Adaptive hybrid optimization of hydrodynamic deep drawing with radial pressure process by combination of parametric design and simulated annealing techniques

Abstract

An adaptive hybrid simulated annealing technique with ANSYS parametric design language is developed to optimize hydrodynamic deep drawing assisted by radial pressure process. This work aims to determine an optimal pressure path by redefinition of simulated annealing parameters and creating an adaptive finite element code using ANSYS parametric design language for any cylindrical, conical, and conical–cylindrical cups. The simulated annealing algorithm is developed adaptively with respect to hydrodynamic deep drawing with radial pressure process to link with ANSYS parametric design language code using a script in MATLAB. Parametric definition of process parameters enables the optimization algorithm to change the finite element model configuration in each iteration. Defective product is detected by definition of two failure criteria based on thinning and wrinkling occurrence during the optimization process. The proposed optimization method is employed in fractional factorial design of experiment to investigate the effective parameters on final product quality. Also, a regression model is derived to predict the final product quality based on the maximum thinning percentage under the optimal pressure path. Reliability of the optimization procedure and regression model is validated by experiments.