Developing Superperforming Composites by Harnessing Fiber Orientation and Processing Technology

Abstract

The current research on the development and performance evaluation of unique composites focuses on an unexplored combination of fibers of PBO (polybenzoxazole Zylon) as reinforcement and PEEK (polyetheretherketone) as a matrix. Their fibers were braided in equal ratios and then compression-molded to develop composites by manipulating all long fibers in one direction (unidirectional, a UD composite) and for the other one half in one direction and the remaining half in the perpendicular direction (bidirectional, a BD composite). The performance was evaluated by hardness (micro- and scratch) studies followed by tensile and impact strength. Furthermore, the lap shear strength of the adhesive joints developed with these fibers in various orientations was also analyzed. The performance properties were compared with those of neat PEEK. It was observed that an exceptional increase compared to neat PEEK in impact strength of 10,000% and tensile strength of 662% was achieved in the case of the UD composite. Scanning electron microscopy (SEM) and microcomputed X-ray tomography were used on failed specimens to analyze the reasons for failure.

1. Introduction

High-performance polymer composites, especially long fiber-reinforced unidirectional (UD) or fabric-reinforced bidirectional (BD), are well known for their high specific strength, modulus, toughness, and ability to withstand high temperatures.¹⁻⁹ However, their processing becomes challenging, primarily when the base matrices are polyaryletherketone (PAEK), polyetheretherketone (PEEK), etc. The solution impregnation technique to make prepregs is not applicable for such polymers, since they do not have appropriate solvents. The very few alternatives, such as film stacking and powder sprinkling, have major issues of poor wetting of filaments in the case of UD composites, with the additional problem of poor wetting at crossover points in the case of BD composites. Hot-melt impregnation leads to resin oxidation problems. The commingling/braiding technique can be used to overcome this issue. It involves two or more yarns of thermoplastic-matrix fibers (such as polyetheretherketone (PEEK), ultrahigh-molecular weight polyethylene (UHMWPE), nylon, etc.), which serve as a matrix after melting and reinforcement fibers (carbon, glass, aramid, etc.) mechanically intertwined so that the load is evenly distributed throughout the structure. The matrix and reinforcement’s multifilament yarns are dispersed at the fiber level in the commingled form. Depending on the proposed application of the composite, it is crucial to focus on how the reinforcement fibers are arranged—unidirectional, bidirectional, or randomly oriented (coventivecomposites, 2022. https://coventivecomposites.com/explainers/choosing-the-right-reinforcement-format-for-composites/).¹⁰ However, for this process to be adopted, it is necessary to have fibers of the matrix available, which is not always possible.

Ye et al.¹¹ studied a commingled yarn system of carbon fibers (CF) 60 wt % with PEEK fibers to investigate the mechanisms of consolidation and impregnations by establishing a theoretical model. An optimum processing window was suggested based on the minimum void content level achieved. Tewari et al.¹² studied the solid particle erosion wear behavior of CF-PEEK composites developed by the commingling technique (carbon fibers 65% by vol) for different fiber orientations (0, 45, 90°) and impingement angles (15–90°) and impact velocities. PEEK as a matrix and carbon fibers as reinforcement are the most favored combinations in the literature for developing such composites using the commingling technique. Poly(p-phenylene benzobisoxazole) (PBO), under the commercial name Zylon, is an emerging polymeric fiber produced by the dry-jet wet-spinning process. Like aramid, it possesses an exceptional strength-to-weight ratio and dimensional stability, excellent heat resistance (100 °C higher decomposition temperature than that of the p-aramid fiber), low flammability, and excellent chemical resistance to organic solvents. The typical applications of Zylon-based composites range from protective clothing, cable cover materials for welding machines and sports equipment, base fibers for conductive textiles coated with nickel, copper, gold, or silver, and suspension lines on parachutes (Fiberbrokers,¹³ Wikipedia,¹⁴ Said et al.¹⁵). In the case of composites, it is still in the exploration stage, and little is reported.¹⁶⁻²³

Although it is well accepted that the orientation of fibers affects the performance of composites significantly, hardly anything is reported on the comparative performance aspects of UD and BD composites using biobased thermoset resin and regenerated cellulose fibers.²⁴ Since wt % of fibers in two composites was not identical, performance was not comparable in the true sense. It was thought to be important to examine the difference in the performance of two composites (UD and BD) by keeping the exact contents of the same fibers in the same matrix. Hence, the UD composite was developed using PEEK and Zylon-commingled fibers. Performance was compared with the earlier work on a BD composite, which had the same amount of fibers but 50% in one direction and 50% in a perpendicular direction.²⁵ In this work, a UD composite using a PEEK matrix in the fibrous form (50 wt %) and Zylon fibers as a reinforcement (50 wt %) was developed, and details of the performance are presented in the subsequent sections. The addition of such high-performance fibers to PEEK can improve the performance of applications such as bush bearings, fasteners, gears, cams, and structural elements.

2. Experimental Section

2.1. Materials

PEEK fibers (ZYEX550/30) were procured from APEX Comnet Ltd., India, and Zylon HM/poly(p-phenylenebenzo-bis-oxazole) (PBO) multifilaments were supplied by Toyobo Co., Ltd., Japan. The chemical structures of Zylon multifilaments and specifications are shown in Table 1. Injection-molded PEEK specimens (Z₀) were procured to have the same grade as the fibers as the reference sample. The process of braiding the fibers is discussed in our published paper.²⁵

Table 1. Properties of Zylon Fibers (Toyobo 2022. www.toyoboglobal.com)²⁶.

2.2. Methodology

2.2.1. Fabrication of Composites

The two composites contained PEEK (50 wt %) as a matrix and Zylon (50 wt %) as a reinforcement. As shown in Figure 1a, in Z_50UD (unidirectional composite with unidirectionally oriented PBO fibers), all fibers were in one direction, while in Z_50BD (bidirectional composite with a braided fabric of PEEK and PBO), half the fibers were in one direction with half in a perpendicular direction.

Figure 1.

(a) Schematic of the fiber orientation in UD and BD composites and (b) schematic of the development process of the UD composite.

The performance of these composites was compared with that of neat PEEK (Z₀). The entire process for developing a unidirectional composite is shown in Figure 1b. The details of developing the bidirectional composite (Z_50BD) are described elsewhere.²⁵ The braided yarns were wound parallel around a metallic plate of dimensions (25.5 cm × 19.5 cm) with 70 turns in 8 layers. The entire assembly was kept inside the mold cavity. The composites were developed by compression molding. The whole assembly was heated to 380 °C at a rate of 2 °C/min under 8 MPa pressure. Five intermittent breathings were used to remove moisture, followed by keeping under pressure and temperature for 20 min. The heating was then turned off, and the composite was allowed to cool at room temperature under pressure. The composite’s final thickness (Z_50UD) was 5 mm (approximately).

2.3. Characterization of Composites

The composites were characterized for physical, thermal, and mechanical properties (hardness, tensile, and impact), lap shear strength of adhesive joints, hardness, X-ray diffraction (XRD) analysis, X-ray microcomputed tomography, etc., as per the methods discussed in subsequent sections.

2.3.1. Physical Properties (Density and Void content)

The practical density of the composites was determined by Archimedes’s principle on a Mettler Toledo weighing balance ME155DU by the following equation

1

where ρc is the density of the composite, ρliquid is the density of liquid used as medium (water), W_air is the weight of the sample in air, and W_liquid is the weight of the sample immersed in water. The composites’ theoretical density was computed using the principle of the rule of mixtures. The void content of the composites was calculated as the difference between the theoretical and practical densities of the composites.

2.3.2. X-ray Computed Tomography

The interior three-dimensional (3D) structure of a composite, such as fiber orientations, voids, diameters, and so on, can be examined using microcomputed X-ray tomography. The experiments were carried out with a Carl Zeiss Versa 510 with an X-ray energy of 80 kV. Each projection was exposed to X-rays for 2 s using 4× optics. ORS Dragonfly software was used to perform segmentation, pore analysis, pore volume distribution, and 3D reconstruction.

2.3.3. Thermogravimetric Analysis (TGA)

The thermal stability of the composites was studied using a Linseis 1000 PT simultaneous thermal analyzer (STA) in an air atmosphere. The temperature range was from room temperature (RT) to 900 °C, with a 10 °C/min ramp rate.

2.3.4. X-ray Diffraction (XRD)

X-ray diffraction studies were carried out on a Philips XPERT-PRO instrument. The wavelength used to obtain a diffraction pattern was 1.5418 Å (Cu Kα) for a 2θ range varying from 10–80° with a scan rate of 2°/min.

2.3.5. Mechanical Properties

Mechanical testing specimens were cut with minimal kerf loss from the fabricated composite sheet utilizing an Omax Waterjet cutter, Maxiem, 1530. For each property, five samples were evaluated and the average value was taken into account.

2.3.5.1. Microhardness of Composites

The scratch hardness of the developed composites was determined using the ASTM standard G 171 on a tribometer (CETR UMT-3MT (Bruker)). A load of 30 N was applied to scratch the surface of the composite for a scratch of 5 mm length at 0.002 m/s using a diamond hemispherical tip indenter with an apex angle of 120°. The scratch on the surfaces of a composite was studied and measured under an optical microscope and was determined by eq 2

2

where P is the load applied in N and w is the scratch width in mm. Vickers hardness was calculated by probing the sample using a square-based pyramidal diamond indenter with a face angle of 136° as per ASTM E384 under an applied load of 30 N. The Vickers hardness was calculated using the following equation

3

where F is the force applied and d is the mean diagonal diameter of the impression.

2.3.5.2. Tensile Properties

Tensile strength, modulus, and elongation at break were evaluated as per the ASTM D3039 with 2 mm/min crosshead speed. The tests were done on an Instron 5882 UTM (Universal Testing Machine) at (23 ± 2) °C and (50 ± 5) % relative humidity.

2.3.5.3. Flexural Properties

Flexural strength and modulus were calculated as per the ASTM D790 standard on an Instron 5582 UTM with a three-point bending configuration. The span length was kept at 64 mm, and the crosshead speed used was 2 mm/min.

2.3.5.4. Izod Impact Strength

The impact strength of the composites was evaluated on an Izod impact tester (Model: IT 504 plastic impact, Tinius, Olsen) as per the ASTM D256. The composites with a notch of 3.2 mm (45°) were struck by a pendulum with an energy of 25.4 J. Sample dimensions were maintained as per the standard.

2.3.5.5. Lap Shear Strength

Adhesive joints for lap shear strength were developed according to the ASTM D 1002 standards. Stainless-steel coupons were selected as adherents and were degreased with acetone. The piece of braided fibers/fabric was placed carefully between the coupons’ overlapping section (15 mm × 15 mm) in the required orientation, followed by the mold closure and placement in the compression-molding machine. The molding process was further carried out by heating the unit to 420 °C. After the temperature was attained, five intermittent breathings were provided, and a final pressure of 7.35 MPa was applied for 10 min. Cooling was done under pressure under ambient conditions. Lap shear testing of the developed joints was done as per the ASTM standard D1002 on a Shimadzu UTM machine with a crosshead speed of 1.3 mm/min and a gauge length of 80 mm. Five specimens of each sample were tested to ensure repeatability.

2.3.6. Fractographic Analysis by Scanning Electron Microscopy (SEM)

The fractured samples were analyzed in-depth for the mechanisms involved by a Zeiss EVO MA10 scanning electron microscope. The samples were gold-coated before analysis by a Cressington sputter coater 108 under an acceleration voltage of 20 kV and a high vacuum to make them conductive.

3. Results and Discussion

3.1. Physical Properties

The composites’ practical and theoretical densities and void content (%) are shown in Figure 2.

Figure 2.

Practical and theoretical density and void content (%) for the composites

The density of the composites improved by adding PBO fibers in PEEK due to the higher density of PBO fibers (density of PEEK is 1.32 g/cm³ and that of PBO is 1.57 g/cm³). However, a difference in density improvement was observed in the cases of Z_50BD (9%) and Z_50 UD (11.5%). The void content was lesser for Z_50UD (9.38%) than for Z_50BD (11.4%). The higher amount of voids in Z_50BD can be due to the loosely woven plain weave fabric by handloom, which might have led to the misalignment of warp and weft in the stacked fabrics during molding. Generically, BD fabric comes with interlacement points, leading to voids and flows in composites.

3.2. Thermogravimetric Analysis (TGA)

TGA analysis of composites (Figure 3) was done to determine the thermal stability of the composites, and the extracts are given in Table 2.

Figure 3.

Comparative TGA thermograms of the composites (RT to 900 °C at 10 °C/min in an air atmosphere)

Table 2. Details of Thermal Stability from TGA Curves^a.

composites	T0 (°C)	T5 (°C)	T10 (°C)
Z0	543	545	558
Z50UD	555	558	600
Z50BD	560	578	610

^aT₅: temperature at 5% weight loss and T₁₀: temperature at 10% weight loss.

Overall thermal stability of composites was higher than that of PEEK, and Z_50BD showed the highest. This was also justified by Yan et al.¹⁹ that incorporating PBO fibers in the blend of PEEK/PI improved its thermal stability. T₀, T₅, and T₁₀ improved with Zylon as observed for Z_50BD and Z_50UD. Z_50UD showed a degradation temperature lower than that of Z_50BD. This is caused by the presence of Zylon fibers in both warp and weft in Z_50BD, acting as a barrier and delaying heat transfer to degrade the PEEK matrix. However, the degradation is early in the case of Z_50 UD, which consists of fibers in only one direction. Hong et al.¹⁶ reported that the onset of degradation of PBO takes place at around 883 K (609 °C). This was evident in the second baseline shift in the cases of Z_50BD and Z_50UD, corresponding to the initiation of degradation of Zylon. However, polymers’ degradation mechanisms are complex and comprise several simple reactions that are difficult to analyze separately. The quantitative contributions of these reactions to the degradation process are impossible to evaluate.²⁷ Since PEEK and Zylon are polymeric phases, analyzing the complex degradation mechanisms in UD and BD composites is challenging.

3.3. X-ray Diffraction

The XRD diffractograms for Z₀ and the composites are shown in Figure 4a,b. PEEK shows intense peaks at 18.9, 22.8, and 28.8°, and Zylon fibers show intense peaks at 15.34, 27.23, and 38.2°.^28,29 The high-intensity diffraction peaks of Zylon are due to the highly oriented fibrils. The peaks observed in the case of Z₀ were in tune with the literature, whereas for Z_50UD and Z_50BD, the peaks were observed for both PEEK and PBO. However, a peak shift was also observed for the Z_50BD, and Z_50UD composites, which could be due to the change in the crystallinity of Z_50BD (36%) and Z_50UD (38%) and crystal rearrangement (possible trans-crystallinity).

Figure 4.

XRD spectra of (a) Z₀ and (b) Z_50BD and Z_50UD.

Zylon fibers behave differently from inorganic fibers such as carbon, glass, and basalt because they are polymeric fibers (polybenzoxazole). As a result, the crystal nucleation and growth processes are distinct,¹⁹ leading to a decrease in the crystallinity of PEEK, as shown in Table 3. The higher crystallinity in the case of Z_50UD led to improved mechanical properties of Z_50UD compared to those of Z_50BD.

Table 3. Crystallinity of Composites.

composites	Z0	Z50BD	Z50UD
crystallinity (%)	41	36	38

3.4. Cryo-fractured Samples: Cross-Sectional Analysis by SEM

The Z_50UD composite was cryo-fractured in liquid nitrogen to observe the damage to the reinforcing fibers (PBO) in the matrix. SEM micrographs at various magnifications are shown in Figure 5. These indicate that the Z_50UD contains a dense network of PBO multifilaments, which were fibrillated at the ends due to fracture. The fracture was at cryotemperature, and the type of fibrillation confirmed the least elongated fibers.

Figure 5.

Cross-sections of cryo-fractured Z_50UD at (a) 200×, (b) 500×, and (c) 1 kX.

3.4.1. Micro-CT Tomography of the Composite

The tomography was done on Z_50UD to understand details of the bulk on composites, voids, matrix-rich areas, alignment of fibers, etc (Figure 6).

Figure 6.

(a) 2D tomogram showing voids (2D tomogram slice obtained from the side indicating the extent of inter-tow voids. The voids around the tow of Zylon (red-colored zone) as well as in the middle of tow (purple-colored zone). Different colored voids are indicative of variation in sizes of voids); (b–d) 3D tomograms showing (b) fibers and voids: Gray-colored Zylon fibers oriented in one direction. 3D tomograms show the efficiency of the matrix in covering the surface of fibers; red-, blue-, and purple-colored regions indicate the different sized voids along the tows of fibers,. The voids are fairly continuous. However, in the case of mechanical loading, the applied force was in the direction of the fiber axis. Hence, the voids were not able to jeopardize the performance properties under uniaxial tension, (c) matrix and voids: bulk of the matrix (PEEK) with voids in colored regions; red, blue, green, and purple volume indicates the different sized voids with its continuity. The golden and light yellow inside the tow indicate the monofilaments of Zylon and the matrix. The presence of such variable-sized voids could be due to highly viscous PEEK (which was unable to flow into complex space in tows), and (d) only voids: porosity analysis of a sliced specimen showing only voids; colored voids of various sizes are oriented in the fiber direction, indicating the inefficiency of PEEK to penetrate inter-tow gaps and wetting monofilaments in fibers. This inefficiency could be more inclined toward damage in transverse loadings as in the cases of BD composites (since voids play a major role in crack initiation and propagation). Uniaxial tension parallel to the fiber length and oriented voids in the UD composite do not play significant roles; occurrence frequency vs. void volume for (e) Z_50BD and (f) Z_50UD.

3.5. Mechanical Properties

3.5.1. Microhardness of Composites

3.5.1.1. Vickers Hardness

The Vickers hardness of the composites (Figure 7) was significantly higher than that of PEEK (Z₀). This could be because PEEK is a relatively flexible chain with two ether groups in its backbone. In contrast, PBO has the heterocyclic benzoxazole moiety, imparting stiffness and rigidity, leading to improved hardness of composites.

Figure 7.

Vickers hardness of composites.

3.5.1.2. Scratch Hardness

The scratch hardness data depend on the fibers’ direction in the composites (Figure 8).

Figure 8.

Scratch hardness of the composites.

Hence, the additional orientation of fibers was selected with respect to the scratching direction as follows.

• Z_50UD-AP is the composite Z_50UD whose fibers were antiparallel to the scratching direction.

• Z_50UD-P is the composite Z_50UD whose fibers were parallel to the scratching direction.

The composites showed a substantial difference in behavior, and the order was Z_50UD-AP (57%) > Z_50UD-P (24%) Z_50BD > (20%) > PEEK (Z₀), where P stands for the parallel orientation of fibers and AP stands for the antiparallel orientation of fibers. The increased hardness can be attributed to the presence of PBO fibers, which might have undergone fibrillation during scratching by the indenter as they are efficient energy absorbers. This is evident from the scratched surface micrographs of Z_50UD AP in Figure 9b. The antiparallel direction proved to be the best since the indenter had to cut the fiber across, which resisted the maximum. In the case of Z_50UD P, the fibers undergo buckling instead of fibrillation, Figure 9a.

Figure 9.

SEM micrographs of scratched surfaces of (a) Z_50UD P and (b) Z_50UD AP.

3.5.2. Tensile Properties

Figure 10 shows the tensile strength, modulus, and elongation at break for the composites. It was observed from Figure 10 that the tensile strength of the composites increased significantly by 160% for the BD composite and 500% for the UD composite. Addition of all 50 wt % Zylon fibers in a loading direction in the UD composite (Z_50UD) had benefited significantly, while 25% fibers in the loading direction and 25% in the perpendicular direction in Z_50BD had reduced the benefits drastically.

Figure 10.

(a) Tensile strength, (b) tensile modulus, and (c) elongation at break of the composites.

At the break of the composites, elongation decreased drastically by 95% (Figure 10c). Zylon fibers are stiff (270 GPa Young’s modulus) than PEEK (3.6 GPa Young’s modulus), which is more flexible due to the two ether groups in the backbone.

Zylon fibers improved the tensile modulus of PEEK (Figure 10b) significantly (197% for Z_50BD and 662% for Z_50UD), which was due to the difference in the moduli of PEEK and Zylon (3.6 GPa for PEEK and 270 GPa for Zylon fibers). The benefits of Z_50BD were lower because only half the amount of Zylon fibers contributed to the enhancing modulus.

3.5.3. Notched-Izod Impact Strength

The orientation of fibers with respect to the hammer direction is shown in Figure 11a, while Figure 11b shows the impact strength of the composite, which increased significantly by adding Zylon fibers and fabric in PEEK. A significant improvement in impact resistance was observed, and the performance order was as follows.

Figure 11.

(a) Orientation of fibers with respect to the loading direction and (b) impact strength of composites.

Zylon fibers are well known for their excellent energy absorption capacity, twice that of p-aramid fibers, which are used in ballistic applications and are responsible for composite high impact strength. The reasons behind such an outstanding improvement could be due to the presence of voids and fibrillation of Zylon in both Z_50UD and Z_50BD. Nevertheless, the most dominating mechanism for increased impact strength remains the same, i.e., efficient energy absorption. A fractographic analysis of impact-fractured samples was done to understand the difference in the behavior of the two composites.

The micrographs in Figure 12 in the left column for Z_50UD at various magnifications show excessive fibrillation of filaments into microfibrils, confirming that the major energy damage mode was fibrillation. On the contrary, the micrographs in the right column for Z_50BD show excessive fiber breakage rather than fibrillation. The broken weave of the fabric (highlighted) can also be seen. Thus, the higher impact resistance of Z_50UD was due to excessive fibrillation of fibers, thereby absorbing the maximum energy. The impact-fractured portions of specimens at the central part of notches were analyzed with SEM to understand the reasons for the difference in damage and are shown in Figure 12.

Figure 12.

SEM micrographs of impact-fractured samples at the central portion of the notch: (a–c) different magnifications of impact-failed samples of Z_50UD and (e–f) different magnifications of impact-failed samples of Z_50BD.

3.5.4. Lap Shear Strength

The lap shear strength test of the adhesive joints was evaluated as per the standard ASTM D1002, and the schematic of the loading direction with respect to fibers in a joint is shown in Figure 13a,b, showing the lap shear strength of joints using braided fibers.

Figure 13.

(a) Schematic of developed joints and (b) lap shear strength of developed joints using braided fibers.

The following was the performance order:

It was observed that the maximum LSS was achieved when the UD fibers were antiparallel to the direction of tensile loading. These fibers offered maximum friction when a layer adhering to coupons was pulled from the coupons during tensile stretching.

The surfaces of Z_50UD P, Z_50UD AP, and Z_50BD fractured after LSS testing showed adhesive failure since all of the developed joints showed the whole material on one substrate and a neat substrate on another (Figure 14). The materials transferred on the steel substrate were studied for Z_50UD P and Z_50UD AP to observe the mechanisms. In the case of Z_50UD P, more fibers are exposed and detached from the matrix due to the orientation of the fibers parallel to the loading direction. In contrast, fewer fibers are exposed and remain protected by the layer of PEEK in the case of the Z_50UD AP.

Figure 14.

Fractured surfaces of the Z_50UD P and Z_50UD AP at (a, b) lower magnification (500×) and (c, d) higher magnification (1500×).

4. Conclusions

The development of novel composites of braided fibers of PEEK and Zylon (50:50 wt %) in both BD and UD forms to achieve maximum wettability of reinforcing fibers was successfully done. Zylon fibers excelled immensely in various properties, and the UD composite performed better than the BD composite. The performance properties of UD composites were specific to the orientation of fibers, which was best when fibers were in the loading direction. Overall, it was concluded that Zylon fibers have the vast potential to improve the mechanical and triboproperties of PEEK in all wear modes. The orientation of fibers with respect to loading was the performance-controlling parameter since both the composites contained 50 wt % fibers. Performance was best when all were in one direction (UD composite). When 25% were in one direction and 25% were in the antiparallel direction (BD composite), performance improvement with respect to PEEK lessened. These composites can be used for structural and tribological applications.

The authors declare no competing financial interest.