MR-GVNO: A Geometry-Aware Variational Physics-Informed Neural Operator for Mindlin-Reissner Plates on Irregular Domains

The study introduces MR-GVNO, a geometry-aware variational neural operator designed for predicting the behavior of Mindlin-Reissner plates on irregular domains. It utilizes boundary point clouds to represent complex geometries and employs a cross-attention mechanism for integrating spatially varying material properties and loads, enabling accurate predictions of transverse deflections and rotations without labeled data. This approach significantly reduces computational costs associated with traditional finite element methods and demonstrates rapid inference times, making it a valuable tool for practitioners dealing with complex plate structures in engineering applications.