BUG: Comparison of Darcy's law with porousSimpleFoam

I guess there is a problem with solving flow through a porous medium. A comparison of porousSimpleFoam with Darcy's law fails and the computed pressure drop is too low in comparison to theory. It seems that there is always one discretization cell missing. I have attached a report to show the issue openFoam_bugReport.pdf

The problem was already reported here: https://www.cfd-online.com/Forums/openfoam-solving/183947-poroussimplefoam-comparison-darcys-law.html

The test case is attached Darcy_test.zip

I am using version v1606+ with a virtual machine on Windows 7.

## Reattaching the author to the issue ticket: @sepp.zell ##

Thanks for your query. I'll check this further.

assigned to @Prashant

Hi, I just wanted to know, if there is something new on this issue. Is the setup of my example correct?

By Sebastian on 2017-05-31T15:16:17 (imported from GitLab project)

## Reattaching the author to the issue ticket: @sepp.zell ##

Dear Sebastien,

Sorry for the delay. I should look into this soon.

/Prashant

Dear Prashant,

you do not have to apologize. It is only for my curiosity.

Regards

Sebastian

By Sebastian on 2017-06-02T07:59:29 (imported from GitLab project)

Dear Sebastien,

The issue is due to centered interpolation in pressure equation with implicit porous media.

With explicit porous media and adjustment in schemes, we can get desired pressure drop (without loss of one cell etc).

We have few ideas to correct this behavior. Would you be interested to get this developed?

/Prashant

Dear Prashant,

thank you for your message and sorry for my late reply. Attached you can find a document with comparison results of the explicit and implicit version of porousSimpleFoam. In my example there is still a discrepancy to the theoretical pressure drop value also for the explicit version. Additionally, also the fluid velocity is not as expected. Usually for an one-dimensional porous medium the fluid velocity is expected to be constant, when a constant inflow velocity is described. However, some spurious oscillations in the velocity occur.

You are mentioning that we can get the correct pressure drop with special interpolation schemes. Maybe you can tell me which scheme I should chose to get the desired results.

In principle I am interested in fixing this issue, but I have no idea where to make the necessary changes in the code.

Regards

Sebastian

Simulation_of_flow_through_a_porous_medium_with_OpenFOAM.pdf

By Sebastian on 2017-07-31T14:02:08 (imported from GitLab project)

Dear Sebastien,

Thank you for the amount of efforts you have done. Regarding schemes, you could try pointCellsLeastSquares grad scheme (for example)

We have other ideas, but this project is uncosted at the moment hence bears low priority.

/Prashant

Dear Prashant,

thank you for your suggestions. I tried the scheme, however there is not really an improvement. I will test different other schemes and let you know, if I can achieve better results.

I understand that there are much more important issues and at least it is good to know that it is not my fault that there is a discrepancy between theory and the results. Maybe there is time some day to fix the issue.

Kind regards and thank you for all your efforts

Sebastian

By Sebastian on 2017-08-03T12:21:46 (imported from GitLab project)

mentioned in issue #570 (closed)

This scheme: https://authors.elsevier.com/a/1WqWr508HeRhh seems to solve the problem of spurious oscillations.

By Santiago Marquez Damian on 2018-04-04T16:09:50 (imported from GitLab project)

## Reattaching the author to the issue ticket: @santiagomarquezd ##

Thanks @santiagomarquezd. We will have a look.

changed the description

Dear @santiagomarquezd , is it possible to share the name of the paper since the link you had shared was broken. Thank you.

Hello @kuti, no problem, the link was temporary one with free dowloading for 50 days. The journal link is:

https://www.sciencedirect.com/science/article/pii/S0021999118301931

"An oscillation-free flow solver based on flux reconstruction"

See Section 5.

It is also the first reference (to my knowledge) to the derivation of fvc::reconstruct method.

Regards!

Dear @kuti, have you had the chance to dig into this lately? I would be very interested Regards, Marco

assigned to @kuti and unassigned @Prashant

changed milestone to %v2112

added bug label

changed title from Comparison of Darcy's law with porousSimpleFoam to BUG: Comparison of Darcy's law with porousSimpleFoam