Direct numerical simulation of reacting scalar mixing layers

Citation
Smd. Kops et al., Direct numerical simulation of reacting scalar mixing layers, PHYS FLUIDS, 13(5), 2001, pp. 1450-1465
Citations number
41
Categorie Soggetti
Physics
Journal title
PHYSICS OF FLUIDS
ISSN journal
10706631 → ACNP
Volume
13
Issue
5
Year of publication
2001
Pages
1450 - 1465
Database
ISI
SICI code
1070-6631(200105)13:5<1450:DNSORS>2.0.ZU;2-R
Abstract
Understanding the passive reaction of two chemical species in shear-free tu rbulence with order unity Schmidt number is important in atmospheric and tu rbulent combustion research. The canonical configuration considered here is the reacting scalar mixing layer; in this problem two initially separated species mix and react downstream of a turbulence generating grid in a wind tunnel. A conserved scalar in this flow is, with some restrictions, analogo us to temperature in a thermal mixing layer, and considerable laboratory da ta are available on the latter. In this paper, results are reported from hi gh resolution, direct numerical simulations in which the evolution of the c onserved scalar field accurately matches that of the temperature field in e xisting laboratory experiments. Superimposed on the flow are passive, singl e-step reactions with a wide range of activation energies and stoichiometri c ratios (r). The resulting data include species concentrations as a functi on of three spatial dimensions plus time, and statistical moments and spect ra of all species. Several aspects of the flow are investigated here with t he conclusions that (1) reactions in which r not equal1 are more accurately modeled by frozen and equilibrium chemistry limits than are reactions in w hich r=1, (2) an existing definition of a reduced Damkohler number that inc ludes temperature and stoichiometry effects is a useful measure of reaction rate, and (3) existing theoretical models for predicting the coherence and phase of fuel-oxidizer cross-spectra and the spectrum of the equilibrium f uel mass fraction when r=1 yield accurate predictions. (C) 2001 American In stitute of Physics.