Comparative Optimization of Cutting Parameters in Turning Process

Abstract

While metal cutting is a method that had been practiced for many years, it still holds its importance nowadays. Turning which is one of the common branches of metal cutting, especially the changes in the processing parameters (cutting / chipping depth, progressing speed, rotation speed) are directly effecting the quality of the product. These parameters must be optimized in the manners of applied force, surface roughness as per the measured values. In this study, it is aimed identify the optimized input values (head type, progressing speed, rotation speed and chipping depth) which are minimizing the output parameters of force and surface roughness. In respect to this, genetic and particle swarm optimization algorithms are applied.  Input and output values which are comprised of 240 experimental measurement values are applied Non-Dominated Sorting Genetic Algorithm , Optimal Multi Objective Particle Swarm Optimization and Speed Constrained Multi objective Particle Swarm optimization from the multipurpose optimization techniques and minimum output values tried to be found. In order to obtain the area suitability coefficient in optimization algorithm artificial neural networks technique is applied and weights within the layers are noted. Minimum value had been obtained by the comparison of the obtained optimal values. It had been observed by the conducted analysis that, minimum value for applied force optimal value 0,007329394 reached in SMPSO algorithm, while minimum value for surface roughness reached in NSGA-II algorithm as 0,031769667. Detailed results and comparison of these experimental values are presented in results and discussion topics.