Progress in numerical combustion

Citation
S. Candel et al., Progress in numerical combustion, COMB SCI T, 149(1-6), 1999, pp. 297-337
Citations number
136
Categorie Soggetti
Mechanical Engineering
Journal title
COMBUSTION SCIENCE AND TECHNOLOGY
ISSN journal
00102202 → ACNP
Volume
149
Issue
1-6
Year of publication
1999
Pages
297 - 337
Database
ISI
SICI code
0010-2202(1999)149:1-6<297:PINC>2.0.ZU;2-O
Abstract
This article begins with a synthetic presentation of key issues in the nume rical description of combustion phenomena. Different levels of combustion m odeling are identified and characterized. It is indicated how these modelin g levels may be used to deal with fundamental questions or technological ap plications. Important advances have been made in detailed numerical modelin g of complex flames and in direct simulation of flame/turbulence and flame/ flow interactions. Results obtained in these areas have been employed to im prove physical modeling methods which are currently used to calculate react ive flowfields in practical combustors operating in the turbulent regime. A s physical modeling relies on average Navier-Stokes equations it requires c losure rules for turbulent fluxes and for mean reaction rates. Considerable effort has been expanded to devise novel closure schemes or improve curren t models. Progress has been accomplished in the development of probabilisti c methods in which the probability density function is calculated. Recent d evelopments have concentrated on the incorporation of complex chemistry usi ng various reduction schemes. Advances have also been made in descriptions based on flame surface density concepts. Transport equations for the flame surface have been refined by processing flame data bases generated by direc t numerical simulations. Detailed experiments have also provided new insigh ts on the fundamental mechanisms of turbulent combustion and observations h ave led to novel submodels. Another area of considerable interest for the f uture is that of large eddy simulation (LES). Progress accomplished in nonr eactive LES provides a good starting basis but some challenging problems ar e encountered when dealing with combustion. The main difficulties are relat ed to the determination of the position of the large scale flame on the coa rse computational grid and to the definition of subgrid models for reaction in the small scales. Some novel schemes for combustion LES are reviewed an d illustrated by typical examples of calculations.