Machine learning based cooling time prediction in plastic injection molding process

This thesis presents a new approach to develop a machine-learning model for the cooling profile prediction of a planar part produced through the plastic injection molding. Design parameters related to the cooling stage of injection molding, such as distance between the cooling channel and plastic part surface, the distance between cooling channels, the cooling channel diameter, the thickness of the part, as well as plastic material properties including density, mass, thermal conductivity, and specific heat are considered. A wide range of scenarios are created considering the design parameters, ensuring the creation of a comprehensive dataset. To manage this extensive collection of cases, scripts are used to automate designs and simulations. The scripts first generate the required cooling channels for each scenario, then simulate the cooling stage of the injection molding process and collect the relevant results. Then, the physics of the cooling is discussed to estimate the time-dependent temperature of the part in the plastic injection process. An LSTM machine-learning model is used to forecast the simulation results and a regression model is used to predict the cooling profile of the parts. The validation of the developed machine-learning model is done by estimating the cooling time of two industrial parts, produced by plastic injection. One of the parts is cooled with the conformal cooling channels and the estimated cooling time is found with a 2.67% error. On the other hand, the second part is cooled with conventional cooling channels and the estimated cooling time is found with an 11.44% error. In addition, the impact of each design parameter on the cooling time was examined. After the examination, it was noticed that the plastic part should be designed as thin as possible during the design to reduce the cooling time. Also, designing cooling channels as close to the surface of the plastic material as possible was very important to reduce the cooling time.