[f_mahmood]

Hi

I am trying to simulate natural convection of water in cylinder, approx with 2d cartesian, Ra=10^12 and Pr=7, using Ansys Fluent12. I have used K-epsilon model , with all default settings. And applied non-dimensional analysis, setting arbitrary vales for density and k to account for Ra and Pr. with this setting I used air as fluid, boussineque approx amd gravity of -1, to account for Pr, and Ra. BC are bottom wall=291k, upper wall=290k ( to have non-dimensional temp from 0 to 1), left and righ boundaries as adiabatic. But soultion remains unconverged. by changing turbulent model to k-omega, it diverges. i can't even change temp higher than 300k; it leads to divergence in temp. In operating conditions, I didn't use operating density in boussinesq parameter.

I have no idea how to get it resolved. I am newbie in Fluent. I even tried solution controls parameter of energy, but of no gain. any help and idea. appreciated.