Qualitative enhancement in machining efficiency of SNCM8 alloy through hybrid ann-taguchi optimization approach

The present study optimizes hard-to-machine materials using hybrid modeling involving artificial neural networks (ANN) and the Taguchi method. The main objective of this work is to reduce tool wear and improve the material removal rate (MRR) along with lowering surface roughness (SR) in the wire electrical discharge machining (WEDM) of SNCM8 alloy steel. The model combines ANN’s predictive capacity with Taguchi’s robustness to forecast machining outcomes as process factors are combined. For this research, an L27 OA is adapted for experimentation; independent variables include current (5 A, 10 A, 15 A), pulse duration (30 µs, 60 µs, 90 µs), and feed rate (FR) (2 mm/min, 4 mm/min, 6 mm/min). The investigated output metrics are MRR, SR, and dimensional accuracy. From the analysis, it is possible to increase the MRR by 20%, from an average of 1.0 g/min to 1.2 g/min, and reduce SR by 15%, from 2.0 µm to 1.7 µm. In addition, the dimensional deviation (DD) was reduced to a minimum of 18%, which reduced from 0.11 mm to 0.09 mm. ANOVA data analysis showed pulse duration and current as the most relevant factors affecting machining performance, accounting for 45 and 35% of the variance. The hybrid model predicted and optimized machining reactions; the ANN predictions were closely aligned with experimental values, with an R-squared value exceeding 0.95. Optimizing parameter settings increased machining efficiency, reduced tool wear by 25%, and improved surface quality, revealing sustainable production techniques.