*3.3. Evaluation of Wear Resistance and Long-Term Friction Behaviour*

As mentioned in Section 2.4, the determination of the wear resistance was performed only for the 3 samples having an antifriction-coating: coated benchmark, 3D microtextured (90◦) coated and 3D microtextured (45◦) coated samples. The wear resistance was determined as the wear rate (mm3/h), which was calculated from wear volume values, measured using laser optical microscopy after the tribological tests, divided by the test duration. It is worth noting that since disc specimens possess an extremely high number of micropillars (and microvalleys) on their surfaces, it is almost impossible to precisely quantify their wear volume after the wear-based tribological tests, because wear measurements were performed using a normal height threshold technique using a defined level, which also encompasses intact microvalleys that did not experience any wear during the tribological tests. Therefore, reported wear values for the disc specimens should not be interpreted as absolute, but as relative, values, which still enable a valuable comparison to be made between the investigated tribological pairings.

But first of all, before presenting the measured wear rates of the 3 aforementioned sample variations, coefficients of friction during these wear resistance tests were also monitored and are presented in Figure 13a for a direct comparison to the friction coefficients measured during the load-varying tests (short-term tests; Figure 12). For this comparison, one should bear in mind that the wear tests were performed at a normal load of 125 N; therefore, on the corresponding diagrams of Figure 12, the friction coefficient value at a load of 125 N should be used as a comparison point. Furthermore, the first values of the coefficients of friction measured and presented in Figure 13a are after a test duration of 15 min, which is about 3 times longer than the short-term load-varying tests. In order to facilitate the direct comparison of the coefficients of friction related to both tribological tests, the results of Figure 12a,c were partially combined and are presented in Figure 13b, and the corresponding coefficients of friction are listed in Table 7.

**Figure 13.** Average coefficients of friction measured: (**a**) during wear resistance tests (long-term tribological tests: 125 N, 2 h) and (**b**) during load-varying tests (short-term tribological tests: 25–200 N, 5 min. per load level) (Figure 13 = partial combination of Figure 12a,c).

**Table 7.** Comparison of coefficients of friction measured during wear resistance (long-term: 125 N, 2 h) and load-varying (short-term: 25-200 N, 5 min. per load level) tribological tests.


From Figure 13 and Table 7, it is obvious that the friction coefficients obtained either from wear resistance tests (@ 15 min) or from load-varying tests (@ 125 N) are quite similar, which gives a certain degree of validation for the friction coefficients obtained from both different tribological tests. The small discrepancy observed for the 3D microtexture with a 45◦-angle may be due to the fact that the time-range of the compared values is different (15 min for wear tests, 5 min for load-varying tests).

It is worth noting that the error bars shown in Figure 13b are larger than those in Figure 13a due to different procedures used for friction measurements. The friction values shown in Figure 13b (short-term tribological tests with varying loads) were measured and averaged for a time duration of 0.2 s at the beginning and end of each load level (*t*begin = 0 s; *t*end = 5 min.); therefore, for each load level, and due to the short time period at a new constant load value, the friction values were partially measured when the system was still in the running-in phase (in which the friction may still be relatively unstable and may relatively differ from test to test), thus resulting in larger error bars when averaging values measured from 3 short-term tribological tests. On the other hand, the friction values shown in Figure 13a (long-term tribological tests at a constant load) were measured and averaged for a time duration of 0.2 s every 15 min (*t*<sup>1</sup> = 15 min.; *t*<sup>2</sup> = 30 min., etc.); therefore, for the first friction measurement at t1 = 15 min., the running-in phase of the system was already finished and the system was already in a friction regime where the friction was relatively stable and relatively reproducible between different tests, resulting in smaller error bars when averaging 3 long-term tribological tests.

Typical top view pictures of wear scars after the performed wear resistance tests obtained using laser microscopy are shown in Figure 14. At first sight, one may observe that the scars of the benchmark samples are somewhat larger than those of the 3D-textured samples, with the smallest scars for the 3D-textured samples with a 90◦-angle.

For the coated benchmark samples, it is obvious that the antifriction coating was completely depleted in the center of the wear scars on the disc specimen, and thus, that the cylinder came into contact with the base material of the disc specimen, which greatly influenced the friction and wear behavior of the coated benchmark samples. At the periphery of the wear scars on the cylinder, small areas may be observed in which material transfer of the antifriction coating from the disc specimens occurred.

**Figure 14.** Typical wear scars on cylinder and disc specimens after wear resistance tests obtained using laser microscopy.

For the 3D-textured samples, the cylinder surface shows linear grooves produced by the lower 3D-texture for the samples aligned with an angle of 90◦, but no linear grooves were visually observed for the 3D-textured sample aligned at an angle of 45◦, which is simply due to the orientation angle of the 3D-texture with respect to the motion direction of the tribometer during the wear tests, as shown in Figure 15. Actually, linear grooves were also produced on the cylinders for the 3D-texture with an angle of 45◦, but the extremely small distance between each groove renders their visual observation difficult: grooves produced on the cylinders when the angle was 90◦ had a width equal to the plateau side dimensions (~ 37 μm as reported in Table 5), while grooves produced on the cylinders when the angle was 45◦ had a width equal to the plateau diagonals (~52 <sup>μ</sup><sup>m</sup> <sup>=</sup> <sup>√</sup><sup>2</sup> <sup>×</sup> <sup>37</sup> <sup>μ</sup>m).

**Figure 15.** Explanation of (**a**) presence of visible linear grooves (width of grooves = sides of plateaus ~37 μm) on cylinders for the 3D-texture with an angle of 90◦ to the direction of motion and (**b**) near absence of visible linear grooves (width of grooves = diagonals of plateaus = ~52 μm) on cylinders for the 3D-texture with an angle of 45◦ to the direction of motion.

The averages of wear rates (mm3/h) of discs and cylinders measured using laser optical microscopy are shown in Figure 16a and 16b for all 3 sample variations respectively. From both diagrams shown in Figure 16, it may be seen that the use of a 3D microtexture with an orientation of 90◦ to the direction of motion in combination with an antifriction coating may reduce the overall wear rate of the tribological system (upper and lower specimens) from an average of 0.5 mm3/h to approximately 0.25 mm3/h for the discs and from an average of 0.0125 mm3/h to approximately 0.0075 mm3/h for the cylinders. By orienting the 3D microtexture at an angle of 45◦ relative to the direction of motion, the wear rate of the discs stays at a value similar to the benchmark, while the wear rate of the cylinders increased slightly above 0.02 mm3/h.

**Figure 16.** Average wear rate (mm3/h) calculated from laser microscopy measurements after long-term tribological tests (2 h) of: (**a**) discs and (**b**) cylinders.

The increase in wear rates for the 3D-textured and coated samples with an angle of 45◦ in comparison with the similar samples with an angle of 90◦ may be explained by the fact that the contact width of the plateaus having a 45◦-angle is significantly higher (~52 μm) than for a 90◦-angle (~37 μm), as shown schematically in Figure 15, and therefore, the amount of worn material through ploughing is higher for both disc and cylinder specimens. Furthermore, the lower wear resistance of the microtexture with a 45◦-angle (in comparison to the samples with a 90◦-angle) is also believed to be due to the difference in real contact pressures between these two textures. For a defined unit surface area, the number of micropillars encompassed by this area is always about 20–25% lower for the samples with a 45◦-angle than for the samples having a 90◦-angle; thus, the real contact pressures experienced by the 45◦-angle samples is 20–25% higher than the contact pressures encountered by the samples having a 90◦-angle (*p* = *F*N/*A*, where *p* = contact pressure in N/mm2, *F*<sup>N</sup> = load in N and *A* = contact surface in mm2). Table 8 gives a short overview of the formulas used for the calculation of the ratio of the contact pressures for both microtextures having different orientation angles to the direction of motion during long-term tribological tests.

**Table 8.** Calculations of ratio of real contact pressures for both microtextures with different angles to the direction of motion during long-term tribological tests.


For the textured samples with a 45◦-angle, the fact that the wear rate of the disc specimens is similar to the coated benchmark samples (Figure 16a) while the wear rate of the cylinders is significantly higher than the coated benchmark samples (Figure 16b) needs to be analyzed separately in more detail; this will be the focus of a future study and publication.
