An innovative regression model-based searching method for setting the robust injection molding parameters

This work offers an innovative searching method for setting the robust process parameters based on a regression model. This robust method could effectively reduce the influence of environmental noise on part quality in the injection molding process. This method firstly selects the main parameters of affecting part quality as experimental factors. Secondly, a two-level statistically designed experiment coped with the least squared error method is developed to generate a regression model between part quality and process parameters. Based on this mathematic model, the steepest ascent method is used to search for the optimal process parameters, which produces qualified products and resists the interruption of environmental noise. This innovative approach has two advantages: (1) the regression model used in this method is so simple that the numerical computation is fast; (2) since the model is low order, experimental runs are relatively few. This approach thus intends to meet a balance between experimental cost and robustness performance. To verify the performance, light-guided plate (LGP) molding is applied in this work. By minimizing their volumetric shrinkage as the goal, the mold flow simulation program is initially carried out to find the robust process parameters. Furthermore, many experiments are conducted to verify the replication ability of LGPʹs microstructures. Overall, the experimental results demonstrate that this searching method for robust process parameters is practical indeed.