Combining the level-set based φ-FEM discretization with a Fourier Neural Operator yields a fast surrogate PDE solver for varying geometries, with accuracy comparable to finite elements on the tested cases.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.NA 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Phi-FEM-FNO: a new approach to train a Neural Operator as a fast PDE solver for variable geometries
Combining the level-set based φ-FEM discretization with a Fourier Neural Operator yields a fast surrogate PDE solver for varying geometries, with accuracy comparable to finite elements on the tested cases.