An Experimental Analysis of Taguchi-Based Gray Relational Analysis, Weighted Gray Relational Analysis, and Data Envelopment Analysis Ranking Method Multi-Criteria Decision-Making Approaches to Multiple-Quality Characteristic Optimization in the CNC Drilling Process

The goal of this research is to optimize the input parameters utilized in dry CNC drilling of forging steel to attain sustainable machining. Particular emphasis will be placed on achieving high productivity while minimizing the impact on surface quality. To achieve the aforementioned goal, three Taguchi-based multi-criteria decision-making (MCDM) approaches, such as traditional gray relational analysis (GRA), weighted gray relational analysis (WGRA), and data envelopment analysis ranking (DEAR), were used for simultaneous optimization of the MRR and Ra. In Taguchi's L12 (2⁴) orthogonal array design, the cutting mode parameters—such as cutting speed, depth of cut, feed rate, and point angle—have been chosen as the input parameters for the modeling and analysis of the drilling process characteristics. The process of determining the effect of the input parameters on the output parameters was carried out with the use of analysis of variance (ANOVA). The best results from the studies were Ra = 2.19 and MRR = 375 mm³/s, which corresponded to Taguchi's single optimization levels, S₂F₁D₁A₂ and S₂F₂D₂A₁, respectively. In the next step, the performance values obtained for each MCDM technique were reoptimized using the Taguchi method, and the optimal levels were obtained: for traditional GRA, the level S₂F₁D₂A₁ (Ra = 2.52 µm, MRR = 125 mm³/s); for WGRA, the level S₂F₁D₁A₁ (Ra = 2.31 µm, MRR = 83 mm³/s); and for DEAR, the level S₂F₂D₂A₁ (Ra = 4.42 µm, MRR = 375 mm³/s), respectively. Lastly, in order to compare the experiments' performance, validation tests were carried out. The results of the experiments using multi-objective optimization show that traditional GRA improved the overall quality response characteristics by 29.86% compared to the initial setup parameters, while weighted GRA improved them by 34.48%, with the DEAR method providing an improvement of 96%. Based on the findings of this investigation, the DEAR optimization method outperforms the GRA method. As a result, the proposed methods are useful tools for multi-objective optimization of cutting parameters. [ABSTRACT FROM AUTHOR]