Optimization of Flash, Billet Dimensions and Friction Factor in Closed Die Cold Forging Process

Three important parameters in designing a closed die for forging process are ratio of width to flash thickness, ratio of  height to billet diameter and the friction factor. In this paper the influences of these parameters on the required force for the forging and percentage of die filling were investigated. It was found that by controlling the flash dimension, the material loss is reduced and the percentage of die filling is increased. Also, an experimental model was simulated and analyzed by finite element method. To validate the numerical results obtained by this research, value of gained force from finite element method was compared with the obtained experimental results. In order to coordinate and connect between the mentioned parameters and obtain a performance function, a two layer neural network was used. Finally, by using neural network and genetic algorithm, the optimum sets of parameters which minimized the force and maximized the percentage of die filling were found. These values were compared with the experimental results of other researchers. The genetic algorithm has good correlation with the experimental method as well as it has presented acceptable estimation for effective parameters in the forging process.