A sectional soot formation kinetics scheme with a new model for coagulation efficiency

Abstract A sectional scheme for soot formation is combined with a novel model for coagulation efficiency based on the thermal rebound concept and involves the minimisation of the Lennard-Jones potential energy between two colliding particles. Here, a novel generalisation of the interaction potential well depth is formulated for particles of any size or material composition that contains a free parameter λ which is related to the soot particle void fraction. A simplification that expresses the coagulation rate in an Arrhenius-like form is proposed so that the entire gas-phase and soot kinetics can be easily coupled and integrated using existing chemical kinetics solvers. The model is included in the sectional scheme, which accounts for nucleation, surface growth/oxidation, coagulation and fragmentation, and is validated against experimental data for a series of ethylene burner stabilised stagnation (BSS) premixed flames and a methane laminar coflow diffusion flame. The soot kinetics is discretised into lumped species according to soot particle size (or number of carbon atoms) and the soot evolution follows a 23-step abstract reaction scheme that is a reduced form of an existing multisectional soot kinetics scheme (Sirignano et al., Energy & Fuels 27, 2013). Overall, the model predictions produce a satisfactory agreement with the experimental data but there is considerable sensitivity to λ . In particular, λ has a strong effect on the soot particle size distribution (PSD) in the BSS flames. λ = 1 or 1.25 lead to an overprediction of large particle concentrations but result in the correct transition to a bimodal PSD as the stagnation plate separation distance increases, whereas λ = 2 is found to produce accurate predictions for nascent soot but suppresses the transition to a bimodal PSD. In the methane coflow diffusion flame increasing λ reduces the soot volume fraction but the predictions are also found to be very sensitive to the chosen numerical threshold for the lower soot size detection limit and results are presented for 2 nm and 7 nm thresholds for comparison against both laser-based and probe-based experimental data.