Projekt
Artificial intelligence for simulation in glass manufacturing/Künstliche Intelligenz für Simulationen in der Glasherstellung
Predicting temperatures and residual stresses is essential in the glass tempering process. Classical numerical solvers are accurate but slow in large domains. This study develops a variant of a Multi-Input Fourier Neural Operator (MIFNO) as a fast surrogate, trained on FEM data. It predicts temperatures and stresses f…
Predicting temperatures and residual stresses is essential in the glass tempering process. Classical numerical solvers are accurate but slow in large domains. This study develops a variant of a Multi-Input Fourier Neural Operator (MIFNO) as a fast surrogate, trained on FEM data. It predicts temperatures and stresses from process parameters with strong accuracy and generalization, achieving 7.7× speedup on seen cases and 15× on unseen cases.
Themengebiete
- Industrie 4.0 Industrie 4.0 – Überblick über Forschungsprojekte, Patente und Akteure im TechnologieAtlas.