Computational assessment of soot models in ethylene/air laminar diffusion flames

To improve the accuracy of soot formation and evolution predictions, several physical and chemical models have been developed over the last decades. These models include (i) detailed chemical kinetic mechanisms describing both gas-phase chemistry related to combustion processes and reaction pathways leading to large-sized aromatic molecules, which are needed for modelling soot formation, and (ii) soot models providing a comprehensive description of soot particle dynamics and interactions with gas-phase chemical species. Accordingly, in this work, two detailed soot models, the method of moments (MOM) and the discrete sectional method (DSM), are evaluated in ethylene/air laminar diffusion flames, and their corresponding results are compared with experimental measurements. Furthermore, the NBP and KM2 chemical kinetic mechanisms are assessed and compared with each other by examining key chemical species related to soot formation and evolution. To compute gas mixture's radiative properties, the weighted sum of grey gases model considering a grey medium is also utilised. Finally, the contributions of the soot precursors known as PAH (polycyclic aromatic hydrocarbon) to soot formation are also analysed. The main results show that the discrepancies in PAH concentrations obtained with different chemical kinetic mechanisms can be significant. In addition, compared to MOM ones, DSM results obtained here show a better agreement with experimental data. Finally, the analysis of PAH shows that those with two (A2) to four (A4) aromatic rings impact the most on soot modelling. Specifically, contributions of A4 were found to be more significant at lower heights above the burner, whereas A2 was found to be more impactful downstream as the flame develops. Maximum contributions of A2 and A4 to the soot inception rate were 66% and 85%, respectively, whereas the maximum summed contribution of PAH with five (A4R5) to seven (A7) aromatic rings accounted for only 13% of the inception rate.