Differential Diffusion Modeling in LES/FDF Simulations of Turbulent Flames

A modeling strategy to incorporate differential diffusion effects on both filter and subgrid scale in large-eddy simulation (LES)/filtered density function (FDF) method is proposed. Differential diffusion at the filter scale is resolved by the mean drift term in composition equations, whereas subgrid differential diffusion is modeled by the proposed modified Curl differential diffusion model in conjunction with a differential scalar mixing timescale model. LES/FDF simulations for a jet-in-hot-coflow methane-hydrogen flame have been performed to investigate the effects of filter-scale and subgrid-scale differential diffusion on flame characteristics. Results show that the predictions of the mean temperature and species distribution at the upstream improve significantly by accounting for filter-scale differential diffusion in the mean drift term. Meanwhile, the impact of accounting for subgrid-scale differential diffusion is in general not important for the predictions of the mean temperature and species distributions, because for the flame considered the controlling flame initiation process occurs mostly in the low shear region where subgrid nonuniformity is low and the resolved filter-scale molecular diffusion dominates over the subgrid-scale mixing. However, the incorporation of subgrid differential diffusion does improve the prediction of the conditional fluctuation of temperature in region near the shear layer where the subgrid nonuniformity is higher.