Modeling of high-speed, methane-air, turbulent combustion, Part II: Reduced methane oxidation chemistry
Rui Xu,Sai Sandeep Dammati,Xian Shi,Ethan Samuel Genter,Zoltan Jozefik,Matthew E. Harvazinski,Tianfeng Lu,Alexei Y. Poludnenko,Venkateswaran Sankaran,Alan R. Kerstein,Hai Wang
DOI: https://doi.org/10.1016/j.combustflame.2024.113380
IF: 4.4
2024-03-06
Combustion and Flame
Abstract:A reduced, 12-species reaction model (FFCMy-12) is proposed for modeling high-speed turbulent methane flames at high Karlovitz numbers. The model was derived from an early development version (FFCMy) of the 119-species Foundational Fuel Chemistry Model Version 2.0. The reduction was carried out by combining direct species pruning, quasi-steady-state assumption, and reaction lumping, targeting a minimum possible set of species that can capture methane combustion over a wide range of thermodynamic conditions. The performance of the reduced FFCMy-12 is compared to that of a 21-species skeletal reaction model (FFCM1-21) generated through conventional directed relation graph theory (DRG) and DRG-aided sensitivity analysis (DRGASA) algorithms. Model testing starts with legacy combustion properties such as homogeneous ignition delay time, laminar flame speed, and extinction/ignition residence time in a perfectly stirred reactor. More importantly, reduced model testing is extended to three-dimensional direct numerical simulations (DNS) of statistically planar, freely propagating turbulent premixed flames at Karlovitz numbers Ka=10 , 102 , 103 , and 104 , which nominally represent conditions from corrugated flamelets to broken reaction zones. Comparisons are made between the DNS results generated by the two chemical kinetic models with respect to turbulent flame structures, turbulent flame speed, and species distributions. Overall, presented results demonstrate the potential of FFCMy-12 for efficient modeling of the methane flames under highly turbulent mixing conditions characterized by a wide range of Ka. As importantly, the one-dimensional turbulence (ODT) model, developed in the companion paper (Part I), is shown to reproduce adequately the mean values of the local thermochemical states observed in the DNS, and as such, the ODT model is a viable DNS surrogate for testing the accuracy and applicability of a reduced model. Novelty and significance We present a 12-species reduced methane oxidation reaction model for the modeling of highly turbulent reacting flows. The reduced model is validated using DNS of premixed turbulent methane-air flames over a wide range of turbulent intensities, from relatively modest corresponding to Karlovitz number Ka=10 to ultra-high intensities at Ka=104 . This represents virtually the entire range of turbulent intensities that could be encountered in any realistic situations. The performance of the 12-species reduced model is evaluated against a 21-species skeletal methane oxidation reaction model. The results show excellent agreement between the two models for DNS at Ka=10−103 . The one-dimensional turbulence (ODT) model is also examined over the same conditions and is shown to be an effective DNS surrogate for evaluating chemical kinetic model reductions.
energy & fuels,engineering, chemical, mechanical,thermodynamics, multidisciplinary