Adaptive Digital Twin for Crack-Free Powder Compaction: From Green State to Fatigue Performance (GeDiRL), EFRE1134, Progr.FESR 2021-2027 + Prediction and prevention of cracks in powder metallurgy (Funds for new projects development Peer)

In Powder metallurgy the transformation of metal powder into solid components using pressure andheat requires precise control over material behaviour and hydraulic actuation since internal cracks and density inhomogeneities can develop. Based on recent research conducted by UNIBZ in collaboration with GKN Sinter Metals a prototype of an industrially implementable methodology for the prediction and prevention of crack risk should be developed. The Research Contract will contribute to this goal by consolidating a previously developed physics-based digital twin, extending it to cover all four stages of the compaction cycle and calibrating it against industrial production signals, and adopting it for the training of an explainable reinforcement learning algorithm capable of minimising crack risk... (see call)