Experimental Investigations and Optimization of Surface Roughness Using Response Surface Methodology Coupled with Genetic Algorithm and Particle Swarm Optimization Techniques in Grinding of Inconel 718

Abstract

Nickel-based superalloy such as Inconel 718 has worldwide applications in the manufacturing of aircraft components and defence industries due to superior properties at elevated temperatures. The importance of this material in high-temperature applications due to its excellent thermo-physical properties is subject to extensive area of interest. The machining of nickel-based superalloy is a challenging task due to generation of high heat in grinding zone which impels the study of improvement of surface quality in the present study. The main aim of the present study is to find the optimum process parameters corresponding to minimum R_a value using different optimization techniques so that the production cost of components and time consumption can be minimized. In the present experimental study, investigation has been carried out on Inconel 718 through a CNC surface grinding machine. Due to the complexity involved in tough-to-machine material, the study focuses on the improvement of surface roughness using down grinding process by the optimization of three influential parameters such as wheel speed, depth of cut and table speed. Response surface methodology based central composite rotatable design is used in this study to illustrate the surface roughness value (R_a) which is greatly influenced by wheel speed followed by depth of cut and table speed. For the optimization of machining parameters, RSM coupled with genetic algorithm (GA) and particle swarm optimization (PSO) is used to reduce the time consumption in the selection of machining parameters and desirous output response in grinding. The R_a value corresponding to GA is found to be 0.2735 µm while 0.2586 µm using PSO technique. The best optimal process parameters corresponding to minimum R_a value using PSO technique are depth of cut = 5 µm, wheel speed = 628 m/min, and table speed = 3588 mm/min. Comparatively, PSO provided better results in terms of minimum surface roughness than GA. The validation of experimental results is done with a statistical model that has shown a fine level of corroboration among them.