The objective of the study is to evaluate the potential of various turbulence models to simulate satisfactorily the influence of freestream turbulence intensity on the development of a cooling film, via a coupled computation, i.e. taking into account the full geometry (plenum, hole and main channel). Isotropic as well as anisotropic turbulence models (for the velocity as well as for the temperature fields) are tested, and an insight on the best suited closure is expected. The question of the respective influences of the various flow parameters (boundary layer characteristics, turbulent length scales, mass blowing ratios…) is also addressed. A low Reynolds number approach gives a correct estimation of the cooling effectiveness after approximately 10 hole diameters, for high or small blowing ratios, and using a k-ε model. The standard k-1 model largely underestimates the mixing in the injection region. The prediction of the injection region still needs to be improved for most configurations, but qualitatively the computation seems more than acceptable, as it exhibits the classically identified counter-rotating vortices that drive the heat transfer phenomena. The study also showed that predicting the influence of the freestream turbulence intensity requires taking into account thermal anisotropies, using an EARSMt (Explicit Algebraic Reynolds Stress Model, t being for Thermal) type model. An increase in freestream turbulence intensity was then shown to diminish the cooling effectiveness for all blowing ratios. The magnitude of the drop has still to be satisfactorily captured.

This content is only available via PDF.
You do not currently have access to this content.