Modification of low Reynolds number k-ε turbulence models for applications in supercritical fluid flows

One of the most important challenges in mathematical modeling of convective heat transfer to the turbulent flow of a supercritical fluid is the turbulence modeling. The turbulence modeling, like other aspects of the supercritical fluid flows, seems to be highly affected by the large variations of the fluid properties. There is no turbulence model available in literature which has been particularly developed for variable property conditions encountered in supercritical fluid flows. In current study, a two dimensional CFD code has been used to establish a modified Low Reynolds Number (LRN) k-epsilon turbulence model. The governing equations have been considered in their most general form to comply with the variable property conditions associated with supercritical fluid flows. Various areas of modifications have been tried and a new LRN k-epsilon turbulence model has been introduced. Incorporation of the present turbulence model into the numerical simulation of supercritical fluid flows leads to the flow predictions which are in good agreement with the experiments for enhanced mode of heat transfer. For deteriorated regime of heat transfer, considerable improvements have been achieved. (C) 2011 Elsevier Masson SAS. All rights reserved.