Performance assessment of flamelet models in flame-resolved LES of a high Karlovitz methane/air stratified premixed jet flame
Cunha Galeazzo Flavio Cesar,Savard Bruno,Wang Haiou,Hawkes Evatt R.,Chen Jacqueline H.,Krieger Filho Guenther Carlos
DOI: https://doi.org/10.1016/j.proci.2018.09.025
IF: 6.535
2019-01-01
Proceedings of the Combustion Institute
Abstract:Tabulated flamelets are commonly used in turbulent combustion modeling due to their relatively low computational cost, which is attractive in industrial applications. However, these models require assumptions of tabulated chemistry and subgrid-scale models for control variable distributions, both of which may contribute to modeling errors. In the present work, large-eddy simulation (LES) with tabulated flamelets is employed to study a laboratory-scale high Karlovitz number stratified premixed jet flame that was investigated recently using direct numerical simulation (DNS). Particularly, the LES resolves properly the transported control variables at a near DNS level, mitigating the errors from subgrid-scale modeling of control variable distributions. Five different flamelet tables are tested in the current work, including the conditional mean from the DNS, counterflow stratified premixed 1D flames with and without differential diffusion, freely propagating premixed 1D flames, and 0D autoigniting plug-flow reactors. The LES results show that although the flamelet tables perform differently for the instantaneous distributions of the progress variable source term, their mean distributions are similar. The mean and rms (root mean square) radial profiles for axial velocity and temperature from the LES with different flamelet tables are in good agreement with those from the DNS; more evident discrepancies are observed for the CH2O mass fraction radial profiles. Finally, the flame structures are examined in temperature space with the table from conditional means of the DNS having the best performance, as expected. (C) 2018 The Combustion Institute. Published by Elsevier Inc. All rights reserved.