Finite element simulation-based predictive regression modeling and optimum solution for grain size in machining of Ti6Al4V alloy: Influence of tool geometry and cutting conditions

Abstract

The present research study was aimed at studying the impact of machining parameters on grain size alterations in machining of Ti6Al4V alloy using finite element analysis (FEA) based on the Johnson-Mehl-Avrami-Kolmogorov (JMAK) recrystallization model. The machining parameters consisted of both cutting conditions and tool geometry including cutting speed, feed rate, tool edge radius and rake angle. Three series of JMAK constants were used in the simulations and the degree of accuracy of the obtained results were compared. Finite element (FE) simulations were conducted for machining conditions designed using a D-optimal Design of Experiment (DoE). Analysis of Variance (ANOVA) was carried out to identify the most effective machining parameters on the responses of average grain size (AGS), machining temperature (MT), cutting force (CF), and feed force (FF). Next, Response Surface Method (RSM) was used to predict regression models of AGS, MT, CF, and FF. The optimal values of machining parameters were obtained to improve AGS, MT, CF, and FF. The ANOVA results showed that rake angle and cutting speed were, respectively, the most and least significant parameters affecting AGS in the design space. Rake angle was the most effective parameter influencing all of the responses. The single-criterion optimization of AGS provided a small improvement in AGS with a high increase in material removal rate (MRR). The results of the multi-criteria optimization of AGS, MT, CF, and FF demonstrated that machining of Ti6Al4V alloy with a tool having a medium cutting edge radius and high positive rake angle at a medium cutting speed and a low feed rate in the design space produced finer grains along with reduced CF and FF and an unchanged MT.