Combustion and Flame, Vol.110, No.3, 298-318, 1997
Modelling finite-rate chemistry effects in nonpremixed turbulent combustion: Test on the bluff-body stabilized flame
The validity of the turbulent combustion M.I.L. model has been investigated in the case of an unconfined bluff-body burner. This model allows for a distribution of turbulent timescales and makes use of a chemical delay time library; it is specially directed toward nonpremixed combustion with finite-rate chemistry effects. Because of the extreme complexity of the bluff-body flow, the study focuses on the ability of the model to simulate various regimes of the burner rather than on a detailed calculation of the flow structure. Four regimes, related to different fuel-to-air velocity ratios V-j/V-a are investigated: namely, cases N (V-j/V-a = 21/25), M (V-j/V-a = 21/15), L (V-j/V-a = 21/7.5), and H (V-j/V-a = 60/15). The general features of the flow, as well as the effect of heat release on the dynamic held, are reasonably well predicted. In agreement with experiments, the computed N and M flames are found to be detached, while flame L is found to be attached. The H flame, which relates to the same region of the stability map as the L flame, is also predicted to be attached. It is proved that the calculated flame structure is sensitive to the prescribed air-inlet conditions, to the chemical delay time library (that is, whether it is derived from a global or a detailed reaction mechanism) and to the scalar-to-velocity timescales ratio R. The length and temperature of the flames in the different regimes are rather well predicted, provided that: the air-inlet boundary conditions are extrapolated from the experimental ones, the chemical time library is derived from a detailed mechanism) and a value slightly lower than the standard one is assigned to R.