Disagreement between SST k-omega and Realizable k-epsilon
 

Hi all,

I'm working on modeling a small-channel heat transfer experiment. The working fluid is supercritical carbon dioxide. FLUENT has been able to match the measured data well, so I'm trying to use computational output to examine some of what is going on in the boundary layer (especially near the point where the bulk fluid passes through the pseudo-critical temperature).

Odd thing is, the SST k-omega and k-epsilon models don't agree with each other, particularly in the turbulent core (y+ > 200). In this region, the two models are supposed to be the same mathematically, right? Any ideas what is going on?

P.S. The k-epsilon model agrees better with the universal profiles, so I think I trust it to be the "correct" result.