In this paper, short carbon fiber-reinforced nylon (SCFRN) composites were fabricated using the fused deposition modeling (FDM) technology. In particular, different surface textures, namely convex squares and triangles, were created by using the printing method. It was found that fiber reinforcements could effectively enhance the load-carry capacity of the printed polymeric materials. Moreover, the tribological performance of SCFRN can be further improved with the surface textures. Microscopy observations revealed that the surface textures are particularly beneficial for the wear reduction by collecting hard wear debris such as broken fibers. The work has demonstrated that 3D printing technology has great potential for developing new wear-resistant engineering materials by controlling and creating desirable compositions and geometric structures/textures simultaneously.
Over the past decade, additive manufacturing (AM) technology, also known as 3D printing, has attracted increasing attention due to its high flexibility and low cost for rapid prototyping [1–3]. Besides metals and ceramics, polymeric materials have been increasingly developed for various industrial applications in AM, due to their synthetic versatility and lower cost processability, as well as their excellent performance to weight ratio . For example, polylactic acid (PLA) has been used for ophthalmological implants and regenerative medicine such as tissue and organ fabrication [5,6]. Nevertheless, most neat polymers cannot fulfill the requirements for mechanical engineering applications because of their relatively low stiffness and strength. In practice, reinforcement fillers such as fibers are commonly used to improve the property profiles of polymeric composites to meet the industrial requirements for different applications, e.g., in aerospace and automotive industries . Up to now, a number of research papers on printed polymer composites have concentrated on their functional and basic mechanical properties such as strength and toughness characteristics [8,9]. Little has yet been published, however, on composites’ friction and wear behavior, although it has been shown that fiber-reinforced thermoplastics and thermosets are often used as gears, rollers, and dry slide bearings [10,11]. Prusinowski et al. are among pioneers to explore the tribo-applications of the printed fiber-reinforced polymer composites. They have reported that the tensile strength of the fiber-reinforced composites increases proportionally to the percentage of fiber in the 3D-printed composites . Further, it was proposed that fiber volume content above 9% provides lower wear intensity compared to the unreinforced 3D-printed materials .
Among the various printing technologies, fused deposition modeling (FDM) becomes the most promising approach for developing high-performance polymer composites, which allows introducing different fillers (including continuous fibers) into the filaments . Nevertheless, in spite of the wide applications and high cost-efficiency of FDM printers, FDM printing methods still have drawbacks in making composite materials such as the limitation of control over the fiber placement, poor interface . In particular, the mechanical and tribological properties of FDM printed plastic parts can be sometimes comprised due to the internal voids and the relatively low interfacial bonding strength between printed layers [9,15]. Different methods have been introduced to overcome such shortcomings, e.g., by using pressure rollers  and the laser-assisted printing method . However, these additional treatments would also bring more complexity and costs of fabrication processing. In this paper, attempts were made to enhance the tribological performance of printed polymeric samples, with fiber reinforcements and the designed surface textures. The friction-reducing effect of surface textures was first reported by Hamilton in 1966 . Since then, a number of research articles have been published on this topic. Braun et al.  confirmed that surface structures could contribute to the improved tribological properties, together with the other mechanical properties of the materials. Zeng et al.  indicated that the adhesive wear of PTFE-based materials could be reduced significantly with the surface texture because the discontinuous textures would effectively reduce the adhesion force in dry sliding contact. Vilhena and Ramalho  reported that 3D-printed untextured specimens showed a higher friction coefficient than the specimen with a higher texture density because of a reduction in plowing friction and an increase in adhesion. Qi et al.  suggested that wear debris would be collected by surface textures during the dry sliding, thereby reducing the abrasive wear. Lu et al.  concluded that converging triangular textures of steel is beneficial for friction reduction as real contact length variation is a major factor controlling the friction behavior. More recently, Wang et al.  proposed that SLM printed convex squared texture possessed better friction–reducing ability than concave textures and fully sintered surface.
In the present work, we will study the friction and wear behavior of FDM printed nylon and short carbon fiber-reinforced nylon composites with surface textures. Toward that, fully printed surface, convex squared surface texture, and convex triangular surface texture were fabricated by using the FDM technology. Accordingly, the tribological behavior of the printed samples was investigated under different loading conditions. The work aims to explore the potential of the FDM technique for developing high wear-resistant polymeric materials/structures, by taking its advantage of controlling fiber fillers, as well as the unparalleled flexibility of manufacturing 3D structures.
2 Material Preparation and Experiment
In this study, two types of filaments, namely neat Nylon (Markforged nylon white) and short carbon fiber-reinforced nylon (Markforged Onyx), were used. The samples were fabricated using a Markforged Mark II system, with a layer thickness of 0.1 mm. All specimens were printed with solid infill at room temperature with a printing speed of 30 mm/s. The temperature of the printing filament was 275 °C. The mechanical properties of the printed materials were characterized using a universal testing machine (Instron 5567) with a loading rate of 5 mm/min, following ASTM standard D3039. The results are given in Table 1. It is evident that the FDM printed nylon and short carbon fiber-reinforced nylon composites (SCFRN) showed a similar tensile strength. Nevertheless, SCFRN provided a higher tensile modulus and a lower tensile elongation than nylon, suggesting the addition of short carbon fiber improves the stiffness but reduces the ductility. Figure 1 shows the SEM image of the fracture surface of the FDM printed SCFRN, showing that the placement of short carbon fiber in the FDM printing filament is random. The fiber volume fraction of the SCFRN is 9.6% ± 0.3%, which is determined by using the thermogravimetric analysis (TGA) method .
2.2 Tribological Tests.
All the wear tests were carried out at room temperature under dry sliding conditions, using a pin-on-disk configuration on a commercial tribometer (NANOVEA-MT/60/NI). The samples were prepared as cubic pins, with a size of 4 mm × 4 mm × 12 mm. The counterparts were stainless-steel disks (LS2542, SKF) with 25 mm internal diameters and 42 mm external diameters. The hardness of the disk is ∼910 HV with a surface roughness of ∼220 nm. In particular, the polymeric samples were prepared with different textures, i.e., 4 convex squares and triangles with a depth of 2 mm, as shown in Fig. 2. It is worth noting that with the printer used in this paper, different printing paths were provided for nylon and SCFRN filaments. To achieve the best printing quality of the printed specimens, same texture dimensions but different texturing densities (a gap between each texture) were applied for nylon and SCFRN. The detailed dimensions of textures are given in Fig. 3.
After the test, the worn surface of each specimen was observed using an SEM (Zeiss EVO SEM). Further, nano-indentation tests were carried out to study the surface properties before and after wear tests. The tests were conducted on a UMIS ultra nano-indentation system with a diamond Berkovich indenter. The maximum load was 10 mN with the loading/unloading rate of 1 mN/s and 5 s holding time at peak load. At least five indentation tests were repeated for each specimen, and the average values were reported as the measured hardness.
3 Results and Discussions
3.1 Characterization of the Printed Surface Textures.
Following the geometric design shown in Fig. 3, both neat nylon and SCFRN samples with different surface textures were printed. Figure 3 also compares the actual size (dotted line) of the printed specimens with the designed sizes (solid line). As shown in the figure, the actual printed size is always larger than the designed one due to the thermal expansion and recovery of polymeric materials . Accordingly, the size of the squared and triangular textures becomes smaller. It is noticed that SCFRN specimens achieved more accurate texture shapes than those of neat nylon, which can be explained by the higher stiffness of SCFRN filaments . The printing path also plays an important role in determining the printing quality. Overall, for triangular textures which require sharp angles, the samples showed rather poor quality. There is an additional triangular texture on the SCFRN specimen because of the introduction of voids and defects when printing more complex shapes with the FDM printing method [28,29]. Hence, to achieve good dimensional accuracy of printed samples, all the factors such as thermomechanical properties of the polymeric matrix, mechanical properties of filaments, and printing path need to be carefully considered.
3.2 The Friction Behavior of the Printed Composite Materials.
Figure 4 summarizes the friction behavior of the FDM printed nylon and SCFRN under different sliding conditions. The friction coefficient was the average value during the steady stage. As shown in Fig. 5, there is a running-in stage in the first 3–6 h, depending on the type of materials. After that, the pattern of friction coefficient becomes rather steady. Normally, it is believed that the contact surfaces would be significantly changed in the running-in stage due to the friction and wear process. In particular, polymeric transfer film layers would be developed and become stable in the contact region, which governs the friction behavior in the steady stage . The average friction coefficient of the fully printed surface and textured surfaces of neat nylon was similar, with a value of approximately 0.7. It is noticed that the friction coefficient of neat polymers remained a higher value in the steady, with observable variations. This can be explained by the relatively high adhesion force between the polymer sample and the transfer film layer formed on the steel counterpart, which often causes stick-slip behavior, i.e., the variations of friction coefficient [31,32]. With fiber fillers, the friction coefficient was reduced in the steady stage. The average friction coefficient for SCFRN in the steady stage was approximately 0.4. It is known for fiber-reinforced polymers, the stiff fibers would stand out and undertake the most load [33–35]. In this case, the adhesion forced would be effectively reduced, leading to a much smoother friction process (cf. Figs. 5(a) and 5(b)).
With the induced surface textures, the neat nylon showed lower friction in the running-in stage. However, in the steady stage, the average friction coefficient was little affected by the presence of surface textures. The worn surfaces of neat nylon with and without surface textures were examined by the SEM, as shown in Fig. 6. It is clear that the texture patterns were fully lost with the covered wear debris in square or triangular regions. Figure 7 shows the SEM images of the worn surface of convex squared texture and convex triangular texture after 5 h and 10 h wear test. It is noticed that surface textures were not fully covered with debris after 5 h wear test, corresponding to the relatively lower friction coefficient obtained by the sample. After 10 h, however, the textures were almost fully covered with wear debris, which led to a higher, unstable friction coefficient due to the increased adhesion forces. To further examine the status of surface textures, the cross-section of the worn textured surface was examined, as shown in Fig. 8(a). It is noticed that after 20 h wear test, the texture part remained, but was covered with continuous nylon material on the top. This can be explained by the deformation behavior of nylon under sliding conditions. During the wear process, nylon debris tends to be elongated and stretched, caused by shear force and friction heating [36,37]. The deformed nylon would show similar mechanical properties with higher crystallinity and less molecular length due to the thermal-mechanical effect of the friction force .
To verify the surface properties of the localized material on the texture part, nano-indentation tests were conducted in the area (as shown by the dotted line A in Fig. 6(b), compared with the tests on the original printed solid part (as shown by the dot line B in Fig. 6(b). The typical load-depth curves are given in Fig. 9. Accordingly, Table 2 summarizes the measured hardness of the FDM printed nylon with different surface textures after 10 h wear test. The results were slightly lower than the surface hardness of the material before the wear tests, which was 100.8 ± 5.6 MPa. This can be caused by the attached loose debris on the surface, which may be transferred back from the counterparts. Nevertheless, it is clear that the surface hardness of the covered textures is similar to that of the originally printed surface after 10 h tests. Accordingly, the resulted friction coefficient for samples with and without surface textures became comparable. Hence, with surface texture, the friction coefficient of neat nylon was reduced in the running-in stage due to the smaller real contact area and thus lower adhesion force. During the wear process, the surface textures were gradually covered with the deformed materials. As a result, the samples eventually showed similar contact surfaces and friction behavior to the samples having the fully printed surface in the steady stage.
For the printed SCFRN samples, a similar phenomenon was observed, i.e., the surface textures contributed to a lower friction coefficient in the running-in stage, but little affected in the steady stage. As shown in Fig. 10, due to the strong fiber fillers, the SCFRN showed much better wear resistance. The wear debris was also observed in the textures. However, the wear debris could not cover surface textures after 20 h (cf. Fig. 8(b)). As aforementioned, for the fiber-reinforced polymers, the fibers would carry the load and determine the contact conditions. Thus, the real contact would be mostly determined by the volume fraction of fibers. In the steady stage, the friction behavior would be governed by the fibers against the transfer films on the counterparts. In this case, the texture showed rather limited effects.
3.3 Synergistic Effects of Surface Textures and Fiber Fillers on Wear Performance.
Figure 11 summarizes the wear results of the printed samples tested under different loading conditions. The FDM printed nylon only applied for a 20-N wear test. With the further increase of load up to 30 N, the test was stopped after 10 h, as the nylon specimens were broken into two pieces owing to the thermal-mechanical failure of the material. Hence, for neat polymers, the results from only one condition, i.e., 20 N, were recorded. It can be seen that the wear rate of neat nylon was slightly reduced with the induced surface textures. As aforementioned, the textures would collect the wear debris during the wear process (cf. Figs. 10(b) and 10(c)), which consequently contribute to the further loading and wear process. As a result, the measured wear loss for surface-textured samples would be less than the actual material removal. Also, it is noticed for both samples, there was more wear debris in the middle area of worn surfaces (Figs. 10(a)–10(c)). This may be related to the higher adhesion forces and temperatures in the middle contact region due to the frictional heating.
For SCFRN, the wear resistance of the polymer composites was significantly improved with fiber reinforcements. More interestingly, the worn surfaces with textures appeared smoother with less fiber removal, associated with the further reduced wear rate. As shown in Fig. 10(a), the severe fiber removal can be seen in the middle region of the worn surface of fully printed SCFRN due to the high contact temperature and stress concentrations. It is noticed that the polymeric matrix was significantly softened and even melted in the area. Without the support of the surrounding matrix, large-sized fiber pieces were removed. As a result, fibers cannot fully contribute to the wear process. Moreover, the rigid fiber debris may also introduce the three-body abrasive wear and further attenuate the wear resistance of the material. For SCFRN samples, the textures were not fully filled with debris after 20 h and thus can effectively affect the friction and wear behavior in the whole test processing. With the surface textures, the stress and temperature distribution would be more uniform across the contact surface. Moreover, the textures could collect the large, hard fiber debris in the contact region, avoiding the severe three-body abrasive wear. As a result, the worn surfaces are much smoother, without the removal of large fiber debris. It is noticed that the shape of surface texture has no obvious influence on the wear behavior of FDM printed SCFRN. Nevertheless, the tribological properties are not internal materials’ properties but depend on the system where materials have to function. More work will be required to understand the tribo-effects of different surface textures, as well as their interactions with different fillers.
In this study, the friction and wear behavior of FDM printed nylon, and SCFRN composites have been investigated. Fully printed surface, convex squared texture, and convex triangular texture were introduced to determine the tribological behavior of FDM printed surface textures. From the results, the following conclusions can be drawn:
The FDM printed SCFRN exhibited a lower friction coefficient and specific wear rate than neat nylon under dry sliding conditions.
The specific wear rate of the printed SCFRN can be further reduced by inducing surface textures.
In our work, convex squared texture and convex triangular texture showed similar effects on the tribological behavior of the printed SCFRN.
Conflict of Interest
There are no conflicts of interest.
Data Availability Statement
The datasets generated and supporting the findings of this article are obtainable from the corresponding author upon reasonable request. The authors attest that all data for this study are included in the paper. No data, models, or codes were generated or used for this paper.