Optimizing injection molding simulations: comparative performance of Kriging and RSM surrogate models for process efficiency

Abstract This contribution investigates the application of surrogate modeling, specifically Kriging and Response Surface Methodology (RSM), in optimizing injection molding simulation processes. Injection molding simulations are computationally intensive due to the multi-physical interactions involved, making surrogate models an attractive alternative to reduce computational effort. The authors employ Kriging and RSM models to simulate the injection molding of three different geometrical configurations, assessing their performance in predicting parameters for process quality/efficiency like deformation, shrinkage, and cycle time. The study makes use of MATLABs fmincon optimization algorithm with the models, emphasizing cycle time minimization while maintaining deformation and shrinkage within acceptable limits. In addition to the prediction accuracy, this contribution demonstrates that both surrogate models reduce the computational cost per evaluation by several orders of magnitude compared to full injection molding simulations and provide fast, iterative optimization. Findings indicate that Kriging outperforms RSM, especially in complex geometries, by providing more accurate predictions with lower error rates. The validation against digital twin simulations supports the effectiveness of Kriging, making it preferable for applications requiring high precision in process optimization.