This paper presents an AI-based approach for parameterizing complete vehicle models for crash simulations. The focus is on taking into account manufacturing effects from forming processes such as deep drawing, which significantly influence local material properties and thus crash behavior.

The approach uses machine learning models trained on data from coarse- and fine-resolution forming simulations to predict spatially resolved material parameters. These are automatically integrated into complete vehicle crash models and enable a more realistic representation of structural behavior.

The paper demonstrates how AI-supported coupling of manufacturing and crash simulation can improve the significance of virtual crash analyses and reduce modeling effort.