Optimization of process parameters in machining of Inconel 718 super alloy on HMT (NH22) capstan lathe using genetic algorithm subject to minimization of surface roughness

Abstract

Inconel 718 has been machined on HMT (NH22) capstan lathe following central composite design of experiment to develop the response surface model. The surface roughness has been predicted using a first order regression model by varying the cutting velocity, feed and depth of cut. The student’s t test and p-values have revealed that the depth of cut as well as combined effects of feed and depth of cut are significant process parameters. The second order response surface model has been developed to account the effect of curvature in predicting the surface roughness. The response surface plots reflect that the roughness of the Inconel workpiece varies largely with the variation in depth of cut and combination of feed and depth of cut. As a test of reliability Fisher’s F-test has been conducted. The F-test revels that the response surface model converges with the observed data reliably. A python code has been generated for genetic algorithm to optimise the process parameters by minimising the surface roughness. The results of the genetic algorithm have converged with the results of the confirmation run.