Enhancing Turbulence Modeling in RANS Simulations for Film Cooling Applications Using Machine Learning Techniques

The research explores the incorporation of machine learning into Reynolds-Averaged Navier-Stokes (RANS) simulations. Acknowledging the constraints of conventional turbulence models, the study seeks to enhance predictions by utilizing information from high-fidelity data obtained from Large Eddy Simulation (LES) simulations. Methods such as correlation analysis are utilized to select crucial input features for training the model, with a focus on improving both accuracy and computational efficiency. Assessment of the refined model against standard RANS simulations and established high-fidelity data showcases notable enhancements in forecasting turbulent heat fluxes and temperature distributions.

کلیدواژه ها:

Turbulence Modeling ، Data-Driven Heat Flux ، Film Cooling