Inscrição na biblioteca: Guest
TSFP DL Home Arquivos Comitê executivo

HYBRID RANS-LES APPROACHES FOR THE PREDICTION OF TURBULENT HEAT FLUX

Donghwa Jeong
Department of Mechanical Engineering, Korea University 5 Anam-dong, Sungbuk-gu, Seoul, 136-701, Korea

Jong Keun Shin
Department of Automotive Engineering, Hanzhong University, 119, Jiheung-dong, Donghae, Kangwondo 240-713, Korea

Young Don Choi
Department of Mechanical Engineering, Korea University, 1, Anam-dong, Sungbuk-ku, Seoul 136-701, Korea

Resumo

In the engineering applications, various types of thermal boundary conditions can be imposed on the velocity field. It is thus necessary to obtain the practically important quantities, such as the turbulent heat flux and the heat transfer coefficient. Because of some success, it is still believed that the most reliable prediction methods are those based on the LES approaches. However, at high Reynolds numbers, LES is considered too CPU-demanding for complex industrial applications. Therefore computations based on a RANS model in the near wall regions, and on LES in some other regions, where explicit computation of the large scale structure is required, are needed as a hybrid RANS-LES(HRL) computation. This paper reports on application of PITM(Partially Integrated Transport Model), the seamless HRL, to the simulation of heat transfer of non-equilibrium flow. In order to reproduce the sub-grid scale eddy viscosity, PITM are based on v2−f model for the sub-grid scale eddy and temperature field is analyzed by eddy diffusivity model. The proposed models are validated against the DNS data of the fully-developed turbulent channel flow with uniform heat flux.