Optimization and prediction of aluminium composite wear using Taguchi design and artificial neural network

Abstract

This paper analyses wear behaviour of Al-Si alloy A356 (AlSi7Mg) based composite reinforced with 10 wt. % SiC, and compare it with the base A356 alloy. Composite are obtained using the compocasting procedure. Tribological testing have been conducted on a block-on-disc tribometer with three varying loads (10, 20 and 30 N) and three sliding speeds (0.25, 0.5 and 1 m/s), under dry sliding conditions. Sliding distance of 300 m was constant. The goal of the paper was to optimize the influencing parameters in order to minimize specific wear rate using the Taguchi method. The analysis showed that the sliding speed has the greatest influence on specific wear rate (39.5 %), followed by the load (23.6 %), and the interaction between sliding speed and load (19.4 %). A regression analysis and experiment corroboration was conducted in order to verify the results of the optimization. Specific wear rate prediction was done using artificial neural network (ANN).