RESULTS OF 4-EQUATION TURBULENCE MODELS IN THE PREDICTION OF HEAT TRANSFER TO SUPERCRITICAL PRESSURE FLUIDS

The paper summarises the results obtained in the assessment of different turbulence models including low-Reynolds k-e and k- equations, in the attempt to improve the prediction by RANS techniques of heat transfer to fluids at supercritical pressure. The work has been mainly developed in two phases. Firstly, 4-equation models available in literature were applied to a broad range of experimental data making use of the relationships suggested in their formulations for evaluating turbulent thermal diffusivity. These models often make use of (or were herein used with) an Algebraic Heat Flux Model (AHFM), aiming at evaluating the turbulent heat flux, though this is used only in the formulation of turbulence production due to buoyancy. In a second phase, the same models were applied repeatedly to a subset of the addressed experimental information with different calculation options, including constant values of the turbulent Prandtl number, mixing models for k- and k- equations in order to identify possible improvements. The results show that recourse to these models, which are more complex than common 2-equation ones, provides limited improvements in the comparison with experimental data.