The aim of this project is to be able to precisely design thermally loaded parts in gas turbines with the help of 3D computational fluid dynamics. Accurate knowledge of the flow situation (laminar-turbulent transition, boundary layer status) is crucial to quantify the heat that is transported by the flow. In this project higher order anisotropy resolving turbulence models are applied in combination with models for the turbulent heat transport, for the first time in the TRACE solver. Calibration of the numerical models for turbulent heat transport should be done with measurement data. Combined with the prediction tools for heat conduction in solid materials, a precise evaluation of thermal loads of turbines will be possible in steady and unsteady (“High Cycle Fatigue” „Hot Streaks“) operation mode.
Earlier studies, such as HEAT, have shown that the Reynolds Analogy, which is currently the state-of-the-art in TRACE, is inappropriate for the accurate prediction of turbulent mixing of hot and cold fluid. The Reynolds Analogy is based on the assumption that turbulent fluctuation increases the artificial turbulent viscosity and conductivity in the same way, regardless of physical transport processes.
However, for parts where film cooling is necessary, the prediction of hot and cold mixing is essential. The COOREFLEX-turbo project 3.4.3 aims to investigate the potential accuracy improvements in film-cooled flow, and higher order modern models must be tested for their industrial applicability.