Data-driven smart control of injection molding process

Abstract

The most widely used cyclic manufacturing process to produce plastic components, injection molding produces large quantities of scrap materials and consumes resources to mitigate the increasing demands for quality products and sophisticated shapes. Such inefficiencies require a smart quality-control system that can bring consistency in the product quality. Continuous monitoring and control of the injection molding are of excessive importance in providing a better understanding of the occurrence of the failures and their correction measures. In the presented paper, a cyber engineering analysis model is developed to illustrate the real injection molding behavior, and multiple experiments are performed to derive the process window for the product with very good quality. To detect the process disturbances, a control boundary with threshold limits of cavity temperature and pressure is established. If-then rule based control algorithm is derived to detect the failures and make the compensatory adjustments to the process parameters. Finally, the proposed method is illustrated with the manufacturing of an automotive product i.e. car door trim module, and its validity is verified. Study results after the introduction of the developed quality control methodology indicate the dramatic reduction in the quality failure rates and the wastages of resources.