Multi-objective optimization of CNC Milling parameters for Mild steel using an integrated Taguchi-GRA-PCA Method

Abstract

This thesis presents a multi-objective optimization framework for CNC milling of St52 grade mild steel, aiming to simultaneously enhance surface roughness (Ra), material removal rate (MRR), and cutting force (Fc). These three responses, representing surface quality, productivity, and process stability, are inherently conflicting, as improvements in one often degrade the others. While previous studies have often optimized these responses individually or combined Ra with MRR using subjective weighting, this research introduces an integrated framework combining Taguchi design, GRA, and PCA to achieve simultaneous enhancement with objective weighting. The Taguchi method with an L16 orthogonal array provides an efficient experimental design for investigating the effects of spindle speed (N), feed rate (F), and depth of cut (Dc) at four levels each. Single-objective analysis using S/N ratios revealed that N primarily influences Ra, while F has the greatest impact on MRR and Fc, with ANOVA confirming the statistical significance of these parameters. For multi-objective optimization, GRA converts the three responses into a single GRG, and PCA-derived objective weights (Ra= 0.35, MRR= 0.33, and Fc= 0.32), quantifying each response's contribution to overall performance without subjective bias. The optimal parameter combination identified was at 1500 rpm N, 135 mm/min F, and 1.5 mm Dc. Confirmation experiments under these parameters yielded Ra of 2.320 μm, MRR of 1932.44 mm³/min, and Fc of 617.57 N. Comparison with single-response optimal confirms simultaneous enhancement (MRR improved by 27.9% over Ra optimal settings, Fc reduced by 28.5% over MRR optimal settings, and MRR improved by 51.0% over Fc optimal settings. These results validate the framework's effectiveness in resolving conflicts between responses. Confirmation tests validate the optimized parameter combinations with prediction errors below acceptable limits (10%), which demonstrates the robustness of the methodology. The findings offer opportunities for manufacturing industries to enhance machining performance in CNC milling operations. The methodology can be extended to other materials and machining processes, contributing to more efficient and competitive manufacturing practices.