Optimizing parison thickness for extrusion blow molding by hybrid method

Abstract

A hybrid method consisting of finite element method (FEM), artificial neural network (ANN), and genetic algorithm (GA) was used to find the optimal parison thickness distribution for a blow molded part with required thickness distribution. Firstly, numerical simulations on the parison inflation were performed using FEM and the K-BKZ integral type constitutive equation. Based on the simulation results, a back propagation (BP) ANN model was then developed to build the relationship between parison thickness distribution and the objective function, which was used to evaluate the wall thickness distribution of part. The predictive ability of the ANN model was verified through FEM simulation results different from those utilized in the training stage. Finally, a GA was developed and used to search for the optimal parison thickness distribution. The results showed that the hybrid method proposed in this work can effectively obtain the optimal parison thickness distribution for a blow molded part with required wall thickness distribution. Compared with the trial and error method, the hybrid method can shorten the part development time and save a lot of material.