Impact of sub-grid scale models on resolving mixing and thermal shear layers in large eddy simulation of JHC flames

Predictive performance of sub-grid scale (SGS) models is investigated in Large Eddy Simulation (LES) of a methane/hydrogen jet-in-hot coflow (JHC) flame using a conceptual analysis proposed for the mean values of major combustion products. It is shown that mean values of major combustion products consist of valuable information on several characteristics of JHC flames such as flame thickness, flame volume and reaction intensity. In particular, sudden change of CO2 and H2O mass fractions from coflow contents to higher values, predicted by static SGS models, demonstrated a thin flame constricted around the center line. However, uniform ascending manner of CO2 and H2O contents in the coflow region predicted by dynamic SGS models revealed their ability on capturing characteristics of a distributed volumetric flame. For the temperature fluctuations in shear layers, the dynamic Smagorinsky model is also shown to provide better predictions than the constant-coefficient Smagorinsky model, the latter exhibiting significant over-predictions. It is also observed that the dynamic kinetic energy SGS model with its unique assumption of non-equilibrium turbulence is the best fitted SGS model for the JHC flames as it provides improved accuracy on developing mixing and thermal shear layers by solving an extra transport equation for turbulent kinetic energy.