Exploding results

MantiumFlow CFD Software SupportExploding results
From MVRC Staff asked 1 week ago

Hi, the last days it seems unpossible for me to run a good simulation. About halfway the simulation the residuals start increasing a lot and the final drag/lift results are way to large numbers. Also the plots shows strange results, for example the streamlines goes in all directions already before hitting the car. I tried 1 design twice and also another design and for both designs the results seems exploding. Strange thing is I already had good simulations for both designs earlier. Only change is I runned them before with a rake angle and now I tried a simulation without rake. I’m using windows 10. I’ve no idea what is going wrong since I’m doing the same as I did before and then the simulations doesn’t show this problem.

7 Answers
MCAE Support Staff answered 1 week ago

Well that is not good. The issue is probably related to some surfaces that are not closed. Before you send the case, so I can have a closer look. Please try to run the check_closed_stls.sh script in the folder of your case. It should help indicate which files you should have a look at.

From MVRC Staff answered 1 week ago

Indeed there is 1 part where the surface isn’t closed. Is it something I do wrong with modeling? Strange thing in that case is that exactly the same .stl files where used earlier and only the rake angle changed. BTW I used the fast simulation templates.

From MVRC Staff answered 1 week ago

I runned the check_closed_stls.sh on older simulations and also there this part is not closed, while those simulations were working well.

From MVRC Staff answered 1 week ago

After some simulations which were working, now the following error appears during the meshing part. After this error MantiumFlow stops running the case:
[4]
[4]
[4] –> FOAM FATAL ERROR:
[4] Symmetry plane ’auto_wt_sym_plane’ is not planar.
At local face at (-0.63624919 5.9495768e-008 0.23797698) the normal (2.0158688e-005 -1 -1.936453e-005) differs from the average normal (-1.2109729e-009 -1 -1.8825903e-009) by 7.8133364e-010
Either split the patch into planar parts or use the symmetry patch type
[4]
[4] From function virtual void Foam::symmetryPlanePolyPatch::calcGeometry(Foam::PstreamBuffers&)
[4] in file meshes/polyMesh/polyPatches/constraint/symmetryPlane/symmetryPlanePolyPatch.C at line 64.
[4]
FOAM parallel run exiting
[4]
I searched those coordinates in my model, but there should be no bodywork at that point. Anyway it is really close to the point where max/min pressure where during the earlier exploding simulations.

MCAE Support Staff answered 1 week ago

This is a meshing issue at the actual symmetry plan. You should check if you car is realy symmetrical to this plane at Y=0. If the car is indeed symmetrical try having the surfaces at the symmetry plan come out of it at 90deg. If this does not help, just send in the case.

From MVRC Staff answered 1 week ago

Hi, I think I solved the auto_wt_sym_plane error. I placed the heat exchangers at the middle of the car, so that they where touching eachother. It seems this was creating the error. I now added a strip of bodywork between them and at least Mflow started with solving now. I’ve no results yet but I think it is solved, since it did not stop during meshing.

MCAE Support Staff answered 1 week ago

That could indeed have caused the issue as the mesher has to create a special zone for artificial flow resistance in the HX volume and for some reason that drives the mesher crazy. Why? Who knows, more or less the cells in that volume only need an indicator.