Computational studies are performed on a Cambridge Stratified Swirl burner (SwB), a lean premixed stratified flame, by using the Reynolds Averaged Navier Stokes (RANS) model and the transported Probability Density Function (PDF) model. The SwB burner was measured
by Sweeney et al. (Combustion and Flame, 2012, 159: 2896-2911), and comprehensive data are available for model validation, e.g., the mean and root-mean-square values of velocity, temperature, and species mass fractions. The experimental data are available for sixteen different cases to investigate flames in premixed and stratified regimes, with or without swirl. In this study, we consider only non-swirling, premixed and stratified cases. Different
turbulence models are examined in the modeling studies, and the Reynolds Stress model with standard model constant values is found to perform well with the transported PDF model. A joint PDF for enthalpy and species mass fractions allows for the highly non-linear reaction term in the transport equation to be completely closed. The mixing term arising from molecular diffusion is not closed and requires modeling which is a significant challenge. For the SwB, we consider a series of mixing models including the Interaction by Exchange with the Mean (IEM) mixing model with different mixing model constants, the Modified Curl model, and two mixing models designed for premixed combustion from the literature. We first examine the models in the non-stratified/premixed case (SwB1) to isolate the effect
of other conditions from stratification on the model predictions. The stratification is then added in two levels, a moderately stratified case (SwB5) and a highly stratified case (SwB9). The predicted results are compared with the experimental data at various locations, inside and outside the recirculation zone in the burner. In general, good agreement is obtained for the velocity fields inside the recirculation zone. Good agreement is also obtained
of the predicted and measured results is obtained for the mean values of temperature and species mass fractions. The scalar fluctuations are generally underpredicted. Overall, the employed modeling method is able to capture the mean flame structure reasonably well in lean premixed stratified flames. Some limitations are noticed, e.g., the underprediction of scalar fluctuations, and overprediction of CH4 concentration in the stratified cases. These observations are useful for guiding the future research directions.