The wear-resistant thin-film sensors and the connection to machine learning processes enable precise monitoring, automated quality assurance and optimization of the injection moulding process.
The multifunctional thin-film sensors are deposited directly on the mold surface. They enable spatially resolved real-time measurement of the temperature in the main stress zones of the injection molding process. One example is the system developed at the Fraunhofer IST, which was applied to a mold insert. With 13 measuring points, it offers an analysis of the flow front. The thermoresistive sensors are arranged in such a way that they map the component geometry. The data is read out and processed in real time. With the help of machine learning algorithms on edge devices, faults and weak points can be detected immediately. The component quality is determined and output as a color signal even before the tool is opened after the injection process. Due to the high wear resistance of the thin-film system, it is also possible to monitor processes with a high proportion of glass fibers and recyclate.