Imposed P-Outlet

[rbecker]

If I am correct the "Imposed-P-Outlet" Option in the GUI is the same as using dirichlet conditions for the pressure in the "old style " (icodcl(ifac,ipr) = 1, rcodcl(ifac,ipr) = desired value).
I have following simplified setup

, two channels connected by a small pipe, while the static pressure drop between lower and upper channel is approx. 3%.
My Setup is BC_1/B_3: inlet with v = 30m/s, BC_4: outlet, BC_2: imposed P-Outlet with p = 1,03*p0. The reference pressure point is placed on a face at BC_4.
Contrary to my expectation the reduced pressure

is lower at the exit of the lower channel (BC_2) than at the exit of the upper channel (BC_4)
Could you please point me in the right direction? What is wrong with my setup? Which is the misconception?

Thank you in advance.
Ralf

[Yvan Fournier]

Hello Ralf,

If might be a bit "optimistic" to impose the pressure on only one outlet and hope the reference pressure on the other outlet will handle the rest. That would be the expected/intuitive behavior, but the the reference pressure is not actually zone-based, so there may be interactions between the 2 (we need to fix this).

In the meantime, could you try imposing the pressure on both outlets ?

Best regards,

Yvan

[rbecker]

Thank you, Yvan.

Following your suggestion I have changed BC_4 to an Imposed-P-Outlet, setting the Pressure to p = 1e5 and BC_2 to an imposed P-Outlet with p=1.1e5. This should force a pressure difference in the range of 10%.

Have look at the result

Still not the expected pressure field.

Do you have some more hints?

Thank you in advance.
Regards,
Ralf

[Yvan Fournier]

Hello Ralf,

No, I'll check with other members of the team who know this part of the code better.

If you have a small test case you can post here, that would be great.

Best regards,

Yvan

[rbecker]

Dear Yvan,

I have attached my simplified Test-Case.

Kind regards,
Ralf

[Yvan Fournier]

Hello Ralf,

Ok I'll take a look. I'll probably need help from the other developers for this subject, so it might take a few days.

Best regards,

Yvan

[Yvan Fournier]

Hello Ralf,

I ran your case, and it seems that if you postprocess the "total pressure" instead of the pressure variable, you have the desired output.

The total pressure is closer to the real pressure (adding the hydrostatic pressure, and, if I am not mistaken, minus some artificial turbulence-resolution related terms), while the pressure is more a by-product of the solution process if I am not mistaken.

Naming the variables differently might be a good idea here....

Best regards,

Yvan

[rbecker]

Hello Yvan,

sorry for my late reply, I was in holiday.
Maybe I didn't get it. What I want is that in the upper channel the pressure is lower than in the lower channel and the pressure difference is about 10% (setup in Test.zip).
What I get is, independant of the variable "pressure" or "total_pressure" that the pressure in the upper channel is larger.
Could you archive your results and post them? So I could compare your results with mine.

Thank you.

Kind regards,
Ralf

[Yvan Fournier]

Hello,

Here is my run (only on a few time steps, as the pressure should match the boundary conditions immediately, and without the checkpoint files, to reduce the archive size).

Best regards,

Yvan

[rbecker]

Thank you ,

seems to a problem of the code_saturne Version.
Comparing the parameter files I see no difference. Running your setup with my older 5.xx version i get the same false results using a 6.04 version i get the expected output.
thank you.

Kind regards
Ralf