CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   pressure in incompressible solvers e.g. simpleFoam (http://www.cfd-online.com/Forums/openfoam-solving/79411-pressure-incompressible-solvers-e-g-simplefoam.html)

 chrizzl August 23, 2010 08:45

pressure in incompressible solvers e.g. simpleFoam

Dear All,

I have a general question about the setup of the pressure in an incompressible solver like the simpleFoam-solver.

As we all know in simpleFoam the "strange" pressure units are a result of the partition by the density. This is reasonable, because the density will not change in an incompressible flow. Hence the only property of the fluid which can be changed by the user is the kinematic viscosity.

In my opinion in a pressure driven flow (pressure difference between inlet and outlet are known from an experiment) you have to normalise the fixed pressure values at inlet and outlet by the density.

In common commercial flow solvers you have to define a reference density. That is not possible in simpleFoam and why I wrote down my thoughts in this forum. It would be gald if someone could help me in this open discussion.

Regards,
Christian

 chrizzl August 24, 2010 03:23

Another reason for posting this massage is, that I have seen many users who don't care about the normalization of the pressure. I just want to make it clear (for me) in this forum.
There are experimental data like:
Inlet: 1,013,250 Pa
Outlet: 101,325 Pa
--> Delta p: 911,925 Pa

With a density of air round about 1,2 kg/m³ you've got:
Inlet: 844,375 m²/s²
Outlet: 84,437.5 m²/s²
--> Delta p: 759,937.5 m²/s²

Which setup is the correct one? In my opinion it's the second one;)
Thanks,
Christian

 Bjw August 24, 2010 05:41

Changing the scale and/or the unit of pressure should not change the physics of your problem.

 kjetil August 24, 2010 06:02

Well, you could just have a look at the units atop in the p-file. From there it should be very clear if your normalization is correct.

 chrizzl August 24, 2010 06:07

I know, that the physics are the same, but the first setup only make sense in a compressible simulation, the second one in an incompressible simulation.
If you set your pressure difference, which is known from an experiment without density-normalization to your simpleFoam-setup you will get a setup, which is different to the experiment (factor in this case 1.2).
--> The setup does not mirror your experiment. Is this correct?

From your reply I guess, that my thoughts are correct. Thanks a lot.

 sven82 August 24, 2010 08:08

Hi everyone,

I'm with Cristian, was your experimental data's initialization with a pressure difference,
you must dived the pressure through the density ("for a normalization"). After this the results should look like the experimental data's. A other opportunity is the work with a post processor.

best regards
Sven ;)

 Bjw August 24, 2010 08:41

Christian, sorry that I misunderstood your Question.

Quote:
 If you set your pressure difference, which is known from an experiment without density-normalization to your simpleFoam-setup you will get a setup, which is different to the experiment (factor in this case 1.2). --> The setup does not mirror your experiment. Is this correct?
Yes, absolutely. You must divide the pressure (Unit Pa) by density first when using an incompressible solver.

 chrizzl August 24, 2010 11:19

Quote:
 Originally Posted by Bjw (Post 272506) Yes, absolutely. You must divide the pressure (Unit Pa) by density first when using an incompressible solver.
Thanks, that's exactly what I was looking for.

 prasa April 8, 2016 03:47

Quote:
 Originally Posted by chrizzl (Post 272379) Dear All, I have a general question about the setup of the pressure in an incompressible solver like the simpleFoam-solver. As we all know in simpleFoam the "strange" pressure units are a result of the partition by the density. This is reasonable, because the density will not change in an incompressible flow. Hence the only property of the fluid which can be changed by the user is the kinematic viscosity. In my opinion in a pressure driven flow (pressure difference between inlet and outlet are known from an experiment) you have to normalise the fixed pressure values at inlet and outlet by the density. In common commercial flow solvers you have to define a reference density. That is not possible in simpleFoam and why I wrote down my thoughts in this forum. It would be gald if someone could help me in this open discussion. Regards, Christian

Hi,

In simplefoam continuity equation doesn't include the density because incompressible fluid. But in momentum equation density is a parameter(constant). Therefore pressure has introduced as pressure/density.

units [pressure/density] = m2/s3

cheers.

 Jan Brunner June 9, 2016 11:01

Quote:
 Originally Posted by prasa (Post 594031) units [pressure/density] = m2/s3
Hi Prasa

Thanks for wanting to clarify this but the unit should be m2/s2, right? ;)

All the best,
Jan

 prasa June 9, 2016 12:30

Quote:
 Originally Posted by Jan Brunner (Post 604134) Hi Prasa Thanks for wanting to clarify this but the unit should be m2/s2, right? ;) All the best, Jan
Hi Jan,

Yes it should be m2/s2.

Regards,
Prasa.

 All times are GMT -4. The time now is 16:44.