Evaluation of Joint Probability D E N S I T Y Function Models for Turbulent Nonpremixed Combustion with Complex Chemistry

Two types of mixing sub models are evaluated in connection with a joint scalar probability density function method for turbulent nonpremixed combustion Model calculations are made and compared to simulation results for homogeneously dis tributed methane air reaction zones mixing and reacting in decaying turbulence within a two dimensional enclosed domain The comparison is arranged to ensure that both the simulation and model calculations a make use of exactly the same chemical mechanism b do not involve non unity Lewis number transport of species and c are free from radiation loss The modi ed Curl mixing sub model was found to provide superior predictive accuracy over the simple relaxation to mean sub model in the case studied Accuracy to within was found for global means of major species and temperature however nitric oxide prediction accuracy was lower and highly dependent on the choice of mixing sub model Both mixing sub models were found to produce non physical mixing behavior for mixture fractions removed from the immediate reaction zone A suggestion for a further modi ed Curl mixing sub model is made in connection with earlier work done in the eld