NUMERICAL SIMULATIONS OF ARGON PLASMA JETS FLOWING INTO COLD-AIR

Citation
Ch. Chang et Jd. Ramshaw, NUMERICAL SIMULATIONS OF ARGON PLASMA JETS FLOWING INTO COLD-AIR, Plasma chemistry and plasma processing, 13(2), 1993, pp. 189-209
Citations number
47
Categorie Soggetti
Physics, Applied","Engineering, Chemical","Phsycs, Fluid & Plasmas
ISSN journal
02724324
Volume
13
Issue
2
Year of publication
1993
Pages
189 - 209
Database
ISI
SICI code
0272-4324(1993)13:2<189:NSOAPJ>2.0.ZU;2-N
Abstract
Computational results and comparisons with experimental data are prese nted for simulations of axisymmetric turbulent argon plasma jets flowi ng into a cold air environment. The calculations were performed using the LAVA code [J. D. Ramshaw and C. H. Chang, Plasma Chem. Plasma Proc ess. 12, 299 (1992)], and were designed to simulate experiments perfor med by Brossa and Pfender [Plasma Chem. Plasma Process. 8, 75 (1988)] (BP) and by Fincke et al [private communication, 1992] (FSH). To our k nowledge, these are the first such simulations in which multicomponent diffusion and interactions between dissociation and ionization of dif ferent species are consistently accounted for. Turbulence effects were represented by a standard k-epsilon model, both with and without an a xisymmetric jet correction term and for several different choices of t he turbulent Prandtl and Schmidt numbers Pr(t) and Sc(t). Simulations were performed for one FSH experiment and two BP experiments at differ ent values of torch power P and argon flow rate W. The inflow profiles in the FSH simulations were adjusted to match P, W, and the experimen tal data slightly downstream of the torch exit as closely as possible. The same profile shapes were then used to match P and W for the BP si mulations, for which data near the torch exit were not available. Swir l was neglected except in one of the FSH calculations, where it was fo und to have negligible effect, as expected. Best results were obtained with the axisymmetric jet correction term omitted and with Pr(t) = Sc (t) = 0.7. Agreement with the experimental data was then fair overall, but still showed systematic deviations and cannot be regarded as full y satisfactory. Possible reasons for the discrepancies are discussed.