CalculiX: load is not applied to the part of the mesh

About the development of the FEM module/workbench.

Moderator: bernd

Post Reply
PhysShell
Posts: 4
Joined: Mon Sep 20, 2021 2:56 pm

CalculiX: load is not applied to the part of the mesh

Post by PhysShell »

Good day. Recently started using CalculiX for doing some basic FEM and got stuck when applying load to the mesh.

Mesh: https://i.imgur.com/0K9ylYe.png;
With load applied (screenshot during animation): https://i.imgur.com/h3jxDqf.png
Project sources: https://fex.net/s/vzyn7eo

Problem is that the load is not applied to the bottom surface and "orthogonal (to the loaded side) segment" of the mesh. It just goes through.

My best guess is that there might be a not-welded-vertices thing and that little part, as well as the bottom, is not "attached" to the main part geometrically.
Another thing worth mentioning: mesh was created manually. And I'm thinking if it is required to build mesh in FreeCAD (using extrude and such) to get the proper geometry.

If someone could point out the mistake or point in the right direction, I'd be really grateful.
Thank you.
User avatar
NewJoker
Veteran
Posts: 3015
Joined: Sun Oct 11, 2020 7:49 pm

Re: CalculiX: load is not applied to the part of the mesh

Post by NewJoker »

From the picture with the results, it appears that some parts in the model do not have a continuous mesh with the rest of the geometry. You will have to merge the nodes or use tie constraint/contact. Modal analysis will show you which parts detach from the rest.

Can you send a step file with CAD geometry ? It would be best to mesh it in some preprocessor.

Here are the results of the modal analysis showing which parts disconnect:

modal 1 v2.JPG
modal 1 v2.JPG (59.31 KiB) Viewed 2077 times
modal 2 v2.JPG
modal 2 v2.JPG (106.49 KiB) Viewed 2077 times
modal 3.JPG
modal 3.JPG (80.03 KiB) Viewed 2077 times
PhysShell
Posts: 4
Joined: Mon Sep 20, 2021 2:56 pm

Re: CalculiX: load is not applied to the part of the mesh

Post by PhysShell »

Thanks for quick response. I thought it would have something to do with "non-continuous" mesh.
NewJoker wrote: Mon Sep 20, 2021 6:16 pm Can you send a step file with CAD geometry ?
Actually, getstart.fbd was made by hand :roll: . Now that I read the answer, I think that's the source of this behavior. So I'll try doing it in FreeCAD today and post the results here.

But I'm cursious to ask whether it's possible to fix in fbd-file? For example, by using "merg" command or "send areampc" or some other ways?
User avatar
NewJoker
Veteran
Posts: 3015
Joined: Sun Oct 11, 2020 7:49 pm

Re: CalculiX: load is not applied to the part of the mesh

Post by NewJoker »

PhysShell wrote: Tue Sep 21, 2021 3:55 am But I'm cursious to ask whether it's possible to fix in fbd-file? For example, by using "merg" command or "send areampc" or some other ways?
You’d have to refine the mesh, add more nodes and then merge them. Tie constraint can also help. But you will need sets to define this.
PhysShell
Posts: 4
Joined: Mon Sep 20, 2021 2:56 pm

Re: CalculiX: load is not applied to the part of the mesh

Post by PhysShell »

NewJoker wrote: Tue Sep 21, 2021 7:43 am You’d have to refine the mesh, add more nodes and then merge them. Tie constraint can also help. But you will need sets to define this.
Looks like I ran into similar situation when working in FreeCAD 0.19 as I was trying to re-create geometry. Here's the link to project file: https://www.dropbox.com/s/9umatluaamnwm ... FCStd?dl=0.

Geometry:
Image

When building this little box (the red arrow points to) out of sketch mesh building in GMSH fails. Without it everything runs correct.

17:23:01 Error : Invalid boundary mesh (overlapping facets) on surface 13 surface 15
Error : No elements in volume 1
Error : Mesh edges from different curves share nodes: create a finer mesh (curve involved: 32)
Error : ------------------------------
Error : Mesh generation error summary
Error : 0 warnings
Error : 3 errors
Error : Check the full log for details
Error : ------------------------------

Then I used Refine Shape in Part → Create a copy → Part → Refine shape as you suggested and it now shows different error:

Warning : No tetrahedra in region 1
Error : No elements in volume 1
Warning : Surface mesh: worst distortion = -8.37038e+13 (avg = -2.89632e+11, 2 elements with jac. < 0); worst gamma = 1.71996e-43
Error : ------------------------------
Error : Mesh generation error summary
Error : 3 warnings
Error : 1 error
Error : Check the full log for details
Error : ------------------------------

The project after "Refine Shape" if needed: https://www.dropbox.com/s/m1co8ustl93m2 ... FCStd?dl=0
User avatar
NewJoker
Veteran
Posts: 3015
Joined: Sun Oct 11, 2020 7:49 pm

Re: CalculiX: load is not applied to the part of the mesh

Post by NewJoker »

Delete this one line from the second sketch and it should work:
F4D50D06-6193-414F-B205-647A522EA7BA.jpeg
F4D50D06-6193-414F-B205-647A522EA7BA.jpeg (36.53 KiB) Viewed 1921 times
PhysShell
Posts: 4
Joined: Mon Sep 20, 2021 2:56 pm

Re: CalculiX: load is not applied to the part of the mesh

Post by PhysShell »

I think I messed the geometry up at some point, so it didn't help.
Solved this by using boolean operations (difference) on primitives. Showed the warning
Warning : 48 ill-shaped tets are still in the mesh
CCX didn't fail nonetheless. Results seems ok:
Image

Thank you NewJoker for your help.
Post Reply