Transported Probability Density Function Simulations Of Turbulent Premixed Flames

Transported probability density function(TPDF)methods can accurately treat the mean reaction source term without assumptions of low-dimensional composition manifolds,and therefore have advantages towards the accurate prediction of turbulent combustion with strong turbulence-chemistry interaction.However,the effects of molecular diffusion require modeling by mixing models which are considered to be the primary challenge for TPDF methods.The research work aims at improving the predictability of TPDF methods for turbulent premixed combustion by enhancing the modeling of scalar mixing timescale,which is one of the two key ingredients of a mixing model.Furthermore,a systematic method has been developed to quantify the controlling physio-chemical processes in turbulent flames and thereafter to shed light on further improvement for TPDF simulations.Specifically,the following research work have been performed.First,the performance of the interaction by exchange with the mean(IEM)model and the Euclidean minimum spanning tree(EMST)model has been systematically investigated in turbulent premixed flames over a broad range of combustion regimes.The mixing-reaction budget analysis through Lagrangian particle tracking offers a direct physical interpretation for the superior performance of the EMST model for TDPF simulations of turbulent premixed combustion,and highlights the importance of developing more accurate mixing timescale models.Then,the effects of turbulence-chemistry interaction on reactive scalar mixing timescales in turbulent premixed flames have been investigated.It is found that for turbulent flames close to broken reaction zones regime,turbulence dominates the mixing process of reactive scalars.For turbulent flames close to flamelet regime,the mixing rates of reactive scalars are governed by both turbulence and flame structure.The study highlights the necessity to account for flame structure information when modeling reactive scalar mixing.Next,a recently proposed hybrid mixing timescale model for reactive scalars has been implemented in RANS/PDF simulations.The model,accounting for both turbulence and flame induced scalar mixing,can accurately recover the behavior of reactive scalar mixing in the flamelet regime and the broken reaction zones regime.Comprehensive a posteriori tests of the model performance have been performed over a wide range of turbulent premixed combustion regimes.The study shows that the model yields superior performance over the constant mechanical-to-scalar timescale ratio model for turbulent premixed flames close to flamelet regime.The hybrid mixing timescale model has been further augmented for reactive scalar mixing in the LES context.The LES/PDF simulations with the augmented model demonstrate the model potential to yield more accurate predictions for turbulent premixed flames.Finally,the particle-level sensitivity approach has been developed to quantify the controlling physio-chemical processes in the Sydney piloted premixed jet flames,PM1-50 and PM1-200.The study shows drastically different controlling physio-chemical processes in these two flames,and sheds light on the further improvement for the TPDF simulations of the PM1-200 flame.