Hi,

1.Has anyone experience with the simulation of a vertical heated pipe with natural convection (massflow is not known a priori!)? Which bc? Tricks etc.

The example from the 4.5 manual p.14-88 cannot be found in the fluent 5 one. Does this mean it does not work properly with Fluent 5?

2. Can anyone understand the eq. 8.4.6 to 8.4.9 in the user's manual of Fluent 5, where the redifinition of the static pressure and the hydrostatic head is described? Isn't there s.th. wrong with minus in 8.4.8: I know that pressure is decreasing with height so it must be dp/dx=-density*g. And if I derive eq. 8.4.7 I never get 8.4.9. Maybe s.o. from the Fluent guys may help?

1. 'Pressure inlet' at the bottom and 'pressure outlet' at the top had worked for my previous test simulation.

2. For high Rayleigh number, you might use time-marching scheme, even though your interest is steady result.

Sincerely, Jinwook

Thank you for your answer.

I chose the same bondary conditions as you and had to carry out an transient calculation due to my high Raleigh number. But the results looked quite strange:

The direction of the flow was the same like the one of the gravity vector. This seems to be nonsense. Maybe my understanding of defining the gravity vector is wrong?:

When the pipe's up-vector is pointing to pos. x-axis direction (bottom has a lower x-coordinate than the the top), I defined g=-9.81. The result was a complete negative x-velocity field!!!

Did you do the simulation in 3d or 2d? Maybe it has s.th. to do with the axis-symmetric solver That I used.

Volker

Are you sure you specified a density versus temperature relationship for your fluid? If you have a constant density you will get the results you describe.

Yes I did. But the medium is not air but water!

Some additions:

Calculations with air and reference density of 1.2 give the results I am looking for. Doing the same with water not. Somehow I believe that I do not understand the sense of the reference density right now.

Now I try a reference density of 0 and on the other hand I will include hydrostatic head in my pressure boundaries . If there is no heating, there should be a balance of forces and a fluid at rest (no flow). With heating I should get a flow.