Hybrid GRA-PCA and modified weighted TOPSIS coupled with Taguchi for multi-response process parameter optimization in turning AISI 1040 steel

Abstract

The objective of the present study is to optimize multiple process parameters in turning for achieving minimum chip-tool interface temperature, surface roughness and specific cutting energy by using numerical models. The proposed optimization models are offline conventional methods, namely hybrid Taguchi-GRA-PCA and Taguchi integrated modified weighted TOPSIS. For evaluating the effects of input process parameters both models use ANOVA as a supplementary tool. Moreover, simple linear regression analysis has been performed for establishing mathematical relationship between input factors and responses. A total of eighteen experiments have been conducted in dry and cryogenic cooling conditions based on Taguchi L18 orthogonal array. The optimization results achieved by hybrid Taguchi-GRA-PCA and modified weighted TOPSIS manifest that turning at a cutting speed of 144 m/min and a feed rate of 0.16 mm/rev in cryogenic cooling condition optimizes the multi-responses concurrently. The prediction accuracy of the modified weighted TOPSIS method is found better than hybrid Taguchi-GRA-PCA using regression analysis.

Keywords

• grey relational analysis
• principal component analysis
• taguchi method
• analysis of variance
• cryogenic cooling