Arabian Journal for Science and Engineering (21914281)50(4)pp. 2663-2689
Parametric study of ductile material forming processes for optimization of damage that occurs in the production stages of the parts will lead to better quality, performance, durability, and reduced production costs. Additionally, implementing an appropriate criterion that accurately predicts and simulates damage growth in the production of a part is an essential step in the optimization process. This research investigates the optimization and minimization of the final damage in a gudgeon pin produced by a two-stage cold extrusion process. First, numerical simulations are performed using the modified Lemaitre’s damage criterion, which provides an accurate estimate of damage growth in combined tensile and compressive loads. In the following, two methods for optimizing the parameters of each extrusion cold die are examined, and the magnitude of the final damage in the product is compared. In the first method, the design of the Taguchi experiment is implemented. In the second method, by implementing and training an artificial neural network and transferring the results to the genetic algorithm, the optimization of the die parameters is investigated to minimize the final damage in the product. A comparison of the results in the two mentioned methods shows that the final damage in the product is less than the critical damage value in both methods. Also, the neural network method will lead to less final damage in the product in comparison with the design method of the Taguchi experiment. © King Fahd University of Petroleum & Minerals 2024.
