Optimization of Filling Time and Volumetric Shrinkage Rate in Simulation of Plastic Product Injection Molding Process Using RSM and NSGA-II

Abstract

Filling time (Td) and volumetric shrinkage rate (Vs) are critical indicators in plastic injection molding. However, these two indicators often conflict, making it essential to optimize both simultaneously. This study aimed to optimize the td and Vs in simulating a plastic electrical socket product injection molding process. Four process parameters were investigated: melt temperature (Tnc), mold temperature (Tk), injection pressure (Ap), and pressure holding time (Ta). The Box-Behnken design method was used to determine the number of simulation samples, and the Response Surface Methodology (RSM) was employed to develop predictive models for td and Vs. The Non-dominated Sorting Genetic Algorithm (NSGA-II) technique was then applied for multi-objective optimization. The results showed that the RSM-based regression models for Td and Vs had high coefficients of determination (R2) of 0.946 and 0.990, respectively, indicating the significance of the developed models. The NSGA-II optimization generated 21 Pareto solutions, with Tnc ranging from 215.0 to 215.4 °C, Tk from 50 to 60 °C, Ap from 67.09 to 76.53 MPa, and ta from 2.097 to 2.5 s. These parameter values corresponded to td values from 1.260 to 1.389 s and Vs values from 4.810 to 5.497%. To verify the accuracy of the optimization method, the solutions with the smallest td and Vs values were selected for re-simulation, and the results showed a difference of less than 2.03% between the predicted and re-simulated values. These results confirmed the effectiveness of the proposed approach in solving the multi-objective optimization problem for the plastic injection molding process, thereby enhancing both the efficiency and quality of the products.