Multi-parametric Optimization of WEDM Using Artificial Neural Network (ANN)-Based PCA for Al/SiCp MMC

Abstract

Wire electrical discharge machining (WEDM) process optimization is essential when any novel material is discovered. The WEDM process has numerous variables which affect multiple responses. Therefore, multi-response optimization needs to be performed with the help of advanced optimization techniques. Multi-parametric optimization of the WEDM for processing aluminium silicate composite with 15–20% silicate (designated as Al/SiCp) was examined in the present work. A composite principal component was calculated using principal component analysis for multi-parametric optimization. The artificial neural network was employed for enhancing the performance of the process. Analysis of variance was performed to realize the influence of WEDM process parameters on the overall WEDM effectiveness. The WEDM response characteristics such as finish part roughness (Ra), material removal rate and kerf width were considered for this work. From the experimental findings, it is observed that the parameters, viz. the % composition of silicate, the pulse off time (POFF) and current (IP), are the most critical process parameters. The parameters obtained through the present analysis were silicate composition 15%, pulse on time 112 μs, POFF 56 μs, IP-3 A, wire feed rate 4 m/min, wire tension 10 kg and fluid pressure 13 kg/cm3.