Cutting Conditions Modeling and Optimization in Hard Turning Using RSM, ANN and Desirability Function

Abstract

Mechanical manufacturing companies are required to produce parts with high quality, greater accuracy and high productivity to be competitive. For this purpose, the present work develops predictive models for arithmetic surface finish (Ra), flank wear (VB) and tangential force (Fz). The optimization was based on the desirability function (DF). The machining tests were carried out by hard turning the X210Cr12 hardened steel (56 HRC) using a coated ceramic tool (CC6050), according to the Taguchi L27 experimental plan. ANOVA was employed to determine the influence of cutting parameters (cutting speed—Vc, feed rate—f and machining time—t) on the output parameters (VB, Ra and Fz). Moreover, the RSM and the ANN methods were used to model the technological parameters. The DF approach was used to determine the optimal machining conditions minimizing simultaneously (VB, Ra and Fz). The results show that VB is mainly influenced by Vc (Cont.% = 39.96) followed by f (Cont.% = 35.36). In addition, it was indicated that f and t have been found as dominant factors affecting Ra with contributions of 31.71 and 23.78%, respectively. However, t and f are the main factors affecting Fz with contributions of 75.74 and 22.66%, respectively. On the other hand, ANN and RSM models correlate very well with experimental data. However, ANN approach shows better accuracy and the capability of predicting cutting process parameters than RSM. The optimum machining setting for multi-objective optimization corresponds to Vc= 80 m/min, f = 0.08 mm/rev and t = 4 min.