"Warning The number of parts is too large. Only the first 9900 parts were created

General Questions
1

Hello, Im trying to do a static structural analysis on a fairly large model. After the analysis finished I get the warning in the title and results will also be zero.

The setup will be pictured below(sorry for the quality, cant upload more than 1 file as a new user).
I already tried using ordinary BC’s instead of the compression only. In this casse there is no warning but the solution is still zero.
Pls help.

In the mean time im running the setup with a smaller Mesh ( i wont use symmetrie bcos the symmetrie plane is intersecting with a node set used in a rigid body Constraint which caused problems in past analysis)

Troubleshooting
Troubleshooting708×964 163 KB

2

Can you share the .pmx file (just delete the results so that it’s not too large) ? If it can’t fit here, you could share a link to some hosting website.

3

Thx for the quick answer, pmx file should be here:

https://limewire.com/d/Q1Ftv#tBAif7HctF

PS: thx for your posts and your youtube videos, i learned alot from them already

1 Like
4

If you want to use compression-only constraints, you should set the nonlinear option to Yes or just enable Nlgeom. However, it’s better to simplify the model for now. Can you also share the version with boundary conditions instead of compression-only constraints ? Did you check if the model is constrained properly using frequency analysis ?

5

Nice example, I would start with a very coarse mesh and once the bc are ok start to reduce the elements size and only in the stressed components. Could you share the geometry files as well?

6

Analysis with fixed BC;
https://limewire.com/d/6NYPV#TqqMiE1OAx

7

these are the .geo aswells as .inp files. If you open the .geo file with an editor, you can scale the global mesh size using the L variable (alltough i have not tested that yet, changing L might screw up some of the logic of the transfinite functions)
https://limewire.com/d/6NYPV#TqqMiE1OAx

8

The problem with this model is that it has pressure load applied to the same face as rigid body constraint while the latter makes the surface rigid.

But for debugging, you should generate a coarser (and likely lower order) mesh. This one takes quite some time to solve even with Pardiso and 16 cpus.

9

I did some troubleshooting today. The main problem that the solution is equal to zero was solved by either using linear elements or (with quadratic elements) to not merge the parts in prepomax. However when I imported the mesh with the planned mesh resolution the parts are merged automatically in the Results page and the results are still zero. (the warning in the title refers to the gap elements, if i cant find a solution to my results = 0 problem, i will open a new more precise thread about it)

Edit:
the parts always appear as if they were merged in the results page, however the coarser mesh gave a acceptable solution while he fine mesh did not(results=0). However when i run both Analysis the one with the coarser mesh used up all available RAM and took around 15 minutes while the fine mesh never used more than 50% of the available RAM and only took 3 minuntes.