Multi-objective optimization of turning Ti-6Al-4V titanium alloy using design of experiments, artificial neural networks and genetic algorithms

This paper presents an experimental investigation into the machinability of Ti-6Al-4V titanium alloy. It focuses on the influence of cutting parameters on quality and productivity of machined parts. A Taguchi L27 orthogonal array combined with analysis of variance is used to evaluate the significance of machining parameters and their interactions. Multi-objective optimization based on desirability function (DF) is performed to determine the optimal cutting parameters. Moreover, a hybrid artificial neural networks-genetic algorithms (ANN-GA) model based on Taguchi L27 and coupled to DF is adopted to predict and optimize cutting responses. ANN-GA-DF exhibits superior performance, achieving significant improvements in surface roughness (39% in Ra, 19% in Rq, 24% in Rz), dimensional and geometric accuracy (25% and 39%, respectively). Material removal rate decines slightly (5.3%). Optimal parameters are tool radius of 0.4mm, cutting speed of 25 m/min, feed rate of 0.065 mm/rev, and depth of cut of 0.7 mm.