Soot formation and oxidation in normal and inverse diffusion flames

Abstract

A numerical study has been carried out to address the sooting characteristics of normal and inverse diffusion flames. The simulation was performed on the basis of single-step kinetics with a laminar flow assumption and non-unity Lewis number. Five different coaxial normal and inverse flame combinations with different momentum ratios have been investigated and compared. The results were experimentally validated using an identical burner with concentric cylindrical ports, where flame temperature measurements were employed for flame length comparison while a paper filter was utilized for soot concentration measurement. It was revealed that the momentum diffusion between the parallel streams governs the flame length and size. The results also showed that the single-step kinetics is able to predict the maximum soot formation in the normal flame annular region, provided that the Lewis number is less than unity. In inverse diffusion flames, the region of the soot surface growth is too small and the flow residence time is too short to allow significant soot production. For all combined inverse/normal diffusion flames, the soot emitted by the inner inverse flame had almost no influence on the total soot production. The thermal radiation by soot increased with the increase in both inner and outer flame lengths. The inner flame contributed to a 13 per cent increase in the total thermal radiation with keeping the same exhaust soot concentration level.

NOx emission performance of triple flames

“…The non-premixed flame had the largest over-prediction range in the sense that it is the one that produces the highest amount of soot and it is also the one whose combustion efficiency (and hence flame length) is mostly affected by turbulence. A similar deviation of 10 per cent was previously attributed to the kinetics employed [31].…”

Performance enhancement of inverse diffusion flame burners with distributed ports

“…31 A minimum spatial step of 0.05 mm is used at the reaction zones to assure a degree of fineness that is sufficient to adequately simulate the detailed reactions related to NO x formation. A Lewis number of 0.95 is assumed to assure that the intersection of the temperature and fuel mass fraction contours properly predict the position of soot in the annular region around the fuel jet, 13 while both turbulent Schmidt and Prandtl numbers are set equal to 0.7. Both the turbulent viscosity and the corresponding turbulence scale are pronounced via solving the turbulent kinetic energy and dissipation rate by setting the inlet turbulence level at 5%.…”

“…For effectively oxidizing such residual soot particles and simultaneously keeping the high turbulence intensities associated with the central air port positioning, the introduction of an outer air supply was highlighted in some other studies such as that by Kamal. 13…”

Numerical Prediction of Radiation and Air Preheating Effects on the Soot Formation in a Confined Laminar Co-Flow Diffusion Flame

“…Smooke et al [10] studied laminar, sooting, coflow diffusion flames experimentally and computationally and observed that the impact of radiative power loss on temperature had a significant effect on soot formation. Kamal [11] carried out a numerical study to address the sooting characteristics of normal and inverse diffusion flames on the basis of single step kinetics with a laminar flow assumption and non-unity Lewis number. Saji et al [12] studied numerically and experimentally a laminar ethylene-ambient air co-flowing jet diffusion flame.…”