Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
MESH import almost successful
Posted Apr 19, 2012, 9:09 a.m. EDT Computational Fluid Dynamics (CFD), Mesh Version 4.2a 7 Replies
Please login with a confirmed email address before reporting spam
I am new to COMSOL and I would like to use it for a comparison with my Fluent and OpenFOAM results. I generated unstructured as well as structured meshes with ICEM. They are now in *.msh format. Due to the fact that COMSOL requires nastran files for instance, I converted the mesh to a *.nas file. COMSOL was then able to import it very quick, but when I want to set up the boundary conditions, all faces can not be selected (not applicable) to be inlet, outlet, etc.
Everything stays gray.
Another thing is when I try to start a simulation (all patches are walls by default), an error message occurs: "Problem setting up multigrid. Multigrid not supported on imported mesh. Error in multiphysics compilation."
Do you have any idea how I can solve those issues?
Thank you very much in advance.
Please login with a confirmed email address before reporting spam
Regarding the boundary problem, you'd better attach the comsol file to let other help you.
Regarding the multigrid solver, you may try to reduce the coarsening factor from 2 to 1.6 for example, or 1.4 or 1.2, however you first need to delete the temporary mesh the multigrid solver created otherwise you will get again that same error, you can find it under the mesh node, you will notice there is some strange mesh node.
However, changing the coarsening factor may influence the reliability of your solution, so may use a direct solver rather than a multigrid.
Cheers
Please login with a confirmed email address before reporting spam
Im not sure there have any "good" solution, the issue, from my understanding, is that COMSOL builds the model on the geoemetrical Entities it derives from your geoemtric Objects. But when you import a mesh it's no "geometry" in the COMSOL sense. You must find a way to convert it to a geoemtry first. I manage back in 3.5, but I have never tried it out in V4.
Once you have a geoemtry, with an attached mesh then you can solve.
This is an essential difference between COMSOl and (some ? all) other FEM tools, with COMSOL you build your physics on the geometry, the meshing comes thereafteer (and NOT before)
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
After some more reading in the forum probably the best way is to start from the pure geometry. Although the meshing of the complex STL is not working so far, the solving might be more stable at the end.
To get a feeling for setting up a case a watched several tutorials and also created two simple 2D problems.
In both cases I was not able to get results. I have chosen the model, generated the geometry, meshed and defined boundary conditions. Finally the computation was done and the error decreased below 10e-3 but the resulting velocity and pressure fields always remained zero.
Do you have any idea what I could have done wrong. Do I need the initialize the values somehow? I have attached the case so maybe you are able to see what is missing.
Thank you so much in advance!
Philipp
Attachments:
Please login with a confirmed email address before reporting spam
I am not sure I understand your problem, in the model you uploaded the results show both the velocity and the pressure. I even managed to plot the streamlines.
Cheers
Please login with a confirmed email address before reporting spam
that surprises me because in all my results, nothing can be analyzed due to zero velocity and pressure fields.
s1.directupload.net/file/d/2866/xqqxnw43_png.htm
When I click on the circle I get values corresponding to the coordinates but I can not see any colors.
Do I have to switch something special on or should the results appear directly after convergence?
Greets!
Please login with a confirmed email address before reporting spam
Well, that is quite interesting, please find attached an image of velocity + streamlines. Please note that I simply opened your model and just added the streamlines. I don't know what you may need to 'turn on' in comsol.
Cheers
Attachments:
Please login with a confirmed email address before reporting spam
I also simulated another case and it was the same issue. The resulting surface plot stays in an uniform colour.
Very strange....