CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM (http://www.cfd-online.com/Forums/openfoam/)
-   -   Compare 2D axi-symmetric and 3D. Is OpenFOAM calculate the velocity wrong? (http://www.cfd-online.com/Forums/openfoam/77055-compare-2d-axi-symmetric-3d-openfoam-calculate-velocity-wrong.html)

RalphS June 11, 2010 06:02

Compare 2D axi-symmetric and 3D. Is OpenFOAM calculate the velocity wrong?
 
2 Attachment(s)
Hi,

I simulate a swirl flow. A big cylinder (in which I took a displacer) following a small cylinder.
I try 2 dimensional axi-symmetric and 3 dimensional. All settings are equal. I used pisoFoam, Rankine vortex at inlet

I think OpenFOAM have problems with calculating the velocity in a 2D case. Enclose two picture.

U_x at the inlet is 10 m/s.
The continuity equation says: u1*A1=u2*A2

In my case, the velocity in the small passage in the middle should be 16 m/s.

u2 = u1 * (r1²/(r1²-rd²)) (With rd, I mean the radius of the displacer)
3D its right!

But in the 2D case the velocity is to high. Does OpenFOAM calculate the area ratio wrong?
If i would calculate like u2 = u1 *(r1/(r1-rd)). I get the velocity in the 2D axi-symmetric case.

The second problem is, that the flow at the axis makes no sense. Here I think OpenFOAM have problems with radial exchange in a 2D axi-symmetric case.

Have you any idea why the velocity is such different between these two cases?
Could it be, that I have forgot something to set in the 2d axi-symmetric case, or is there an additional setting to control something like area ration etc.

Regards
Ralph

johannesk June 14, 2010 12:09

2d case/wedge, is radial area ratio respected?
 
Hi Ralph,

nice to meet you here :)

If I understand you correctly, your problem with the 2d wedge case is as follows: the inlet is at a smaller radius than the outlet and has a smaller area as the outlet. Yet the results for U from OF look like they were the same area?

Quote:

Originally Posted by RalphS (Post 262601)
Have you any idea why the velocity is such different between these two cases?
Could it be, that I have forgot something to set in the 2d axi-symmetric case, or is there an additional setting to control something like area ration etc.

Could you post your polyMesh/boundary file of the 2d case, or maybe even the complete 2d case?

cheers,
Johannes

l_r_mcglashan June 14, 2010 12:39

Based on what the continuity equation gives you, it sounds to me as if you're not simulating a wedge, but an infinite plane. Are you following this:

http://openfoamwiki.net/index.php/Ma...s/AxiSymmetric

erncyc June 14, 2010 13:08

Hi Johannes...I am a new user to OpenFOAM and I am trying to simulate flow in a cyclone. I imported the mesh from a program caedium which created the .ofm. I get the following errors when i try to run please help thank you. (Ia m using simplefoam). I have also enclosed the offending file (ie RASproperties). thanks!

Errors below:

/*---------------------------------------------------------------------------*\
| ========= | |
| \\ / F ield | OpenFOAM: The Open Source CFD Toolbox |
| \\ / O peration | Version: 1.5 |
| \\ / A nd | Web: http://www.OpenFOAM.org |
| \\/ M anipulation | |
\*---------------------------------------------------------------------------*/
Exec : c:\cfd\OpenFOAM-1.5\bin\simpleFoam
Date : Jun 14 2010
Time : 20:07:31
Host : hostname
PID : 3416
Case : C:\Users\user\Documents\PhD\CFD\Openfoam\My-tutorials\cyclone\cyclone-flow
nProcs : 1

// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create time

Create mesh for time = 0

Reading field p

Reading field U

Reading/calculating face flux field phi

Selecting incompressible transport model Newtonian
Selecting RAS turbulence model kEpsilon
Programme terminated with errors: exit code 1, status 0.
Error messages:


keyword kEpsilonCoeffs is undefined in dictionary "C:\Users\user\Documents\PhD\CFD\Openfoam\My-tutorials\cyclone\cyclone-flow\constant\RASProperties"

file: C:\Users\user\Documents\PhD\CFD\Openfoam\My-tutorials\cyclone\cyclone-flow\constant\RASProperties from line 14 to line 15.

From function dictionary::subDict(const word& keyword)
in file C:\tmp\OpenFOAM-1.5\src\OpenFOAM\db\dictionary\dictionary.C at line 288.

FOAM exiting



Offending file below:

/*--------------------------------*- C++ -*----------------------------------*/
// OpenFOAM 1.6.x
// Generated by Symscape's Caedium on: 14 Jun 2010 16:20:41, www.symscape.com

FoamFile
{
version 2.0;
format binary;
class dictionary;
object RASProperties;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

RASModel kEpsilon;
turbulence on;





novyno November 12, 2010 12:17

Split wedge patch...
 
Hi RalphS,

You probably have defined a single wedge patch (called for example front&Back) in wich you have put every wedge surfaces.

Try to generate two different patchs for every "wedge side" (for example called front and back). This two patch of course will be "specular".

Now when you run your simulations the continuity equation will give you the "naturals" expected results. :-)

Regards!

RalphS November 13, 2010 08:00

Close the thread please.

Why you answered a question from june?

novyno November 13, 2010 21:51

I have answered becouse the thread original question wasn't solved (or the solution was not clearly described). I had the same problem some times ago and I hoped that my answer could help someone who encounter the same problems with wedge patchs.


All times are GMT -4. The time now is 14:03.