Lagrangian investigation of local extinction, re-ignition and auto-ignition in turbulent flames

Lagrangian PDF investigations are performed of the Sandia piloted flame E and the Cabra H2/N2 lifted flame to help develop a deeper understanding of local extinction, re-ignition and auto-ignition in these flames, and of the PDF models’ abilities to represent these phenomena. Lagrangian particle time series are extracted from the PDF model calculations and are analyzed. In the analysis of the results for flame E, the particle trajectories are divided into two groups: continuous burning and local extinction. For each group, the trajectories are further sub-divided based on the particles’ origin: the fuel stream, the oxidizer stream, the pilot stream, and the intermediate region. The PDF calculations are performed using each of three commonly used models of molecular mixing, namely the EMST, IEM and modified Curl mixing models. The calculations with different mixing models reproduce the local extinction and re-ignition processes observed in flame E reasonably well. The particle behavior produced by the IEM and modified Curl models is different from that produced by the EMST model, i.e., the temperature drops prior to (and sometimes during) re-ignition. Two different re-ignition mechanisms are identified for flame E: auto-ignition and mixing-reaction. In the Cabra H2/N2 lifted flame, the particle trajectories are divided into different categories based on the particles’ origin: the fuel stream, the oxidizer stream, and the intermediate region. The calculations reproduce the whole auto-ignition process reasonably well for the Cabra flame. Four stages of combustion in the Cabra flame are identified in the calculations by the different mixing models, i.e., pure mixing, auto-ignition, mixing-ignition, and fully burnt, although the individual particle behavior by the IEM and modified Curl models is different from that by the EMST model. The relative importance of mixing and reaction during re-ignition and auto-ignition are quantified for the IEM model.