3.2. The Linear Tribological Contact
Table 4 presents the tribological characteristics of neat PES and the PES-based composites determined in the linear tribological contact (according to the ‘block-on-ring’ scheme) against the GCr15 steel counterface. For all studied composites, no significant increase in temperature at the tribocontact was observed under the metal–polymer rubbing conditions compared to neat PES. This fact reflected, in part, the high thermal conductivity of the GCr15 steel counterface, which effectively dissipated the frictional heat generated by overcoming the friction force. Loading PES with SCFs 100 µm long at AR of ~10 did not reduce WR values, while CoF levels only slightly decreased to a level of ~0.4–0.5.
According to the authors of [
35], lowering CoF levels to ~0.2 could indicate the tribological layer formation on HPP-based composites, which are characterized by high strength. This should reduce the CoF values under linear tribological contacts. This phenomenon contributed to a significant reduction in WR compared to neat PES. The authors of [
35] refer to the “tribological layer” as a ratcheting layer (up to several microns thick), consisting of the polymer matrix reinforced with fractured SCFs, which was compacted (tamped) and polished due to repeated sliding of the counterface with an
Ra surface roughness of 0.20–0.25 µm roughness. The tribological layer was formed from debris in two stages: a sharp increase in the CoF to a value of approximately ~0.4–0.5 followed by a subsequent decrease in CoF to about ~0.1–0.2. If the CoF level then remained stable, it can be asserted that the formed tribological layer was firmly adhered and fulfilled its functional role. In the following sections of this study, the interpretation of the results will be based on this assumption.
Loading PES with 10 wt.% SCFs 200 µm long reduced WR values by 9.6 times compared to that of neat PES, while enhancing the content of SCFs up to 30 wt.% was not accompanied by a further decrease. The reinforcement of PES with SCFs of 2 mm length in the composites PES/10SCF
2mm and PES/20SCF
2mm led to a significant reduction in both WR values and CoF (WR decreased by 11 and CoF by 2 times), whereas in PES/30SCF
2mm, the wear intensity decreased only slightly (by 1.3 times compared to unfilled PES) with an average CoF of 0.47 (
Table 3). Thus, it can be concluded that no tribological layer was formed on the wear track of the PES/30SCF
2mm composite.
Figure 2 shows the time dependencies of the CoF for neat PES and all studied composites. For neat PES (
Figure 2a), after a brief running-in period, the CoF maintained a consistently high value (~0.57) along with a high level of ‘high frequency’ (HF) oscillations. According to
Figure 2b, the friction coefficient also had a high average value (0.411–0.503) and was characterized by a noticeable level of HF oscillation during the tribological test. According to the assumption made in this work, this indicates that no tribological layers had formed on the wear track surface.
A clear tendency for tribological layer formation was observed only in three composites, namely the PES/10SCF
200µm (
Figure 2c and
Figure 3a,d), PES/10SCF
2mm (
Figure 2d) and PES/20SCF
2mm ones. Their CoF levels quickly reached a consistently low level of 0.20–0.25. In these cases, three stages were evident in the CoF time dependencies: (i) growth (running-in stage), (ii) reduction stage and (iii) stage of stable low-level (or gradually enhancing) CoF. For the PES/10SCF
200µm composite, the formed tribological layer, reinforced with fragmented (fractured) SCFs, is illustrated by the photo in
Figure 3a (and its smooth wear track profile is presented in
Figure 3d). Conversely, in the case of PES/20CF200 µm and PES/30CF200 µm, it is evident that the SCFs remained intact on the friction surfaces (
Figure 3b,c), which did not allow the formation of a smooth (compacted) tribological layer. So, the wear track profiles were characterized by sawtooth shapes (
Figure 3e,f).
According to the authors, when shorter SCFs of 200 μm length were loaded, the formation of a tribological layer was evident as a layer of mixture reinforced with fractured fibers, as illustrated by
Figure 3a (as well as by a smooth tribotrack’s profile,
Figure 3d). The regions outlined in red corresponded to the tribological layer. The latter was fragmented and non-uniform due to instability of adherence. This was manifested by the pattern of corresponding CoF time dependence (
Figure 2c). When PES/20CF
200µm and PES/30CF
200µm composites were tested, the CF maintained their integrity (
Figure 3b,c). The latter did not ensure the formation of a smooth (compacted) tribological layer, while the wear track possessed sawtooth profiles (
Figure 3e,f).
The indicated stages are interpreted as follows: At the beginning of the tests (the running-in stage), a high and unstable CoF was observed. This can be explained by the interaction of asperities on the surface of the GCr15 steel counterface (Ra~0.25 µm) with those on the friction surface of the composite, formed as a result of grinding during sample preparation. A subsequent decrease in CoF to 0.1 occurred after a sliding distance of ~100 m. This is associated with the tribological layer formation and consolidation. A subsequent relatively constant CoF value of ~0.25 was observed at the test distance starting from 200 to 1000 m. The episodic instability of the CoF value may be related to the detachment and subsequent formation of new fragments of the tribological layers on the wear track surfaces. Although the process of tribolayer formation was initiated from the very beginning of the tests by adhered debris, the process was completed only at the test distances of about 100 m. This instability was explained by the high mutual activity of PES [
3] and GCr15 steel components, whose interaction drives the development of tribo-oxidation reactions [
36,
37]. This phenomenon will be further discussed in the corresponding section.
In addition to the episodic short-term sharp increases in CoF values observed in the time dependencies (
Figure 2c for the PES/10SCF
200µm composite, as an instance), a distinctive feature of such graphs during the formation of the tribolayer was the gradual increase in CoF during the third stage; this took place when PES was loaded with SCFs 2 mm length (
Figure 2d for the PES/10SCF
2mm and PES/20SCF
2mm ones). It should be noted that some fractured SCFs 2 mm in length were evident.
The authors believe that, as will be further investigated and discussed in the analysis of the topography of the wear track and other types of counterparts, the gradual increase in CoF with the formed tribolayer was due to more active interaction of the GCr15 steel counterpart and the composites, resulting in the shifting/flowing of the tribolayer relative to the bulk sample. Furthermore, when 2 mm long SCFs break down, their physical contacts with the bulk specimens are lost. The absence of such an effect in the composites with shorter SCFs (PES/10SCF200µm) was attributed by the authors to a more efficient retention (similar to tree roots) of the entire tribological layer, including due to a more stochastic directional distribution (compared to SCFs 2 mm long).
Since the ceramic counterpart is more inert with respect to the PES matrix, further tests were carried out in the ceramic–polymer tribological contacts under the same testing conditions for comparison purposes. It should be also noted that the use of the ceramic counterface was accompanied by two additional effects: its high hardness and lower heat dissipation. However, the role and contribution of these factors were not analyzed in detail within the framework of this study.
Table 5 presents the tribological characteristics of neat PES and PES-based composites obtained in tests against the ceramic counterface. From
Table 5, it can be seen that the average CoF levels were approximately two times lower compared to those for the GCr15 steel counterface (
Table 4). This result was evidently associated with the minimal development of tribo-oxidation processes in the tribological contact zone. At the same time, there was an increase in counterface temperature in the contact zone, which was likely due to the lower thermal conductivity of Al
2O
3. Replacing the GCr15 steel counterpart with the ceramic one reduced WR values by a factor of three for neat PES, although its CoF remained approximately at the same level. For the studied composites, a significant reduction in WR was observed across all aspect ratios, ranging from 7 to 43 times.
In
Figure 3, the dependencies of the CoF for all composites on the testing distance are presented. These results were also associated with the tribological layer formation. For neat PES (
Figure 4a), the CoF level remained consistently high (~0.56) during the steady-state wear stage. After PES was loaded with SCFs 100 µm long at contents of 10 and 20 wt.% (
Figure 3b,c), the running-in stage increased to 400 m, during which more intense wear of the composites occurred. However, the WR values did not exceed (2.3–3.6) × 10
−6 mm
3/N·m. At the same time, on the polished wear track surfaces, some fractured SCFs were seen (
Figure 4a,b,d), which was one of the key criteria for the tribological layer formation. The observed local spikes in the CoF levels were likely related to damage and removal of the tribological layer fragments (during attempts at tribological layer formation and consolidation).
After PES was loaded with 30 wt.% SCFs 100 μm long, the running-in stage was short (within 100 m according to
Figure 4d), accompanied by the lowest WR value of ~0.57 × 10
−6 mm
3/N·m. However, a gradual increase in the CoF level was recorded, similar to the cases shown in
Figure 2d for the GCr15 steel counterface. This might indicate that SCFs 100 μm long were unable to prevent displacement (flow) of the formed tribological layer relative to the bulk sample. However, in terms of the achieved wear resistance for the PES/30SCF
100µm composite, the result was the highest. This could be attributed to the fact that fractured SCFs with minimal length are most effective in reinforcing the tribological layer.
For the PES/10SCF
200µm composite, the CoF changing kinetics was similar (
Figure 4b), although shifted in distance compared to the cases with 10 and 20 wt.% SCFs 100 μm long (where the local peak corresponds to the tribological layer formation). Upon further increasing the contents of SCFs 200 μm long up to 20 and 30 wt.%, the trends of CoF changing kinetics were altered (
Figure 3e). Although the average CoF levels over the entire test distance were within 0.2, it was not possible to conclude that the steady-state wear stage had been reached. As a result, the abrupt CoF changing kinetics was accompanied by WR values of ~(1.15–1.97) × 10
−6 mm
3/N m (which, however, are low values, particularly after PES was loaded with 30 wt.% SCFs 200 μm long).
The photos presented in
Figure 5g–i reveal that non-solid (fragmented) layers were formed on the wear track surfaces (as evidenced by their “torn” profiles, shown in
Figure 5j–l). Only the PES/10SCF
200µm composite exhibited a distant reinforcement with fragmented SCFs, whose CoF changing kinetics followed the above-described interpretation of the conditions for the tribological layer formation.
Regarding the CoF composites with the longest SCF (2 mm long) (
Figure 4d), the changing kinetics of CoF (a sharp increase followed by a gradual decrease to a stable level of 0.07–0.10) clearly indicated the formation of a continuous tribolayer. This means that the composition, fracture pattern of SCFs and the formed structure of the composites provided the ability to form, adhere and retain the wear-resistant tribological layers.
Since the composites loaded with SCFs 2 mm long showed higher tribological characteristics, their wear tracks were subjected to analysis.
Figure 6 and
Figure 7 show profilograms and micrographs of the wear track surfaces and their profiles on these specimens after the tribological tests against both GCr15 steel and ceramic counterfaces.
When tested against the GCr15 steel counterface, neat PES exhibited a high WR value, accompanied by the formation of longitudinal (relative to the sliding direction) grooves on the wear track surfaces, reaching up to 10 µm in width (
Figure 6a–c).
In contrast, in the PES/10SCF
2mm and PES/20SCF
2mm composites, fragmented SCFs were observed on the wear surface, and the reduced wear intensity suggests the formation of fragmented (discontinuous) tribological layers (
Figure 6e,h). The
Ra surface roughness increased two-fold compared to neat PES (from 0.150 µm up to 0.328 µm). As seen in
Figure 6f,i, the wear track depths in the composites decreased by an order of magnitude (compared to neat PES) due to the tribological layer formation.
In the PES/30SCF
2mm composite (
Figure 6j–l), a tribological layer was not formed, although some adhered debris was observed on its wear track surface. In addition, no grinding signs were found there after the tribological test against the GCr15 steel counterface. Nevertheless, several furrows caused (
Figure 6j) by the plowing action of the fractured SCFs were visible on the friction track. Consequently, the
Ra surface roughness increased three-fold compared to the PES/20SCF
2mm composite (from 0.328 µm up to 0.954 µm).
When PES was tested against the ceramic counterface (
Figure 7a–c), the temperature in the tribological contact increased up to 51 °C (
Figure 6j) due to the low heat removal through Al
2O
3. This could have further stimulated the development of tribo-oxidation processes (which were accompanied by the formation of longitudinal furrows along the sliding direction (
Figure 7a), as well as a high CoF level of ~0.56 and the oscillating pattern of its change (
Figure 4a)) [
35].
On the wear track surface of the PES/10SCF
2mm composite, some fractured SCFs were observed that enhanced the
Ra surface roughness up to 0.372 µm (
Figure 7d). In this case, a tribological layer was formed, which included both intact and fractured SCFs (
Figure 7e). The partial preservation of intact SCFs could be caused by both their low quantity and activity in the development of mixing processes due to less active interaction with the ceramic counterface.
When the content of SCFs 2 mm long was increased to 20 wt.% (
Figure 7g), the
Ra surface roughness on the wear track decreased by ~35% (0.273 µm versus 0.372 µm), which corresponded to the formation of a discontinuous tribological layer reinforced with fragmented SCFs (
Figure 7h). This was not accompanied by an increase in the CoF level, which remained extremely low (0.074), while the WR value, conversely, was quite high, WR = 3.77 × 10
−6 mm
3/N m. The authors attribute this to the discontinuous (non-solid) structure of the tribological layer.
A continuous tribological layer was observed only at the content of SCFs 2 mm long of 30 wt.% (
Figure 7j,k). In this case, the
Ra surface roughness decreased by 20%, and the WR value of 0.77 × 10
−6 mm
3/N m was minimal. However, it was slightly higher than in the case for the PES/30SCF
100μm composite (0.57 × 10
−6 mm
3/N m), although with a higher CoF level of ~0.183, as seen in
Figure 5c,f. It should be noted that in the latter case, a large number of SCFs were visible on the wear track surface of the PES/30SCF
100μm composite, most of which were not fractured during friction. This indicates that SCFs 2 mm long could be part of the tribological layer. However, this effect developed only during friction against the ceramic counterpart, i.e., while minimizing chemical interaction between the components of the friction pair.
Based on the above, the formation of the wear-resistant tribological layer depended on the superposition of several factors, not only (i) the three-stage pattern of the CoF changing kinetics, reaching constant low values of 0.1–0.2 at the steady-state wear stage, but also (ii) low WR values, (iii) the formation of a polished continuous surface on the wear track and (iv) the reinforcement of PES with SCFs.
A more detailed analysis of the surface topography of the wear track of the PES/30SCF
2mm composite in terms of tribological layer formation was carried out. For this purpose, the structure of both tribological and subsurface layers in the PES/30 CF2mm composite during friction against a ceramic counterface was investigated using SEM at different accelerating voltages (10, 20 and 30 kV). It is known that changing the accelerating voltage value enables the penetration depth of electrons into a substance to be varied. Consequently, the backscattered electrons from such depths reflected patterns shown in
Figure 8. X-ray computed micro-CT can provide more detailed information on the tribological structure [
38,
39]. However, the thickness of the latter in units of microns might be too small to be duly resolved by micro-CT means.
At an accelerating voltage of 10 kV (enabling the investigation of the wear track surface), in addition to some adhered fragments of debris, a large number of fractured SCFs was observed (
Figure 8a). At 20 kV (
Figure 8b), the wear track boundaries became less defined, and the number of visible fragments of debris decreased. At an accelerating voltage of 30 kV (
Figure 8c), which allowed us to examine the subsurface layer, intact SCFs were visible, remaining undamaged during the tribological test. Thus, it can be “roughly” assumed that the thickness of the tribological layer was limited by this depth (as a distance from the wear track surface).
To assess the depth of electron beam penetration, the methodology described in [
40] was employed. The corresponding backscattered electron and X-ray plumes were calculated using the “Casino” (CASINO_v2.4.8.1) free software, based on the Monte Carlo method [
40]. As shown in
Figure 9, electrons were backscattered at a depth of 610 nm relative to the surface for U = 10 kV, while for U = 20 kV and U = 30 kV, this depth was approximately 1850 nm and 3500 nm, respectively. For this reason, this approach allowed us to study the only thin subsurface layer, whose structure inevitably changed upon testing. Based on the obtained data, it could be concluded that the tribological layer thickness was approximately ~2–4 µm for the PES/30SCF
2mm composite. On the other hand, another estimate of the tribological layer thickness could be obtained based on the characteristic diameter of SCFs, by which the tribological layer was actually reinforced. In this case, its thickness could reach 6–7 µm.
Figure 10 shows Raman spectra of neat PES and the composites loaded with SCFs 2 mm long before and after the tribological tests against the GCr15 steel counterface. For the non-tested samples, both neat PES and the composites (
Figure 10a,c) exhibited characteristic peaks from sulfones and sulfoxides (SO
2—1000–1420 cm
−1; S=O—1225–980 cm
−1; S–O—870–690 cm
−1) and C–S—570–710 cm
−1. In the 1610–1790 cm
−1 region of reciprocate wavelengths, characteristic peaks corresponding to ethers (C=C and C–C) were observed. At the same time, the Raman spectra within the wear tracks (
Figure 10b,d) indicated the development of tribochemical processes during the tribological testing. Changes in the spectrum of neat PES were observed in the 570–710 cm
−1 region of reciprocate wavelengths. Specifically, as the content of SCFs increased, the peaks from sulfones and sulfides, as well as from ether compounds, decreased down to zero. This suggested a change in the chemical composition of the wear track surface layer: in the spectrum of the PES/30SCF
2mm composite, two diffuse maxima were observed in the 1650 and 1350 cm
−1 range, indicating the formation of more complex esters and oxides.
A quite different pattern was revealed for the ceramic–polymer tribological contact (
Figure 11). For the wear track of neat PES (
Figure 11c), the Raman spectrum was identical to the initial one (
Figure 11a). For the composites with increasing the content of SCFs, the amplitude of the peaks from ether and sulfonic compounds decreased. However, the chemical structure did not change: within the wear track, the intensity of the peaks was weaker (
Figure 11b,d) compared to the initial ones (
Figure 11a,c).
Therefore, the fact that tribo-oxidation processes occurred during the tests against the GCr15 steel counterface suggested that the latter interacted more actively (including in terms of shear strains in both tribological and subsurface layers) with the studied composites.
Since metal–polymer contacts are significantly more common than ceramic–polymer ones, the obtained results are discussed below as a comparison with those for the less-“active” AISI321 steel counterpart.