Comparison of direct numerical simulation of lean premixed methane–air flames with strained laminar flame calculations

Abstract

Direct numerical simulation (DNS) with complex chemistry was used to study statistics of displacement and consumption speeds in turbulent lean premixed methane–air flames. The main focus of the study is an evaluation of the extent to which a turbulent flame in the thin reaction zones regime can be described by an ensemble of strained laminar flames. Conditional averages with respect to strain for displacement and consumption speeds are presented over a wide range of strain typically encountered in a turbulent flame, compared with previous studies that either made local pointwise comparisons or conditioned the data on small strain and curvature. The conditional averages for positive strains are compared with calculated data from two different canonical strained laminar configurations to determine which is the optimal representation of a laminar flame structure embedded in a turbulent flame: the reactant-to-product (R-to-P) configuration or the symmetric twin flame configuration. Displacement speed statistics are compared for the progress-variable isosurface of maximum reaction rate and an isosurface toward the fresh gases, which are relevant for both modeling and interpretation of experiment results. Displacement speeds in the inner reaction layer are found to agree very well with the laminar R-to-P calculations over a wide range of strain for higher Damköhler number conditions, well beyond the regime in which agreement was expected. For lower Damköhler numbers, a reduced response to strain is observed, consistent with previous studies and theoretical expectations. Compared with the inner layer, broader and shifted probability density functions (PDFs) of displacement speed were observed in the fresh gases, and the agreement with the R-to-P calculations deteriorated. Consumption speeds show a poorer agreement with strained laminar calculations, which is attributed to multidimensional effects and a more attenuated unsteady response to strain fluctuations; however, they also show less departure from the unstrained laminar value, suggesting that detailed modeling of this quantity may not be critical for the conditions considered. For all quantities investigated, including CO production, the R-to-P laminar configuration provides an improved description relative to the twin flame configuration, which predicts qualitatively incorrect trends and overestimates extinction.