A subgrid model for equilibrium chemistry in turbulent flows

Abstract

A method is presented whereby the fast chemistry reaction, Fuel+(r)Oxidizer →(1+r) Product, may be modeled in the context of a large eddy simulation (LES). The model is based on a presumed form for the subgrid-scale probability density function (PDF) of a conserved scalar. The nature of the subgrid-scale statistics is discussed and it is shown that a beta function representation of the subgrid-scale PDF is appropriate. Data from both laboratory experiments and direct numerical simulations (DNS) are used to show that the predictions of the model are very accurate, given the exact values for the filtered scalar and its variance. A possible model for this variance is presented based on scale similarity.