Artificial Neural Network and Taguchi Analysis of Multi-Objective Optimisation of Wear Behaviour of Zro2 based Aluminium Nanocomposite

Abstract

Metral matrix composites are frequently utilized to replace single, unified materials in the aerospace, manufacturing, and defence industries due to their higher qualities for example strength properties through low weight, significantly greater toughness, excellent wear resistance, and improved head conductance characteristics. The goal of this innovative research was to determine the specific wear rate (SPR) and coefficient of friction (CFR) of zirconium dioxide-filled Al 8014 (Al-Mn alloy) matrix composites ZrO2-). To identify the finest array of process parameters for SWR and CFR of recommended composites, the Taguchi method was applied. The stir casting procedure were used to make composite samples with varied ZrO2 particle additions (5, 10, and 15% wt.). The wear tests were carried out in dry conditions using a pin-on-disk device in accordance with the L27 orthogonal design. The following four control variables were selected at three levels for this test: ZrO2 weight percent, load, disc velocity, and sliding distance. According to the experimental data, the created composite sample has a minimum SWR of 15 weight percent ZrO2, a weight of 29.43 N, a velocity of disc of 0.94 m/s, and a sliding distance of 1000 m. According to the ANOVA results, the weight percentage of ZrO2 content had the second most significant impact on the SWR and CFR, following the load. Neural network model is developed to predict the responses. The model predicts the result with an accuracy of 99.78%.