Dear all, I try to simulate a swirl-stabilized combustor. I employ the non-premixed-pdf-model. After tangeling around with the urfs I finally get a solution. But the flame seems to be unrealistic. It is very very thin. Turbulence modelling is covered by rke, wall treatment is done by enhanced wall treatment, y_plus is around 1. Does anyone has a suggestion regarding the following points: (1) Is there a turbulence model which may be suited better for this case? (2) Is it of any advantage to include the swirl-modification available in the turbulence model(s)? (3) Is it better to ramp the injected mass flow instead of increasing the urfs for the discrete phase? (4) Does high swirling flow eventually requires rsm-modelling? (I know about anisotropy, but I thought rke would take this into account.)

Thank you very much indeed for your answers.

Best regards, joerg

I uploaded a contour plot of total temperature. This doesn't look like a flame to me. But some of you combustion pro's can help me out with this, I hope. Thanks for any advice

http://img137.imageshack.us/my.php?i...ture002uy5.png