Quality Control in Injection Molding based on Norm-optimal Iterative Learning Cavity Pressure Control
Sebastian Stemmler, Marko Vukovic, Muzaffer Ay, Julian Heinisch, Yannik Lockner, Dirk Abel, Christian Hopmann
Abstract
Plastic injection molding is characterized by high design flexibility of the manufactured parts. Consequently, it is one of the most important processes for mass production of plastic parts. The setup of the manufacturing process is very complex due to numerous impact factors. In addition, material fluctuations or changing ambient conditions require the adaption of the setup during manufacturing to guarantee a constant product quality. In order to reduce the setup effort and to control the quality, the concept of model-based self-optimization is applied to injection molding. Therefore, a model-based Norm-Optimal Iterative Learning Controller (NOILC) is used to track a desired reference for the cavity pressure during the entire cycle. This reference is determined by the so-called pvT-optimization which considers the cooling behavior of the melt within the cavity. It is shown by experiments that the cavity pressure can be controlled with high accuracy using the presented NOILC. Furthermore, the accuracy of the quality, especially the part weight is improved by combining the NOILC with an additional pvT-optimization.