Response Surface Methodology Integrated with Desirability Function and Genetic Algorithm Approach for the Optimization of CNC Machining Parameters

In this study, response surface method (RSM), desirability function (DF) and genetic algorithm (GA) techniques were integrated to estimate optimal machining parameters that lead to minimum surface roughness value of beech (Fagus orientalis Lipsky) species. Design of experiment was used to determine the effect of computer numerical control machining parameters such as spindle speed, feed rate, tool radius and depth of cut on arithmetic average roughness (Ra\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R_{\mathrm{a}}$$\end{document}). Average surface roughness values of the samples were measured by employing a stylus type equipment. The second-order mathematical model was developed by using response surface methodology with experimental design results. Optimum machining condition for minimizing the surface roughness was carried out in three stages. Firstly, the DF was used to optimize the mathematical model. Secondly, the results obtained from the desirability function were selected as the initial point for the GA. Finally, the optimum parameter values were obtained by using genetic algorithm. Experimental results showed that the proposed approach presented an efficient methodology for minimizing the surface roughness.