Study of alternative descriptions of methane oxidation for CFD modeling ofturbulent combustors

Citation
A. Brink et al., Study of alternative descriptions of methane oxidation for CFD modeling ofturbulent combustors, COMB SCI T, 141(1-6), 1999, pp. 59-81
Citations number
37
Categorie Soggetti
Mechanical Engineering
Journal title
COMBUSTION SCIENCE AND TECHNOLOGY
ISSN journal
00102202 → ACNP
Volume
141
Issue
1-6
Year of publication
1999
Pages
59 - 81
Database
ISI
SICI code
0010-2202(1999)141:1-6<59:SOADOM>2.0.ZU;2-Q
Abstract
Four alternative ways to describe methane oxidation that all can be used in computational fluid dynamics (CFD) have been investigated. The purpose has been to investigate how advanced a description of the methane oxidation is needed for an accurate description of the main components in practical CFD calculations. Especially the prediction of the intermediate combustion pro ducts CO and H-2 has been stressed, The four alternative approaches are: I) thermodynamic equilibrium, II) a three-step simplified mechanism including infinitely fast in;eversible global reaction steps for the oxidation of me thane and hydrogen, and a global, kinetically controlled, reaction step for carbon monoxide oxidation, III) a four-step simplified mechanism including two global reaction steps for the break-down of the hydrocarbon, a global reaction step for the oxidation of hydrogen and a global reaction step desc ribing the reversible water-shift reaction and IV) a comprehensive reaction mechanism including 32 chemical species and 156 reversible elementary reac tions. The comparison has been made by studying the concentration predictio ns for perfectly stirred reactor (PSR) conditions as a function of nominal residence time and by a CFD study of a diffusion flame. The results from th e PSR computations indicate that the three-step simplified mechanism is a g ood alternative for the fuel lean conditions and for flames far from extinc tion. The four-step simplified mechanism was found to be a good compromise between accuracy and computational speed. The four-step simplified mechanis m also performed well at moderately fuel rich conditions. However, in the C FD study of the diffusion flame, the difference in the results obtained wit h the four-step simplified formulation and the comprehensive mechanism was surprisingly large. The assumption of instantaneous local thermodynamic equ ilibrium in the reacting fraction of the fluid proved to be an unsuccessful alternative in the CFD modeling.