Optimization of wear parameters using Taguchi grey relational analysis and ANN-TLBO algorithm for silicon nitride filled AA6063 matrix composites

Abstract

The current research work aims to synthesis AA6063-Si3N4 composites by stir casting technique and analyzing their wear behavior using optimization techniques. The selected composition of the proposed composites are AA6063, AA6063-5 wt% Si3N4 and AA6063-10 wt% Si3N4. The Taguchi grey relational analysis (TGA) was used to analyze the wear performance of the synthesized composites. Further, using the measured experimental data, an artificial neural network (ANN) model was developed to predict the wear rate and teaching learning based optimization (TLBO) algorithm was implemented to obtain the optimized parameters value with the objective of minimizing the wear rate. In the existing work, the multiple linear regression model (MLRM) was developed using the measured values of both dependent (response) and independent (process parameter) variables and these models are used in the optimization technique to optimize the process parameters. From the TGA technique a minimum wear rate of 0.0002 mm3 min−1 was obtained for 10% wt reinforcement, 29.43N load, 3 m s−1 is sliding velocity and 1500 mm sliding distance and from the ANN model and TLBO algorithm, it is probable to attain 0.000183 mm3 min−1 wear rate for the process parameter values of 9.892% of wt reinforcement, 9.837N of load, 2.936 m s−1 sliding velocity and 1130.7 mm sliding distance. As compared to TGA method, an amount of 8.4% improvement in wear rate is achieved by implementing ANN–TLBO method.