[gu1]

Hello,

I have a small problem and I haven't found a solution until then... I'm running a LES case (WALE model) for a natural convection problem (buoyantPimpleFoam; Co_max = 0.5 and Pr = 0.7) and the velocity results are not good in the center of the domain (the opposite occurs in the profiles computed near the wall).

On the other hand, the temperature profiles are very good!

I attached some images.

I believe that the problem may be related to the number of Prandlt being less than 1 (... since in this case the temperature is carried by the hydrodynamic portion), but I cannot say where I may be wrong.

Best.

[Tobermory]

Guilherme - it's difficult to comment without a little more information. What is the Prandtl number of fluid in the experiments? What are the different charts showing (what do 0.1, 0.5, 0.9 mean in the chart legends?). Long and short of it, though - if you have not matched the Pr number from the experiments, then you are probably not going to match both the T and velocity fields, since they will be diffusing at different rates.

[gu1]

Originally Posted by Tobermory

What is the Prandtl number of fluid in the experiments?

Pr = 0.705
Both the experiment and the CFD tests have the same Prandtl number.
And as you can see, the profiles, both temperature and velocity, are in agreement near to the wall... so the number of Prandtl is equivalent. The problem is at the middle of the domain (in this case, the velocity profile).

Originally Posted by Tobermory

What are the different charts showing (what do 0.1, 0.5, 0.9 mean in the chart legends?)

These are the heights where the profiles were taken from the geometry. If it were 0 (zero) it would be on the lower wall of the domain (bottom), if it were 1, on the upper wall of the domain (top).
0.5 means being in the exact middle of the domain.

Best.

[gu1]

Anyone can help me?