porousSimpleFoam: oscillating velocity in the porous zone
 

Hello!
I try to use a porousSimpleFoam solver to simulate a flow in a 2D channel from two equal parts, a free flow part and a porous part (wall's type set in the 'slip' for simplicity):

http://img-fotki.yandex.ru/get/5601/...61536200_L.jpg

Base case is a angleDictImplicit.

My results are a good p, but strange oscillating U.
On the next sctrutured mesh:
http://img-fotki.yandex.ru/get/5106/...c6b86593_L.jpg

I get this solution:
http://img-fotki.yandex.ru/get/5701/...c78ed4bd_L.jpg

Velocity field is:
http://img-fotki.yandex.ru/get/5802/...b0d0a7fc_L.jpg


One can see what velocity is good but slightly oscillated on porous boundary. So, question is:

Are such oscillations caused by a stepwise porous source? If yes, how can I specify smooth porous source?

But more bad situation have place on non-structured mesh.
Mesh is
http://img-fotki.yandex.ru/get/5602/...b921d596_L.jpg

Solution is:
http://img-fotki.yandex.ru/get/6001/...6741e4eb_L.jpg

http://img-fotki.yandex.ru/get/5001/...3b4da787_L.jpg

One can see very oscillating velocity in the porous region.
So, why this happens? Non-orthogonal correction not helped.

Thanks.

Hi!

I have exactly the same problem. Did you find a solution for that?

kind regards,

marc

This problem has been reported to http://www.openfoam.com/mantisbt/view.php?id=134. A fix is also described there

This is not a problem.


Just change under-relaxation factor to 0.0001, and you will not have this oscillations.

OpenFOAM software is the best software among all CFD, and being unexperienced in CFD, does not meam that OpenFOAM has bugs, that should be reported like that...

This is absolute brilliant: under-relaxation is like alcohol: if you take enough you see no problems anymore. Getting things done will take forever, but who cares?

Hi Bernhard,

You are right it takes a lot of time of calculations.

In industrial application it is not good idea to use it.

However, if you use 10 000 - 17 000 cells (hex) for the case proposed in this topic, and make underrelaxation = 0.0001, it will take 30 minutes to have an Exelent solution.

Unfortunately, tetrahedral cells were used, and seems to me under-relaxation 0.0001 will take a day or two for this simulation.

I feel like it is not industrial application, since it is tetrahedral, and 1-2 days is possible to give to get better velocity profile)) corrected pressure loss))

se9a
btw, I forgot to mension - I always use 7 cells in flow direction of porous zone + 0.0001 underrelaxation.

other regions can have very coarse mesh, since they are of no interest I guess

Thanks Bernhard

It worked like magic :)

Dear Bernhard

Thanks for the help. I tried it, somehow it works, but as it is also mentioned in there it causes a huge conservation error. Do you have any idea where does this error comes from? any help appreciated!

Best regards
Bahram

Have a look at the link http://www.openfoam.com/mantisbt/view.php?id=134

It is not my claim and I never checked how big HUGE is

Thanks a lot for quick respond, I have already checked that page but as I said before that idea causes some errors and there is not any clue in there to solve this fluctuating velocity without loosing or gaining some mass!!!

Hi guys,

I also have an oscillating velocity when simulating flow through different porous media adjacent to each other using fvOptions as well as the implicit treatment in porousSimpleFoam.

So the solution for this problem is to use a small under-relaxation factor like 0.0001? I try this, and it does work, nearly perfectly.

But any other new idea? Anyone know if fluent has the same problem?

Regards,
Yan

Fluen?
 

Dear Assistant Moderator

Do you have experience in fluent too? I have some questions about creating profile, etc.

Your quick help is highly appreciated?

Kindest Regards, Rayman

The instability is due to a co-located (aka Rhie and Chow) scheme being employed. The variable porosity problem is a classic benchmark for such schemes. There have been a number of solutions/modifications proposed in the literature over the years, with varying degrees of success. Staggered schemes do not suffer from this deficiency.

Hello!

Is it possible to "vanish" those oscillations just changing configurations in fvSchemes/fvSolutions? So far I just have found threads related to implementation of new algorithm. I have the same problem and just changing the under-relaxation factors to 0.0001 will not solve my problem. I thought using the solver porousSimpleFoam would help, but it doesn't.

Is there somebody who could give me some hints?

Best regards,

Roberto

how this oscillation can be corrected?. under relaxation factor change didn't helped me.

pororusSimpleFoam
 

Can you please share the link to see the fix?

This is the link to the bug-report: https://bugs.openfoam.org/view.php?id=134

However, I never got correct pressure values with this change.

Fix-porousSimpleFoam
 

Can you please explain how to add this fix in porousSimpleFoam?

Thanks in advance

You can follow the instructions here to compile a new solver:
https://openfoamwiki.net/index.php/H...ure_to_icoFoam

In your case copy porousSimpleFoam and replace the mentioned line of the solver as explained in the bug report.

sorry , i am new to openfoam. Please tell me where to copy this (means which file and where in file)

thank you