Issue 77

Y. C. Arun et alii, Fracture and Structural Integrity, 77 (2026) 316-339; DOI: 10.3221/IGF-ESIS.77.19

ANOVA for wear loss Tab. 8 displays the ANOVA results assessing the statistical relevance of process parameters on wear loss. A higher F value denotes a greater influence on the response variable since it shows the ratio of the variation explained by each factor to the unexplained experimental error. A significant effect at a 95% confidence level is indicated by a p-value of less than 0.05. Sliding velocity (SV) and SiC abrasive particle size had the greatest F-values (13.40 and 12.40, respectively) and very low p- values (≤ 0.001) among the examined parameters, indicating that they are the most statistically significant factors controlling wear loss. This suggests that material removal methods are dominated by both abrasive severity, which is regulated by SiC particle size, and frictional heating effects, which are linked to SV. Although its relative contribution is smaller than that of SV and SiC particle size, load also has a statistically significant impact on wear loss (F = 3.95, p = 0.040 < 0.05). Conversely, axial distance (AD) (F = 0.23, p = 0.797) and filler content (F = 0.30, p = 0.747) show strong p-values (> 0.05), suggesting that their effects are statistically negligible within the examined range. Good experimental reproducibility is indicated by the error term remaining small. Additionally, the lack-of-fit is statistically non-significant (p = 0.299 > 0.05), indicating that the generated model accurately and significantly depicts the experimental data. As a whole, the findings show that sliding velocity and SiC particle size are the main factors influencing wear loss, followed by load; under the conditions under investigation, AD and filler content have very little statistical impact. . Source DF Adj SS Adj MS F-Value P-Value Load (N) 2 0.000517 0.000259 3.95 0.040 SV (m/s) 2 0.001756 0.000878 13.40 0.000 AD (m) 2 0.000030 0.000015 0.23 0.797 SiC Paper (μm) 2 0.001624 0.000812 12.40 0.001 Filler (wt%) 2 0.000039 0.000019 0.30 0.747 Error 16 0.001048 0.000065 Lack-of-Fit 15 0.001037 0.000069 6.54 0.299 Pure Error 1 0.000011 0.000011 Total 26 0.004755 Table 8: ANOVA results for main effect on wear loss Regression equation and residual plots for wear loss The association between wear loss and the input parameters is established by the created quadratic regression model (Eqn. 1), suggesting that wear is positively influenced by all factors. According to the coefficients, sliding velocity (SV) has the biggest impact (0.1022), followed by SiC paper grit and filler content. The impacts of load and abrading distance (AD) are comparatively smaller. This suggests that whereas AD has no effect, an increase in SV greatly speeds up wear because of increased frictional energy and thermal effects. The model has a fair predictive capability, as evidenced by the coefficient of determination (R 2 = 73.49%), which shows that it explains a significant amount of the variability in wear loss. The model is statistically appropriate without undue overfitting, as confirmed by the somewhat lower but still acceptable adjusted R² (67.18%), which takes into consideration the number of predictors. The residual diagnostics used to validate the regression model are displayed in Fig. 9. The normality of the residuals is confirmed by the normal probability plot, which shows a near-linear trend. Random scatter is seen in the residuals vs. fitted values, showing no systematic error and constant variance. In a similar vein, there is no trend in the residuals vs. observation order, indicating independence. For estimating wear loss, the model is often dependable and statistically sufficient. Main effect plots for coefficient of friction The main effects of process variables on the coefficient of friction (CoF) are shown in Fig. 10. With increasing load, a significant declining trend is seen, suggesting better surface conformance and less interfacial shear at higher loads. Sliding velocity only slightly varies, and at higher velocities, CoF somewhat increases, probably because of frictional heating effects. With slight variations in CoF, AD shows a weak, non-linear contribution, indicating a limited impact within the examined range. In a similar vein, SiC particle size exhibits a small effect, with finer abrasives modestly increasing CoF because of increased asperity interaction. ( ) ( ) ( ) ( ) SiC paper m µ ( ) 0.0575 0.00095 0.1022 0.000052 0.00045 0.00364 % m       s Wear loss g Load N SV AD m Filler wt =− + + + + + + (4)

327