure 17.

A schematic diagram of yarn tensile test with rod angle of (a) 0°; (b) 90°; (c) 180°; (d) 270°.

Figure 18.

Comparison of yarn strength at different rod angles.

If the results given by the single yarn pull-out test suggest the possibility of increasing yarn gripping force through structure modification, the data obtained from multiple yarn tests, then, indicate great potential for ballistic protection improvement over conventional plain or even UD fabrics. This could be explained by two reasons. For one thing, our previous work showed that PWs exhibit better energy absorption capacity than UD fabric.³ This conclusion, however, is obtained under the condition that the fabric edges are firmly constrained. In other words, yarn pull-out is eliminated during the laboratory test, and hence, the possibility of yarn windowing is reduced. Nevertheless, yarn pull-out and windowing could not be avoided for a woven fabric in a soft body armour system as the fabric edges are unclamped in practical situations. More often than not, the woven-fabric-based soft body armour fails to take best advantage of the mechanical properties of the high-performance fibers, primarily due to yarn pull-out. If, by insertion of knitted structures, one can resist the yarns being withdrawn, an improvement in ballistic performance might be expected. On the other hand, it has been reported that, for a 9 mm full metal jacketed projectile used for NIJ Standard 0101.04 Level IIA and II, the projectile is subjected to mushroom-like deformation during a ballistic event,⁴⁹ implying that the number of yarns getting into contact with the projectile head will increase with progressive penetration of the fabric panel. When multiple yarns are loaded transversely by the projectile on structure-modified plain fabrics, the knitted structures, especially the rib structure, will lock the weft yarns in the looped area, causing fiber failure induced penetration rather than yarn pull-out induced penetration. Despite the above analysis, one must resort to actual tests to confirm all the hypotheses so that the design could be put to practical applications.

Conclusions

The focus of this research has been to investigate the influence of different insertions on the response of PW subjected to weft yarn pull-out. For plain structures, it was found from the single yarn pull-out test that the value of the JRF and the decreasing oscillating trend of the stick-slip region are associated with crimp exchange and the contact angle between the weft and warp yarns. When leno structures are incorporated, the samples give an approximately 20% increase in the JRF over PW; no significant difference is found between PWL1500D and PWL3000D. As the number of pulled yarns increases, the gripping effect imposed by the leno structure becomes more noticeable on the weft yarns. It was determined that the JRF for PWL3000 is 41.3 and 65.4% higher than PW and PWL1500 in quadruple yarn pull-out tests, respectively. In addition, the forces sustained on each individual yarn are 70.6 (quadruple yarn pull-out test), 74.9 (triple yarn pull-out test) and 42.4% (double yarn pull-out test) higher than the JRF observed in the single-yarn pull-out test. In terms of plain weave with knitted insertions, the fabrics would accommodate more in-plain deformation before allowing the weft yarn to be withdrawn. That is, the knitted gripping line provides significant resistance for the weft yarns to be pulled out. For plain fabric with one wale of single jersey structures (PWI and PWSJ), the value of the JRFs are more than those of the PW and PWLs by a factor of approximately two for single yarn pull-out and by a factor of approximately 11 for the quadruple yarn pull-out test; for plain fabric with two wales of single jersey structure in parallel (PWR) and in “zig-zag” geometry (PWTSJ), the JRFs are around 15 and 5.5 times higher than that of PW in the single yarn pull-out test. The weft yarns are even locked, and consequently stretched to fail in multiple yarn pull-out tests in these fabrics, which could be explained by the squeezing of yarn loops and the compression effect of the loop-forming contact pairs. It can be concluded from this investigation that the ability of a plain fabric to resist yarn pull-out could be greatly improved by inserting knitted structures, indicating the possibility of further exploring the potential of textile-based technologies for ballistic protection. If the impact resistance and total mass of woven fabrics can compete with those of UD structures, the more flexible properties would make using woven fabrics a more comfortable choice for body armour wearers. Additional work is underway to perform ballistic tests on structure modified plain weaving, aiming to verify their ballistic properties and provide alternatives for the engineering design of soft body armour.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is funded by National Natural Science Foundation of China (Grant Number: 11502179); Ministry of Education, Hubei province (Grant Number: D20171602).

References

Yang

Chen

. Investigation on energy absorption efficiency of each layer in ballistic armour panel for applications in hybrid design. Compos Struct 2017; 164: 1–9.

Yang

Chen

. Investigation of failure modes and influence on ballistic performance of Ultra-High Molecular Weight Polyethylene (UHMWPE) uni-directional laminate for hybrid design. Compos Struct 2017; 174: 233–243.

Chen

Zhou

Wells

. Numerical and experimental investigations into ballistic performance of hybrid fabric panels. Composites, Part B 2014; 58: 35–42.

Lee

Song

Ward

. Failure of Spectra® polyethylene fiber-reinforced composites under ballistic impact loading. J Compos Mater 1994; 28(13): 1202–1226.

Jacobs

MJN

Van Dingenen

JLJ

. Ballistic protection mechanisms in personal armour. J Mater Sci 2001; 36(13): 3137–3142.

Karahan

. Comparison of ballistic performance and energy absorption capabilities of woven and unidirectional aramid fabrics. Text Res J 2008; 78(8): 718–730.

Tan

VBC

Lim

Cheong

. Perforation of high-strength fabric by projectiles of different geometry. Int J Impact Eng 2003; 28(2): 207–222.

Nilakantan

Merrill

Keefe

, et al. Experimental investigation of the role of frictional yarn pull-out and windowing on the probabilistic impact response of kevlar fabrics. Composites, Part B 2015; 68: 215–229.

Wang

Miao

Huang

, et al. Effect of the inter-fiber friction on fiber damage propagation and ballistic limit of 2-D woven fabrics under a fully confined boundary condition. Int J Impact Eng 2016; 97: 66–78.

10.

O'Masta

Crayton

Deshpande

, et al. Mechanisms of penetration in polyethylene reinforced cross-ply laminates. Int J Impact Eng 2015; 86: 249–264.

11.

Chen

Zhu

Wells

. An analytical model for ballistic impact on textile based body armour. Composites, Part B 2013; 45(1): 1508–1514.

12.

Kirkwood

Lee

, et al. Yarn pull-out as a mechanism for dissipating ballistic impact energy in Kevlar® KM-2 fabric: Part II: Predicting ballistic performance. Text Res J 2004; 74(11): 939–948.

13.

Briscoe

Motamedi

. The ballistic impact characteristics of aramid fabrics: the influence of interface friction. Wear 1992; 158(1–2): 229–247.

14.

Hoffman

. Explanations for the cause of shear thickening in concentrated colloidal suspensions. J Rheol (N Y N Y) 1998; 42(1): 111–123.

15.

Barnes

. Shear-thickening (“Dilatancy”) in suspensions of nonaggregating solid particles dispersed in Newtonian liquids. J Rheol (N Y N Y) 1989; 33(2): 329–366.

16.

Brady

Bossis

. The rheology of concentrated suspensions of spheres in simple shear flow by numerical simulation. J Fluid Mech 1985; 155: 105–129.

17.

Tan

VBC

Tay

Teo

. Strengthening fabric armour with silica colloidal suspensions. Int J Solids Struct 2005; 42(5–6): 1561–1576.

18.

Lee

B-W

Kim

I-J

Kim

C-G

. The influence of the particle size of silica on the ballistic performance of fabrics impregnated with silica colloidal suspension. J Compos Mater 2009; 43(23): 2679–2698.

19.

Hasanzadeh

Mottaghitalab

Babaei

, et al. The influence of carbon nanotubes on quasi-static puncture resistance and yarn pull-out behavior of shear-thickening fluids (STFs) impregnated woven fabrics. Composites, Part A 2016; 88: 263–271.

20.

Baharvandi

Khaksari

Kordani

, et al. Analyzing the quasi-static puncture resistance performance of shear thickening fluid enhanced p-aramid composite. Fibers Polym 2014; 15(10): 2193–2200.

21.

Feng

Wang

, et al. Effects of different silica particles on quasi-static stab resistant properties of fabrics impregnated with shear thickening fluids. Mater Des 2014; 64: 456–461.

22.

Sun

Chen

. Plasma modification of Kevlar fabrics for ballistic applications. Text Res J 2012; 82(18): 1928–1934.

23.

Sun

Stylios

. Fabric surface properties affected by low temperature plasma treatment. J Mater Process Technol 2006; 173(2): 172–177.

24.

Chu

Chen

Sheel

, et al. Surface modification of aramid fibers by atmospheric pressure plasma-enhanced vapor deposition. Text Res J 2014; 84(12): 1288–1297.

25.

Chu

Chen

Wang

, et al. An investigation on sol–gel treatment to aramid yarn to increase inter-yarn friction. Appl Surf Sci 2014; 320: 710–717.

26.

Hwang

H-S

Malakooti

Patterson

, et al. Increased interyarn friction through ZnO nanowire arrays grown on aramid fabric. Compos Sci Technol 2015; 107: 75–81.

27.

Lee

Wetzel

Wagner

. The ballistic impact characteristics of Kevlar® woven fabrics impregnated with a colloidal shear thickening fluid. J Mater Sci 2003; 38(13): 2825–2833.

28.

McNeil

Standard

. Increased bending rigidity of wool fabric imparted by hybrid organic-inorganic sol–gel coatings. Text Res J 2017; 87(5): 607–616.

29.

Ahmed AY. Physical properties of leno fabrics. PhD Thesis, Victoria University of Manchester, UK, 1972.

30.

Zhou

Chen

Wells

. Influence of yarn gripping on the ballistic performance of woven fabrics from ultra-high molecular weight polyethylene fibre. Composites, Part B 2014; 62: 198–204.

31.

Sun

Chen

Mrango

. Investigating ballistic impact on fabric targets with gripping yarns. Fibers Polym 2013; 14(7): 1184–1189.

32.

Sun

Chen

Wells

. Engineering and analysis of gripping fabrics for improved ballistic performance. J Compos Mater 2014; 48(11): 1355–1364.

33.

Spencer

. Knitting Technology – A Comprehensive Handbook and Practical Guide, 3rd ed. Cambridge: Woodhead Publishing, 2001, pp. 60.

34.

Zhu

Naebe

Blanchonette

, et al. Moisture transfer properties of bifacial fabrics. Text Res J 2016; 87(9): 1096–1106.

35.

Zhu

Naebe

Blanchonette

, et al. Heat transfer properties of bifacial fabrics. Text Res J 2016; 87(9): 2307–2313.

36.

Zhu

Naebe

Blanchonette

, et al. Mechanical properties of bifacial fabrics. Text Res J 2017; 88(12): 1335–1344.

37.

Bilisik

Yildirim

. Weft directional stick–slip force of yarn pull-out in para–aramid fabric for ballistics. J Thermoplast Compos Mater 2014; 27: 1167–1191.

38.

Zhu D, Soranakom C, Mobasher B and Rajan SD. Experimental study and modelling of single yarn pull -out behavior of kevlar 49 fabric. Compos Part A-Appl S 2011; 42: 868–879.

39.

Ravandi

SAH

Toriumi

. Spectral analysis of the yarn-pullout force from plain-weave fabric. J Text Inst 1996; 87(3): 522–531.

40.

Nilakantan

Obaid

Keefe

, et al. Experimental evaluation and statistical characterization of the strength and strain energy density distribution of Kevlar KM2 yarns: exploring length-scale and weaving effects. J Compos Mater 2011; 45(17): 1749–1769.

41.

Sebastian

SARD

Bailey

Briscoe

, et al. Extensions, displacements and forces associated with pulling a single yarn from a fabric. J Phys D 1987; 20(1): 130.

42.

Sebastian

SARD

Bailey

Briscoe

, et al. Effect of a softening agent on yarn pull-out force of a plain weave fabric. Text Res J 1986; 56(10): 604–611.

43.

Kirkwood

Lee

, et al. Yarn pull-out as a mechanism for dissipating ballistic impact energy in Kevlar® KM-2 fabric: Part I: Quasi-static characterization of yarn pull-out. Text Res J 2004; 74(10): 920–928.

44.

Bilisik

. Properties of yarn pull-out in para-aramid fabric structure and analysis by statistical model. Composites, Part A 2011; 42(12): 1930–1942.

45.

Bilisik

Yolacan

. Single and multiple yarn pull-out on E-glass woven fabric structures. Text Res J 2011; 81(19): 2043–2055.

46.

Kawabata

Niwa

Kawai

. 3—the finite-deformation theory of plain-weave fabrics Part I: The biaxial-deformation theory. J Text Inst 1973; 64(1): 21–46.

47.

Zhu

Soranakom

Mobasher

, et al. Experimental study and modelling of single yarn pull-out behavior of Kevlar® 49 fabric. Composites, Part A 2011; 42(7): 868–879.

48.

Majumdar

. Principles of woven fabric manufacturing, CRC Press, 2016.

49.

Bilisik

Korkmaz

. Multilayered and multidirectionally-stitched aramid woven fabric structures: Experimental characterization of ballistic performance by considering the yarn pull-out test. Text Res J 2010; 80(16): 1697–1720.