NIST Time|NIST Home|About NIST|Contact NIST

HomeAll Years:AuthorKeywordTitle2005-2010:AuthorKeywordTitle

Purely Buoyant Diffusion Flames: Some Experimental Results. Final Report.


pdf icon Purely Buoyant Diffusion Flames: Some Experimental Results. Final Report. (3148 K)
McCaffrey, B. J.

NBSIR 79-1910; 49 p. October 1979.

Combustion Institute/Eastern States Section. Chemical and Physical Processes in Combustion. 1978 Fall Technical Meeting. November 29-30, 1978-December 1, 1978, Miami Beach, FL, 8/1-4 pp, 1979.

Available from:

National Technical Information Service (NTIS), Technology Administration, U.S. Department of Commerce, Springfield, VA 22161.
Telephone: 1-800-553-6847 or 703-605-6000;
Fax: 703-605-6900.
Website: http://www.ntis.gov
Order number: PB80-112113

Keywords:

buoyancy; diffusion flames; plumes; scaling

Abstract:

Measurements of temperature and velocity using thermocouples and an impact probe were made in the near field of a purely buoyant diffusion flame produced by a porous refractory burner. Based on time-averaged center line value of V and delta T together with photographic records the flame can be conveniently divided into three distinct regimes: (1) a continous flame region, starting from the surface of the burner with V equal to zero at the surface and rising with the height above the burner. Z, to the 1/2 power. Delta T is constant over this regime. Higher up is (2), an intermittent regime, with pulsating flame (~3 Hz) exhibiting approximately constant V and delta T falling with z to the first power. Still higher is (3) the plume region which is, most of the time, free of flames with V~z-1/3 and delta T~z(-5/3) as predicted by conventional plume theory. Throughout the three regimes and indistinguishable among these is the consistency of the buoyancy relation, [equation omitted] which has a value of approximately 0.9, a factor 2.5 times previous estimates in the flame region and confirming the recent correlation measurements of Cox(1). Different heat release rates, [equation omitted], can be scaled to a "universal" fire if the length is normalized as z/[equation omitted](2/5) and the velocity scale as v[equation omitted](1/5). The flames regime is thus independent of [equation omitted]. In the radial direction for time-averaged quantities only the plume region appears resonably Gaussian. The data in the flame and intermittent regimes do not fall as rapidly as that dictated by a Gaussian distribution. In all three regimes the velocity profile is wider than the temperature profile. Large scale, low frequency p fluctuations are about 35% of the time-averaged signal on the center line throughout the three regions. Radially the fluctuating to time-averaged signal ratio rises from the center line value and approaches 100% in the wings. Elementary spectral analysis indicates that most of this energy is concentrated in a narrow band centered around 3 Hz. Implications of these results for flame entainment calculations and heat release rates will be discussed.