Assessment of semi-empirical soot production models in C1–C3 axisymmetric laminar diffusion flames

Abstract

The objective of this work is to assess the accuracy and limitations of three different semi-empirical soot models: an acetylene/benzene flamelet based two-equation soot model developed by Lindstedt (1994) [4] (in: H. Bockhorn, pp. 417–441) and two laminar smoke point (LSP) based soot models developed by Lautenberger et al. (2005) [15] (Fire Saf. J., 40, pp. 141–176) and by Yao et al. (2011) [14] (Fire Saf. J., 46, pp. 371–387). This task is performed by simulating twenty-seven C1–C3 hydrocarbon-air flames, consisting of normal (NDF) and inverse (IDF) diffusion flames and covering a wide range of residence times and fuel sooting propensities. The acetylene/benzene based soot model predicts the maximum soot volume fractions and the peaks of the integrated soot volume fraction within a factor of 2 of measurements. However, two pre-exponential factors for the surface growth process, one for the C1–C2 hydrocarbons and the other for the C3 hydrocarbons, were necessary in order to obtain this agreement. LSP soot model results are also within a factor of 2 of the measurements for most of the propane, ethylene, propylene and acetylene NDFs. For weakly sooting fuels, such as methane and ethane, the LSP concept overestimates soot production rates. In addition, the model of Lautenberger et al. was found to provide a better description of the soot production in the ethylene IDFs than that of Yao et al. Eventually, computed radiant fractions were compared with experimental data for propane, ethylene and propylene NDFs, showing that predictions using the models of Lindstedt and Lautenberger are on the whole within 10% of the experimental data.