A physics-constrained multi-objective CFD-driven model training framework and its application to SWTBLI flows

Denggao Tang

doi:10.1016/j.ast.2025.110750

A physics-constrained multi-objective CFD-driven model training framework and its application to SWTBLI flows

2025-08-12

0

PRE

AI

DOI：10.1016/j.ast.2025.110750

Original

Original request for help

Save

Abstract

En 中文

• A physics-constrained, multi-objective CFD-driven training framework is proposed. • Symbolic corrections are trained with CFD solver feedback for model-consistent learning. • Inversion fields are used to guide correction domain and improve interpretability. • The SST-MO model outperforms classical FISR in accuracy across SWTBLI test cases. • The method maintains better interpretability than single-objective CFD-driven models.

Keywords:

Turbulence model

SWTBLI

Gene expression programming

Field inversion

Model-consistent training

A physics-constrained multi-objective CFD-driven model training framework and its application to SWTBLI flows

Abstract

Journal

Organization