Estimating and parametrically improving the microstructure, hardness, and wear resistance of SiC-CeO2 reinforcements on hot rolled Al7075 hybrid composites

Abstract

This research employed stir casting to fabricate hybrid aluminum matrix composites (MMC) by mixing different weight proportions of silicon carbide (SiC) with a fixed weight percentage of cerium oxide (CeO₂) and adding it to Al7075 alloy. Hot rolling process was carried out for the developed hybrid composites and mechanical and wear behavior were studied. The effect of wear parameters like applied load (N), sliding distance (m) and wt. % of SiC were studied using statistical approach. The obtained results indicate that, significant improvement was obtained in the grain refinements with minimum porous structures. Similarly, increases of toughness (80–120 KJ/m2), tensile strength (115–136 N/mm2) and hardness (55–71 VHN) with increasing in 0–6 wt. % of SiC reinforcements were obtained. The statistical analysis results indicate that, SiC reinforcements significantly influence the wear resistance of the hybrid composites followed by applied load and sliding distance. Lastly, a feed-forward & backward propagation neural network employing the Levenberg-Marquardt algorithm was used to study COF & wear loss based on three input parameters. For both combinations, the coefficient of correlation was found to be 0.9516 & 0.9956 for training & 0.9907 & 0.9736 for testing, with a confidence interval of 95 %. The mean square error performance achieved was 1.6010^-5 & 1.3210^-5, respectively.