Characterization of extinction/reignition events in turbulent premixed counterflow flames using strain-rate analysis

Abstract

We investigate extinction/reignition events in two contrasting turbulent premixed flames of the Yale turbulent counterflow flame (TCF) burner, that are both qualitatively and quantitatively different. One of the two chosen flames is a high-burning (HB) flame with a low probability of local extinction while the other flame is a low-burning (LB) flame with a high probability of local extinction. In recent work, we successfully studied the turbulent premixed flames of the Yale TCF burner using the large-eddy simulation/probability density function (LES/PDF) methods. In this present study, the main motivation is to investigate how the compositional structure of the two turbulent flames are related to that of laminar flames. To this end, steady, one-dimensional, strained, opposed-jet laminar flame calculations are performed to investigate the effect of strain rates K on the laminar counterparts of HB and LB, and to evaluate the extinction strain rates Sext for the two flames. Subsequently, a normalized distance Z is defined in terms of the mixture fraction ξ, which is calculated based on the mass fraction of N2. The scatter plots from the particle data of (a) CH2O mass fraction YCH2O*vs. progress variable p* and (b) temperature T*vs. the normalized distance Z* are quite different for the two flames with more samples close to the extinguished laminar profile for the LB flame than for the HB flame. The cell-mean profiles of T vs. Z resemble the laminar profiles at different strain rates even though the LES/PDF are non-trivially 3D and unsteady. These cell-mean profiles are used to evaluate the instantaneous equivalent steady strain rate (ESSR) S for the two flames. The cumulative distribution function (CDF) of S, conditional on S < Sext (i.e., burning samples), is somewhat similar, with the distributions being broad without a peak close to the bulk strain rate. However, comparatively, more samples of the LB flame have the ESSR values above the extinction strain rate, i.e., S > Sext. The scatter plots of the ESSR S vs. the fresh product layer thickness Δf quantify the thinning of the product layer as the strain rate S increases. The scatter for the HB flame follow the corresponding laminar profile quite closely, whereas, the thicknesses observed for the LB flame are higher compared to its laminar prediction.