CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM (https://www.cfd-online.com/Forums/openfoam/)
-   -   Framework of the Mesh (https://www.cfd-online.com/Forums/openfoam/88540-framework-mesh.html)

Eren10 May 19, 2011 12:33

Framework of the Mesh
 
Hi,

I have mesh generated with Pointwise and this program generates circular mesh frameworks. This means that I can't specify inlet and outlet conditions, so I have the same conditions around the radius of the mesh.
I thought , the circular shape doesn't matter for simulating flows around airfoils.

Now I have doubts about this :confused:. I think it should matter, doesn't it ???
generating mesh is easy but I do not know how to change the framework of the mesh.

Eren10 June 7, 2011 10:27

problems due to framework of the mesh
 
I have convergence problems due to the framework of mine mesh. I am trying to solve this issue with the icoFoam solver.

The framework of the mesh is represented with one patch, so I can't specify inlet/oulet separately ( this would solve my problem).

How should I define this one important patch ? I have used inletOutlet but the system doesn't converge.U file

Flow
{
type inletOutlet;
inletValue uniform (25 0 0);
value uniform (25 0 0);
}


p file

Flow
{
type outletInlet;
outletValue uniform 0;
value uniform 0;
}


How should I define this patch :confused:

maddalena June 7, 2011 10:54

Hello,
here is a long thread I started some time ago: http://www.cfd-online.com/Forums/ope...parison-2.html Go through it, around post 35 I started to use a circular domain (generated with fluent, but this does not matter). There is a lot of information on how your bc should be set and some more about convergence issues. Exactly what you are looking for, aren't you ;)?

mad

Eren10 June 8, 2011 06:22

Quote:

Originally Posted by maddalena (Post 310881)
Hello,
here is a long thread I started some time ago: http://www.cfd-online.com/Forums/ope...parison-2.html Go through it, around post 35 I started to use a circular domain (generated with fluent, but this does not matter). There is a lot of information on how your bc should be set and some more about convergence issues. Exactly what you are looking for, aren't you ;)?

mad

Thank you for your reply. :)

Indeed it is almost what I want, only you have 3D case.

For testing I am using icoFoam and than I will go to kepsilon. I am now using inletOutlet for U and p. But still it is not converging. maybe there is a bug in openfoam version 1.7.1 :confused:

maddalena June 9, 2011 03:23

Quote:

Originally Posted by Eren10 (Post 311016)
I am now using inletOutlet for U and p. But still it is not converging. maybe there is a bug in openfoam version 1.7.1 :confused:

No bug! Read the thread a little bit more carefully: you should use zeroGradient for p!

mad

Eren10 June 9, 2011 16:06

Quote:

Originally Posted by maddalena (Post 311200)
No bug! Read the thread a little bit more carefully: you should use zeroGradient for p!

mad


Thank you :) . I found an other very important problem: the trailing edge is perpendicular to the flow, which gives convergence problems. It seems Openfoam can't handle that part with normal mesh size :D . Maybe it is a good idea to choose an other airfoil.


All times are GMT -4. The time now is 20:16.