Warpage optimization of the GFRP injection molding process parameters

Warpage deformation is a common defect in the glass fiber-reinforced plastic (GFRP) injection molding process. In this study, a case study of the GFRP product is proposed. Firstly, with the minimum warpage deformation as the optimization objective, three significant factors affecting product quality (the first stage packing pressure, injection time, and melt temperature) are selected from nine process parameters by using Plackett Burman. Then, the three selected significant factors are considered as design variables, and Box-Behnken design (BBD) is used to design the experiment. Warpage analysis is performed based on Moldflow simulation software. Secondly, on the basis of the BBD experimental samples and results, a regression model by compound optimization, back-propagation neural network based on adaptive boosting and genetic algorithm (AdaBoost-GA-BP) model is established. To further illustrate the prediction accuracy of the model, response surface methodology model, back propagation neural network (BPNN) model, back-propagation neural network based on genetic algorithm (GA-BPNN) model are taken as comparative algorithms. The results show that the prediction system of the AdaBoost-GA-BP model has good stability and accuracy. Finally, the particle swarm optimization approach is used to search for the minimum warpage. It can be concluded that the quality of the plastic part after optimization has been improved.