When processing post-consumer recyclates (PCR) considerable differences between the properties of different material batches have to be accounted for. In order to be able to achieve consistently high component quality in injection moulding processes even with such batch fluctuations, the cross-phase process control concepts for injection moulding using modern control strategies developed in the first phase of the project were further developed for the use of recycled plastic granulate.
The underlying phase-combining process control strategy solves the switching problems of conventional injection molding processes for the first time by means of continuous cavity pressure control in the injection and holding pressure phases. Using model predictive control (MPR), a high degree of reproducibility of the process and component quality was demonstrated. This process control concept has now been further developed for the processing of recycled plastic granulate.
© IKVFirst, batch differences were recorded in the test laboratory and in injection molding tests and the interactions between material, process variables and component quality were mapped. No clear correlation was found between the material parameters of the recyclates and the component quality, nor between the process parameters of the recyclate processing and the component quality, as there are interactions with the material composition. However, the key figures of the screw pressure and cavity pressure curve can be monitored in order to detect changes in the injection molding process and thus take batch fluctuations into account in the process control concept.
The alternative approach of direct quality tracking was therefore pursued as part of the project. Bayesian optimisation (BO) was used for this purpose, which does not require explicit modeling of the correlations between cavity pressure reference and component quality. Instead, the interdependencies are mapped indirectly using a data-driven surrogate model. Based on an extended model-predictive cavity pressure control and a superimposed reference optimisation, a tracking of the component quality over different PCR batches and components was successfully demonstrated using Bayesian optimisation. The reproducibility of the process and component quality achieved surpasses conventional plastic injection molding processes.
Project data and funding
We would like to thank the DFG for funding the project (funding code HO 4776/44-2) and the project partners for their cooperation.
Project duration: 05.2022 – 04.2024
Promotion:
