Research on methods for determining heat transfer coefficients in aluminum alloy wheel casting processes

Aluminum alloy casting is critical for manufacturing key components such as A356 automotive wheels, and numerical simulation is essential for predicting casting quality—with thermal boundary conditions being the determinant of simulation accuracy. Conventional thermal boundary modeling methods suffer from low efficiency and poor generalization, failing to meet the precision demands of A356 wheel mass production. The purpose of this study is to optimize the method for determining thermal boundary conditions to enhance the accuracy and efficiency of casting simulations. This work first reviews four mainstream approaches for aluminum alloy casting thermal boundaries: experimental inversion, data-driven surrogates, multi-physics modeling, and multiscale modeling. Taking A356 wheel casting as a case study, we compared interfacial heat transfer coefficient (IHTC) values from different studies and proposed two targeted methods: one is a zone-specific IHTC method that classifies mold-casting interfaces to improve localization accuracy, and the other is a hybrid framework integrating computational fluid dynamics (CFD) and neural network (NN) for evaluating the mold’s external heat transfer coefficient, which enables automated data processing. The main results indicate that compared with uniform IHTC settings, the zone-specific IHTC method reduces the simulation-experiment deviation to within 25 °C; meanwhile, compared with pure CFD simulations, the CFD-NN hybrid framework shortens the average calculation time per task to 0.18 s. This study provides a fast and accurate scheme for determining thermal boundary conditions in casting simulation, and this scheme is also applicable to other similar casting scenarios.