Genetic Algorithm Optimization for Enhanced Turning Processes in Steel Rolling Mills

Abstract: Optimizing turning processes in steel rolling mills is critical for enhancing productivity and reducing operational costs. This study proposes a genetic algorithm (GA)-based multi-objective optimization framework to minimize machining time (Mt) and maximize tool life (Tl) during the dry turning of hardened cast iron rolls (50-55 SHC) using tungsten carbide inserts (RCMX25). A Waldrich Seizen CNC lathe (90 kW) was employed to conduct experiments under varying spindle speeds (10-18 rpm), feed rates (1.2-1.6 mm/rev), and depths of cut (8-10 mm). Regression models derived from an L9 orthogonal array quantified the impact of parameters on Mt (RMSE = 5.26) and Tl (RMSE = 2.95). The GA optimized these conflicting objectives, achieving a 15.8% reduction in machining time and a 22.3% improvement in tool life compared to baseline Taguchi methods. Results demonstrate that GA effectively balances trade-offs between productivity and tool longevity, offering a data-driven solution for industrial CNC machining. This work bridges the gap between theoretical optimization and real-world implementation, providing actionable insights for steel rolling mills.

Keywords: Genetic Algorithm (GA), Multi-objective optimization, Steel rolling mills, Machining time (Mt), Tool life (Tl), Dry turning, Hardened cast iron rolls, Tungsten carbide inserts, Spindle speed, Feed rate, Depth of cut, L9 orthogonal array, Regression models