Turbulence–equation of state interaction modeling in large-eddy simulation at supercritical pressure conditions

Abstract

The development cost of high-pressure energy conversion systems can be substantially reduced by using predictive computational fluid dynamics simulation tools. Under such conditions, the non-linearity of the equation of state (EoS) increases, and one may be required to provide a closure model for the interaction between the EoS and the unresolved turbulence scales in large-eddy simulations (LESs). The use of such a model has been limited to a few previous studies, and it is largely ignored by most others. This study explores the need for a closure model under different thermo-hydrodynamic conditions. A presumed probability density function approach is used to provide a closure model for the filtered density field in a pressure-based LES solver. The simulation results compare very well with the experiments, both qualitatively and quantitatively. This closure model is then used to prescribe a regime diagram that determines the necessity of using this model under different injection conditions. Multiple LES runs were conducted, both with and without the closure model, under various injection conditions. The findings demonstrate the efficacy of the regime diagram in identifying the conditions necessitating turbulence–EoS interaction closure modeling. In general, using such a closure model is recommended for transcritical injection—where the injection temperature is less than the pseudo-boiling temperature of the supercritical fluid—even in highly resolved large-eddy simulation cases.