Online adaptive injection molding process and quality control

An online adaptive control methodology was developed to control the process and quality variables in injection molding. A process variable, such as the cavity pressure or mold separation, was selected for the process control, which adaptively changed the fill-to-pack switchover point from shot to shot, and adjusted the hydraulic pressure during the pack/hold phase within one shot. Further, a more sophisticated online adaptive quality control scheme was investigated that incorporated direct online quality control and neural network models in the feedforward loop. The neural network models were extracted from computer-aided engineering (CAE) predictions. Incorporating CAE results in the online quality control fills the gap between a science-based product/process design and an empirics-based process/quality control on the shop floor. This methodology can also be applied to monitor and control other quality variables such as part dimensions.