Problem with mill STL

Submitted by Sapsan on Mon, 02/27/2012 - 15:23

Dear community,

weeks ago I posted some questions regarding a set-up of a mill simulation.

Now, after much help from richti83, the ball mill simulation progressed.

But one problem still remains: the balls are "glued" to the lower(!) part of the lifting bar (see attachment).

I couldn't fix this problem. I tried to mesh it via GMSH, I drew it with ProE with zero wall thickness (bars had still thickness) then exported it to STEP and finally to STL, I tried to increase the wall thickness but it did not help either - the balls are still under the lifting bar when rotating (nonsense).

Can somebody help, please? What should I do to fix it?

Further, I want to know whether I can create such geometry directly via gmsh without drawing it in ProE first and then exporting it? Maybe this should help.

Thank you!

Best,

Eugen

AttachmentSize
Plain text icon in.Kugelmuehle_massrate.txt3.96 KB
Image icon Mill.PNG238.05 KB
Binary Data stl.tar_.gz52.08 KB
Binary Data km_stp.tar_.gz13.24 KB
Image icon Mill_01_03.PNG262.75 KB
Binary Data stl_01_03.tar_.gz269.93 KB

Forums:

ckloss's picture

ckloss | Tue, 02/28/2012 - 12:35

Hi Eugen,

to me the STL file you posted does not look like it comes from GMESH, as it contains triangles with really very high aspect ratios, and I guess this is the reason for the problems you are experiencing (you should get a warning about that when importing the mesh in LIGGGHTS). Typically, STL files from CAD look like that, but this is not a proper mesh for a numerical simulation. You could try simplifying the geometry if possible and necessary, export iges or step from proE and import it in gmsh and to mesh it there - I don't see why this should not work in your case.

Hope I could help!

Cheers, Christoph

Sapsan | Tue, 02/28/2012 - 15:53

Dear Christoph,

thank you for your reply! You are right, the stl-mesh is from ProE. I tried to proceed the way you describe it: exporting the STEP and then mesh it in GMSH but I have some problems with the lifting bars: these are semi circles which are meshed as wedges. I twiddled with different parameters and options in GMSH->Mesh but it did not really help. Then I noticed that meshing on different computers produced different results.

I will attach a STEP file, maybe you can try to mesh it in GMSH and say if the lifting bars are meshed acceptably on your computer.

Which parameters are the most important when creating a mesh? How can I make divisions where the mesh is very fine compared to the rest?

Thank you very much!

Best regards,

Eugen

ckloss's picture

ckloss | Tue, 02/28/2012 - 16:18

>>How can I make divisions where the mesh is very fine compared to the rest?
It should not be a problem to have local refinement anywhere

>>Which parameters are the most important when creating a mesh?
The aspect ratio of the triangles should not be too odd.

>>I will attach a STEP file, maybe you can try to mesh it in GMSH and say if the lifting bars are meshed acceptably on your computer.
Sorry, I do bugfixing, but not other people's work. If you find gmsh inappropriate for your application, try another one, there are tons of FE meshers out there which should do a pretty good job

Cheers, Christoph

Sapsan | Wed, 02/29/2012 - 08:57

Hi Christoph,

thanks for your reply. I understand that you can't do the work for other people :)

Can you please name a few good meshers, which I can try? From your site I know GMSH and Salome.

Thank you!

Kind regards,

Eugen

Sapsan | Thu, 03/01/2012 - 09:52

Hi Christian,

yes, I tried this mesher. The simulation is still running...

I'm very curious to see the results.

Regards,
Eugen

Sapsan | Thu, 03/01/2012 - 12:59

Hello altogether,

I have improved the mesh, but the problem with the ball adhesion underneath the lifting bar still remains.

Does anybody have any suggestions?

I will attach the new STL-file and the picture of the new simulation.

I will appreciate your help.

Regards, Eugen