426 COMBUSTION AND FLAME 82:426-434 (1990) Soot Formation in Long Ethylene Diffusion Flames D. R. HONNERY and J. H. KENT Department of MechanicalEngineering, University of Sydney, N.S. W. 2006, Australia Soot measurements are made in long laminar ethylene diffusion flames to determine the effect of long residence times. Flames with lengths greater than 1 m are obtained by stab'tlizingthem in a round grid. Although soot volume fraction distributions for the long flame are different to the shorter flames, the maximum conversion fraction of fuel into soot remains about the same for the whole range of flame sizes. An important finding is that the total soot mass flow (kg/s) at any height in the soot growth region of the flame is only a function of the particle trajectory time from the reaction zone. This is true irrespective of the flame length and relative position in the flame. The surprising result implies that soot particle growth rates averaged across the flame section are not dependent on the gas-phase environment. NOMENCLATURE INTRODUCTION ag gravitational acceleration In the absence of an established chemical kinetic fo soot volume fraction model of soot formation, empirical relations in M s mass flow rate of soot terms of known parameters is sought [1, 2]. Soot Q volumetric fuel flow rate at room tempera- behaves differently than other species. Particle ture diffusivity is much smaller than gas-phase species r, R radial coordinate and so the widely used relationship between gas u, U axial component of velocity species mass fractions and a conserved scalar, x, X axial coordinate mixture fraction, cannot be applied [3, 4]. Such a lib mass fraction of inlet fuel burned relationship requires that chemical reaction is fast Ys mass flow of soot/carbon in fuel nozzle compared with gas mixing rates in the flame. mass flow This is true for the major species such as CO 2 and H20, but it is not clear whether it applies to Greek Symbols soot reactions. In order to establish an empirical model for soot formation, a prerequisite is to F transport coefficient determine whether soot formation rates are lim- /~ dynamic viscosity ited by chemical kinetic rates or by the mixing mixture fraction rates in the flame. In the former case, soot forrna- p density tion rate data may possibly be correlated with ~k stream function local mixture fraction, temperature, and other relevant parameters to build up a model. How- Subscripts ever, if soot formation is limited by the mixing rates, then the fluid mechanics determines the st stoichiometric local soot yield for a given fuel and soot forma- s soot tion rates cannot be related to local stoichiometry. Copyright © 1990 by The Combustion Institute Published by Elsevier Science Pubfishing Co., Inc. 0010-2180/90/$3.50 655 Avenue of the Americas, New York, NY 10010