[random_ran]

Greeting O.F.er:

I have a question regarding with no-slip boundary condition.

The simulation case is really simple: flow over a circular cylinder. My confusion is that why velocity contour on the surface of the cylinder is not zero? Is that violate the no-slip b.c?

I am using pisoFoam with O.F. v4.1. ParaView is v5.2.0.

Can someone elaborate a little bit of it?

Thanks,

Ran



Q1.jpg
Q2.jpg

[jherb]

You selected the point values of U to be displayed. What happens if you select the cell values (in the menu at the top select U with a box in front of it and not a point).

[random_ran]

Hi, jherb:

Thanks for your reply.

I have also checked the cell value, but it still gives me zero velocity contour on the surface of the cylinder. Is that something wrong with my understanding of the boundary condition? or it is a bug?

q22.jpg
q11.jpg

[Tobi]

The problem should be related to the reader. You are using paraview and not paraFoam, right? So paraview cannot handle the BC. Either use paraFoam or check out the files in your time directories. If for the BC you have »noSlip« and no other value, you are fine and it is really just a problem with displaying the stuff.

A workaround is to rename noSlip to fixedValue. Maybe in the latest ParaView version they fixed it.
Maybe you also can load an empty file named *.OpenFOAM

[wagnergaluppo]

Thanks for the answer

I also experienced that this is a dummy exhibit in ParaView. I tried to view the results with the paraFoam -builtin command usage, to check the solution while it was running, and I've got values for the velocity field on the wall. So when I reconstructed it and just checked the results with the paraFoam command, the velocity values for the wall were correct with value zero as expected for a wall with no slip condition. A figure is attached!

Best regards,
Wagner Galuppo

Screenshot from 2017-05-25 12-02-42.jpg

[bentkj]

Originally Posted by Tobi

The problem should be related to the reader. You are using paraview and not paraFoam, right? So paraview cannot handle the BC. Either use paraFoam or check out the files in your time directories. If for the BC you have »noSlip« and no other value, you are fine and it is really just a problem with displaying the stuff.

A workaround is to rename noSlip to fixedValue. Maybe in the latest ParaView version they fixed it.
Maybe you also can load an empty file named *.OpenFOAM

But noSlip and fixedValue (0 0 0) really means the same thing right? i guess as long as we've ensured one or the other for our BC we should be fine?

EDIT: Yes, I've experimented a little and like Wagner mentioned, the problem lies with the .foam and .OpenFOAM format for paraview. As long as the noSlip condition is ensured as the BC, it should be alright.

Thanks!
Ben

[random_ran]

Originally Posted by Tobi

The problem should be related to the reader. You are using paraview and not paraFoam, right? So paraview cannot handle the BC. Either use paraFoam or check out the files in your time directories. If for the BC you have »noSlip« and no other value, you are fine and it is really just a problem with displaying the stuff.

A workaround is to rename noSlip to fixedValue. Maybe in the latest ParaView version they fixed it.
Maybe you also can load an empty file named *.OpenFOAM

Thanks Tobi. Yes, I used ParaView instead of paraFoam. I did not fully understand how the ParaView really worked. ParaView helps us to read the data. I checked the raw data OpenFOAM calcuated for me. It only contains the value at cell center, which means that the other locations within the cell are interpreted from the the cell center data. ParaView must know the interpretation rule. In this thread, maybe I was incorrectly follow the rule, so I confused ParaView, which eventually caused ParaView to calcuate a non-zero velocity at the no-slip B.C..

Originally Posted by wagnergaluppo

Thanks for the answer

I also experienced that this is a dummy exhibit in ParaView. I tried to view the results with the paraFoam -builtin command usage, to check the solution while it was running, and I've got values for the velocity field on the wall. So when I reconstructed it and just checked the results with the paraFoam command, the velocity values for the wall were correct with value zero as expected for a wall with no slip condition. A figure is attached!

Best regards,
Wagner Galuppo

Attachment 56245

Hi Wagner,

From the provided information, paraFoam had the capacity to identify the no-slip B.C. and changed the interpretation rule at the cell whose faces are no-slip B.C.s. Thanks for your information.

Originally Posted by bentkj

But noSlip and fixedValue (0 0 0) really means the same thing right? i guess as long as we've ensured one or the other for our BC we should be fine?

EDIT: Yes, I've experimented a little and like Wagner mentioned, the problem lies with the .foam and .OpenFOAM format for paraview. As long as the noSlip condition is ensured as the BC, it should be alright.

Thanks!
Ben

Hi Ben:

Noslip is equal to fixedValue (0 0 0). They are the same thing. One is a jargon and the other is a mathmatical description.

I have tried the name trick, but it fails again (ParaView 5.4.0 64bit on Windows).

Thank you guys,

[dileeps]

Hi all, i tried to use foamToVTK and then opened vtk results. Thus i could obtain 0 wall velocity. I got it from a website

[Tobi]

Hi, the problem is that ParaView from Kitware does not handle the noSlip condition from OpenFOAM. If you want to obtain it, you need to run paraFoam which includes the builtin OpenFOAM reader that comes with OpenFOAM. Hence, the boundaries are displayed correctly. For a standalone paraview application without the builtin reader from OpenFOAM, one can convert it to VTK as already pointed out or you ignore that stuff.

[febriyan91]

What about opening result using paraFoam inside WSL while the Paraview is installed on Windows. since WSL doens not support GUI. Anyone has experience in it?

[otaolafr]

Originally Posted by febriyan91

What about opening result using paraFoam inside WSL while the Paraview is installed on Windows. since WSL doens not support GUI. Anyone has experience in it?

hello,
you have two options for this,
if your computer is capable, (I am still waiting for it with mine... thanks microsoft), you can update windows 10 to the windows 10 vr2004 (you can see your vr by properties in my pc icon in your desktop (if you do not have the my pc icon, go to desktop-> right click->personalise->themes->desktop icon settings ->select computer and save)). if you succesfully have a vr2004 you can use WSL2, that has a build in capability of use GUI. (note: WSL2 is noted to be better performance than WSL but takes more time in writing/reading big chuncks of data from the shared files with w10, i can not remember where i read this, but as i can not try it myself for the moment, i do not know how OF performs better or worst...

if you can not use WSL2, you can install xming for use GUI interface with WSL, I have not used myself as i did not bother to much as my w10 pc is only for preparation of the case, but this is a possibility, so a quick research in the subject will help you (sorry i can not help more with this... but at least having the right keywords will ease the search)

best regards!