BC set at fixedValue (0 0 0), but I get a non-zero velocity @ wall
 

Hi Foamers,

I have the following problem:

At the moment, I'm performing simulations of the flow through a pipe by using rhoTurbFoam. I made an axi-symmetric mesh in Gmsh and converted it to OpenFoam.

At the wall, which I've defined as "wall" in the boundary dict, I've specified the velocity as: fixedValue, uniform (0 0 0)

Everything works fine, but when I look at the resulting data in paraview I see a non-zero velocity at the wall. Why? How?

I've tried a couple of things, but nothing seems to work.
- specifying the sides as symmetryPlane
- speciying the sides as wedge
- refining the mesh
- different rotation (extruding) direction

The steps I did:
- Surface in Gmsh
- Extrude surface + mesh around the axis
- gmshToFoam
- define patches
- define BC's

Who can help me out here?

You are probably looking at interpolated values.
Try to visualize the "U" field in the cells (it is the one with a sort of cube as icon in the drop-down list in paraview)

Francesco

Thanks for the quick reply!
But I'd already tried that and it actually makes it worse ;)

Ok, an update!

When I look at my complete "part", I see non-zero velocities at the wall.
When I only import the wall in paraview, I see zero velocity at the wall.

So, everything seems to work fine....but the post-processing is a bit strange imo.

What is causing this behaviour?
How can I make sure that paraview shows me the correct values at the wall?

I'm still not satified with the results.
When I look at the results of the pitzDaily tutorial, I clearly see that the velocity at the wall is 0. So I don't think I can blame Paraview

Is the problem causes by Gmsh and the conversion to OpenFoam?
How can I get rid of it? (don't say: "stop using Gmsh")

Hi Squadron,
have you solved your Problem yet?
I´ve got the same Problem here. I converted my Mesh from Gambit to OF and used the makeAxialMesh utility.
All worked fine, but i´ve got a non zero velocity at my Wall same as you. Any suggestions?

Linh

Hi Linh,

I'm still not sure, but I think this problem is caused by Gmsh or by me, doing something wrong in Gmsh. Maybe you can try blockmesh.
I'm now working on an other project, I won't be working with Gmsh anymore.

Age

Hallo,

I built a mesh with blockMesh and converted it with makeAxialmesh to axissymmetric. I plotted the velocity-profils and I get nonzero values at wall boundaries, too. My inlet velocity is about 0.265 m/s and directly at the wall the vellocity is 0.08 m/s by using the sample utility.

I even changed the writePrecision from 6 to 10 in the controldict, but nohting changed. Anybody an idea?

Thanks a lot Thomas

Hello Age,

It's been quite some time since you reported your problem in cfd-online. I am getting the same wrong data at wall with OF-2.1.x, were you ever able to fix that problem?

I need to run some other code on a 2D plane with the data of my axisymmetric simulation, with the proper BC's values. I don't want to just simply substitute those values with the proper ones.

thanks!

I am facing the similar problem. Did any one solved the issue?

My mesh was created using GAMBIT and it's a wedge geometry.