Managing distortion in high pressure die casting: Evaluating a digital twin approach for real-time adjustment of quenching parameters

High pressure die casting (HPDC) is prone to distortion especially in the case of components featuring large area and low wall thickness. Counteracting these tendencies is either costly (mechanical straightening) or risky for lack of flexibility (adapting die cavity geometry). The alternative is introducing a deliberately inhomogeneous cooling after casting or solution heat treatment (SHT). In order to reduce distortion in this way, local heat extraction must be adapted to the state of the part immediately before quenching. Assuming a HPDC process chain including SHT, a digital twin of the process chain up to the final stages of SHT is foreseen to provide the input data for derivation of the required local cooling conditions. Experimental validation is carried out using Bühler SC/N 66 HPDC equipment. Actual distortion of parts produced at different parameter settings is measured using optical 3D scanning before and after SHT. Both experimental and simulation data is utilized to train a data-driven model linking process parameters to part properties – namely distortion and residual stress -, as latency of physics-based simulation is not acceptable. Practical implementation of heterogeneous cooling is envisaged via a spraying field composed of individually controlled spraying nozzles. Local heat transfer coefficients (HTC) delivered by these are adapted based on models of spraying characteristics previously parameterized experimentally. Finite Element Analysis (FEA) of position-dependent effects of individual nozzles on residual stress and distortion is used to establish suitable combinations of nozzles at different positions as well as parameter settings capable of alleviating a specific deformation or stress pattern.