Generalization of machine-learned turbulent heat flux models applied to film cooling flows
This work addresses a domain-specific issue in engineering design for film cooling systems, but it appears incremental as it builds on existing machine learning approaches without introducing a new paradigm.
The paper tackled the problem of inaccurate turbulent heat flux models in film cooling flows by training machine learning models to predict a non-uniform turbulent Prandtl number field, exploring their generalization ability across different datasets.
The design of film cooling systems relies heavily on Reynolds-Averaged Navier-Stokes (RANS) simulations, which solve for mean quantities and model all turbulent scales. Most turbulent heat flux models, which are based on isotropic diffusion with a fixed turbulent Prandtl number ($Pr_t$), fail to accurately predict heat transfer in film cooling flows. In the present work, machine learning models are trained to predict a non-uniform $Pr_t$ field, using various datasets as training sets. The ability of these models to generalize beyond the flows on which they were trained is explored. Furthermore, visualization techniques are employed to compare distinct datasets and to help explain the cross-validation results.