Propagation characteristics of turbulent premixed flames in nearly isotropic turbulent flows

Direct comparison of the turbulence flame speed (obtained from flame propagation velocity) and flame perimeter ratio has been made in turbulent premixed flames. The turbulent premixed flame is propagating freely downward for propane/air mixtures at various equivalence ratios, with u '/SL of ranging from 1.2 to 5.3 by using a pulsed flame flow reactor which generates flames propagating in nearly isotropic turbulent flow field. The measurement involved a high-speed digital imaging system at 1000 frames/second to capture the flame propagation motion. The turbulent flame speed ranged from 2.6 times the laminar flame speed to about 7 at high-turbulent intensities, while the flame perimeter ratio ranged from 1.4 to 3.3. In these freely propagating flames, the global flame curvature can lead to an enhancement of the flame speed by a factor of up to 3.5. For flames involving the coupling of globally curved flame geometry with flow divergence or any flow non-uniformity, correcting for this geometrical effect may require a careful consideration of the flame topology and flow field. The difference between the observed flame speed and the two-dimensional flame perimeter ratio, after correcting for the global flame curvature effect, is attributed to the fact that the flame wrinkles in three-dimensions are associated with a larger flame surface area more so than what is observable in the flame perimeter ratio data. This also points to a need to better understand the three-dimensional geometrical effects, including the global flame curvature and the local flame wrinkle structure in turbulent premixed flames. The observed turbulent flame speed data for the most part follows the flame speed models of Bray and Damkohler, wherein the flame surface area increase is modelled as a function of turbulence and thermochemical properties. The, above results, taken together, indicate that the fundamental assumption of the turbulent flame speed arising from the increased flame surface area is valid and can be used to estimate the turbulent flame speed within reasonable accuracy, provided that the three-dimensional flame geometry effects are correctly accounted for the global flame curvature and local flame wrinkle structure both. The flames observed at extreme equivalence ratios exhibit intermittent propagation in that only a small fraction of ignited flame kernel resulted in full propagation of the flame. Also, at low equivalence ratios, the flame speed decreased substantially even at high-turbulence intensities.