Modeling and optimization of hard turning: predictive analysis of surface roughness and cutting forces in AISI 52100 steel using machine learning

This study addresses the critical need for high-strength, corrosion-resistant materials in renewable energy, biomedical and maritime applications, necessitating effective heat treatment processes. Focusing on AISI 52100 steel, the research employs finish hard turning with coated carbide inserts under dry cutting conditions. Five machine learning methods are applied to model surface roughness (Ra) and cutting forces (Fx, Fy, Fz) using a Taguchi L36 orthogonal array. Results indicate SVM, XGB, DT, and XGB are superior algorithms for Ra, Fx, Fy, and Fz prediction. Key findings highlight feed rate predominant influence (96.55%) on surface roughness, while depth of cut significantly affects cutting forces. Optimal cutting parameters, 0.1 mm depth of cut, 0.15 mm/rev feed rate, 160 m/min cutting speed, 0.4 mm nose radius, and 40.9 HRC hardness are identified via response surface methodology (RSM) and desirability function. The study underscores the importance of optimizing cutting parameters to enhance surface quality and machining efficiency in challenging material processing scenarios.