Joint scalar probability density function modeling of pollutant formation in piloted turbulent jet diffusion flames with comprehensive chemistry

A transported joint probability density function (pdf) approach was applied to model three (Re approximate to 8200, 22,400, and 44,800) piloted CH4/O-2/N-2 turbulent jet diffusion flames investigated experimentally by Barlow and co-workers. The flames covered conditions from weakly turbulent to close to extinction. The chemistry is a systematically reduced variant of the 48-species and 300-reaction C/H/N/O mechanism of Lindstedt and co-workers. The applied form features 16 independent, 4 dependent, and 28 steady-state scalars. The mechanism takes full account of the C-2 chemistry via the inclusion of C2H2, C2H4, and C2H6 as solved species. The formation of oxides of nitrogen is treated by a 5-independent-scalar (NO, NO2 NH3, N2O, and HCN) submechanism. The level of detail retained in the chemical mechanism is novel in the present context and permits an assessment of the accuracy of other closure approximations. The computational method for the solution of the joint-scalar pdf features moving particles in a Lagrangian Framework. The second moment closure by Speziale et al. is used for the velocity field, and molecular mixing is modeled using the ubiquitous modified Curl's model of Janicka et al. It is shown that excellent flow and scalar field predictions are possible across the range of Reynolds numbers and that conditional pdf values are reproduced with good accuracy even close to extinction. A sensitivity to the mechanical! scalar timescale ratio is, however, observed for flames under conditions of the latter type. Levels of NO become progressively overpredicted at downstream locations, though the magnitude of the discrepancy is arguably consistent with the increasing influence of radiative heat losses. The latter are not accounted for in the present work due to current uncertainties surrounding the accuracy of radiation submodels in the present flames.