Assessment of sub-grid scalar flux modelling in premixed flames for Large Eddy Simulations: A-priori Direct Numerical Simulation analysis

•Sub-grid scalar flux models for premixed turbulent combustion have been assessed.•Assessment conducted using a-priori analysis of Direct Numerical Simulations data.•Sub-grid scalar flux shows counter-gradient transport in all cases considered here.•The extent of counter-gradient transport increases with increasing filter width.•Models based on the alignment of local velocity and scalar gradients perform well.

The performances of different models for sub-grid scalar flux for premixed turbulent combustion in the context of Large Eddy Simulations (LES) have been assessed based on a Direct Numerical Simulation (DNS) database of freely propagating turbulent premixed flames with a range of different values of RetRet where Damköhler and Karlovitz numbers are altered independently of each other to bring about the variation of RetRet, whereas the heat release parameter ττ is kept unaltered. It has been found that the sub-grid scalar flux exhibits local counter-gradient transport for all cases considered here. However, the extent of counter-gradient transport decreases with decreasing values of filter width ΔΔ and for increasing values of the ratio of the root-mean-square turbulent velocity fluctuation to the unstrained laminar burning velocity u′/SLu′/SL. The performance of several algebraic models has been assessed with respect to explicitly filtered DNS data. The standard gradient hypothesis based model does not adequately capture both the qualitative and quantitative behaviours of sub-grid scalar flux for all cases for all filter widths. The models which account for local flame normal acceleration perform better than the standard gradient hypothesis model. In general the performance of the models, which account for the alignment of local resolved velocity and scalar gradients, remains relatively better than the performance of the other existing models. Detailed physical explanations have been provided for the observed model performances.