A new approach to dynamic forecasting of cavity pressure and temperature throughout the injection molding process

Abstract The injection molding process is very sensitive to ordinary environmental alterations, as the numerical simulation is limited to within one injection cycle, and it cannot predict transient regimes. The present study presents a new approach based on SARIMAX models developed to predict the temperature and pressure inside the mold cavity. The proposed approach was developed in Python language, and it can identify the behavior of the process, allowing preventive actions. Experimental data of temperature and pressure obtained in real‐time inside an injection mold were accessed to use and to validate the proposed model. The results showed its efficiency and its high accuracy for predicting variations in temperature and pressure inside the mold, even when using a small number of samples to be trained. The proposed model can be very useful for monitoring the production of mechanical parts, under an Industry 4.0 environment. For future works, the model enables a contribution toward digital twins of a molded part, considering all the alteration on the parts' properties due to the disturbance on the injection molding process. Furthermore, it lays the groundwork for a new injection machine control system architecture.