Physics-Informed Neural Networks for Shell Element Analysis: Development and Validation - Astraea-Software Co. LTD

Physics-Informed Neural Networks for Shell Element Analysis: Development and Validation

Astraea Software developed a PINNs surrogate model of shell elements using our own AI model structure and verified its performance. Please refer to the PDF file below for the verification results.

Shell elements are elements used when creating FEM models of thin plate structures such as automobile bodies and are frequently used in actual FEM analysis. Since shell elements have rotational degrees of freedom in addition to translational degrees of freedom, the physical equations of PINNs also need to be formulated to include rotational degrees of freedom. The paper reports on the formulation of this physical equation and examples related to disks and rectangles.