CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM (http://www.cfd-online.com/Forums/openfoam/)
-   -   SonicFoam simulation probleme (http://www.cfd-online.com/Forums/openfoam/79805-sonicfoam-simulation-probleme.html)

bruxellois September 4, 2010 17:22

SonicFoam simulation probleme
 
1 Attachment(s)
Dear Foamers!

I'm trying to simulate the exhaust of a muzzle, my mesh and my boundary conditions are ok (transient boundary condition (time-depended) ), i'm using sonicFoam (with perfect gas).

the simulation do not give any result (0 iteration all the time).

You can find my case directory here

http://www.4shared.com/file/hpAUp6tE/test64tar.html


Any help please.

Alex

wyldckat September 5, 2010 16:29

Hi Alex,

What version of OpenFOAM are you using and which gcc version did you use to build it?

To know the gcc version, run:
Code:

gcc -v
I haven't tried to run the case yet, but knowing the versions would help reducing the chances of "works for me, but not for you" ;)

Best regards,
Bruno

bruxellois September 5, 2010 16:35

Hi Bruno,

I'm using OpenFoam 1.7.0 and gcc version 4.4.3.

Best regards

wyldckat September 5, 2010 16:52

Hi Alex,

Well, that was easy... I didn't even need to run it. The problem is that you didn't notice that your write time is larger than your end time:
Code:

endTime        0.00007 ;
(...)
writeInterval  0.0005;

Therefore, if it did write a time folder >0, then there would be something veeeery wrong ;)

And yes, I've made this mistake too in the past :rolleyes:
Always keep in mind that with OpenFOAM (and probably other CFD applications too), you must handle input and output data with eagle eyes :) Otherwise, you might end up with a GIGO case (Garbage in - Garbage out).

Best regards,
Bruno

bruxellois September 5, 2010 16:59

Hi Bruno,

I've notice this error ans i modified it

ndTime 0.007 ;
(...)
writeInterval 0.0005;

But there is no change, all the simulation is with O iterations (p,e, Ux..) so the result is the same of my initials conditions. I let it working for 10 hours without any change.

Any idea.

Thanks a lot.

wyldckat September 5, 2010 19:14

2 Attachment(s)
Hi Alex,

Let me first just warn you: I have a lot of experience in installing OpenFOAM in Linux and Windows, but sadly I don't have much experience actually using it :rolleyes:

Nonetheless, I changed a bit the the times in the controlDict to:
Code:

endTime        0.0007 ;

deltaT          0.000008;

writeControl    runTime;

writeInterval  0.0001;

But I only let it run until it gave me a new time snapshot. The new snapshot was at 9.6e-5 seconds. Attached are the two snapshots, looking at the entrance zone and they compare U vectors with magnitude for color scale. The residues indicated that convergence is almost immediate, at least if deltaT allows for it to converge.
I can only assume two things from these two pictures:
  • the mesh could use some more refinement near the entrance... but I'm not sure about this.
  • the mesh doesn't have an outlet! If you are testing the exhaust of a muzzle, I can only assume that you don't want to compress the fluid inside the mesh volume.
Either way, if you run checkMesh, it reports a point that isn't exactly on the surface of one of the patches. Well, it's barely an error, when compared with the simulation domain...

My suggestion (as a pseudo-amateur) is to first test the mesh with icoFoam or simpleFoam, and see if the air can simply flow from one place to the other, without taking into account accuracy in turbulence and all that. Simply test the mesh if is a valid simulation domain.

Best regards and good luck!
Bruno

bruxellois September 5, 2010 21:03

Hi Bruno,

Thanks for your response.

I had my mesh from a fluentMeshToFoam conversion, i've run the case under fluent and it gave a good result.


Best regards.

kiran October 30, 2010 11:06

Hi bruxe

how about your simulations. explain me your problem which your are facing now. snd a private msg.

regards
Ambilpur Kiran


All times are GMT -4. The time now is 19:14.