Micromechanics Transformer for Predicting Material Responses

Micrometer AI framework addresses complex mechanics challenges through operator learning, leveraging advanced architectures like FNO and Transformers. Key contributions include applying operator learning to elastic tensors, developing the CSMBench/CMME dataset, and improving strain field predictions with synthetic data techniques. Performance insights highlight surpassing baseline models and enhancing efficiency in homogenisation modelling, integrated with FEM for multiscale analysis. Future...