I'm doing some calculation on natural convection in cavities at high rayleigh numbers (Ra=10^10-10^16) and Pr = 0.71. I have serious convergence difficulties using a standard k-eps turbulence model. Does' anyone have any suggestions for me ??

Thanks you in advance

change the scheme to upwind and try to see if it will converge or not.

Hi,

The documentation has a discussion about obtaining convergence and you might get some ideas here: http://www.cfd-online.com/Wiki/Ansys...gence_criteria

Note that it is common for flows at these Rayliegh numbers will converge steady state as they are almost always transient. Are you running in transient mode?

Glenn Horrocks

Hi, I have done very similar simulation using fluent 2 yrs back and I see that certainly one has to be very strategic to get convergence for natural convection simulations . I would suggest that you solve the case steady for few iterations but soon change to transient since as Glenn says and I agree most of such flow are transient in nature.

Best of lucky