Physics-informed MeshGraphNets (PI-MGNs): Neural finite element solvers for non-stationary and nonlinear simulations on arbitrary meshes.
Tobias WürthNiklas FreymuthClemens ZimmerlingGerhard NeumannLuise KärgerPublished in: CoRR (2024)
Keyphrases
- non stationary
- finite element
- mesh generation
- finite difference
- mesh refinement
- finite element analysis
- autoregressive
- geometric modeling
- soft tissue
- numerical solution
- adaptive algorithms
- gpu accelerated
- empirical mode decomposition
- multiresolution
- d mesh
- boundary element
- material properties
- multiscale
- change point detection
- finite element method
- numerical simulations
- finite element model
- three dimensional
- biomechanical model
- continuous wavelet transform
- shear stress
- computer aided design