RANS modelling of turbulent non-premixed flames

Bhargav

Hi, Am trying to model an axi-symmetric turbulent jet diffusion flame using the standard k-epsilon model, and closure provided by the steady laminar flamelet model through the densities. To match the axial evolution of velocity with the experiment, I had to increase the constant C_eps1 in the epsilon equation. By doing this, I got the correct decay of the axial velocity and mixture fraction. But, my variance of mixture fraction was severely underpredicted. The original constant predicted a faster decay of the axial velocity and mixture fraction, but, also resulted in the correct prediction of peak value for the variance. If anybody has come across a similar situation please send in your comments. Regards Bhargav