„Lebendige“ FE Superelemente mit Machine Learning

Werner Moretti's presentation focused on integrating Machine Learning (ML) with Finite Element (FE) Superelements, a concept he referred to as ‘Living’ FE Superelements. The talk highlighted the synergy between ML and Computer-Aided Engineering (CAE), emphasizing how ML can enhance CAE processes like design of experiments (DOE) and reduced-order modeling (ROM). A significant part of the presentation was dedicated to the basics and applications of FE Superelements. These are used to reduce the complexity of FE models by dividing them into substructures, each represented by reduced matrices. The introduction of Smart ‘Living’ Superelements (SSE) marked a pivotal point in the discussion. SSEs incorporate ML to embed design variations directly into the system analysis, thus enabling dynamic and adaptable FE analyses. Moretti showcased practical applications of SSE in fields like aircraft structure analysis. These examples demonstrated how SSEs could handle parametric variations, leading to more efficient and adaptable simulations at the system level. The integration of ML with SSEs allows for real-time predictions and enhances the flexibility of the models.