I have a large (~250K shell and solid elements) assembly that I am seeking to do a dynamic time analysis on. I have successfully performed several static analyses and a modal analysis on the model. However, the dynamic time analysis always fails with an internal engine error on the "begin element calculations, pass 0" step after 200-300 seconds. Opening the *.pas file gives the following information:
*** A fatal error has occurred. ***
An internal error has occurred in file S:\sim\engine\src\as\dynamics\DYN_ModalStressRecords.cxx, line 973.
I did a full search of my computer C: drive for "DYN_ModalStressRecords.cxx" but found nothing. Command I used: dir /b /s *DYN_ModalStressRecords*
I have a few questions, answers to any would be extremely helpful!:
1. If you have any ideas in general, please let me know.
2. Where is the "S:\sim\engine\src\as\dynamics\" directory? I can't seem to said directory.
3. There were several temporary files made: "mpc.tmp", "mel1.bas", "oel1.bas", "kblk1.bas", and "kblk2.bas". Does anyone know their significance?
4. Are *.mda and *.mma files both mesh filetypes?
FYI, other references regarding similar errors are:
Thank you for any assistance!
Well, I am not sure why you cant find those directories. I haven't tried looking for those directories whenever I have such an error.
Yes, Simulate does create a lot of temporary files, and I do not know their significance.
It always creates mda files for analyses, but I don't know what those do. I know mma files are mesh files (so whenever you run AutoGEM and it asks if you want to save the mesh, if you choose Yes, and you have an assembly, it saves the mesh as an mma file, and mmp for parts).
Your best bet would be to file a Case with PTC support and send them the data set. I did a quick search of "An internal error has occurred" on PTC Support, and found many different articles, with varying issues, and varying solutions, so there's no hard and fast solution to this.
Hope this helps.
Thank you for the response!
So I did some researching and found out that those files are PTC source code files, so they mean nothing to use because we don't have the source code.
Also, it seems that mda files and mma files are equivalent, and mdb and mmp files. See How to view Mechanica Structure mesh.. WF2 . Moreover,
(1) I made a part
(2) meshed the part and saved the mesh elsewhere
(3) Simulated the part without using the mesh in step 2.
(4) copied the mdb file to the part folder and changed the extension to mmp.
(5) AutoGEM detected and loaded that new mmp file successfully as a mesh.
Also, I still am not sure why my model was failing in the time analysis, but I created a much simpler model (replacing many of my components with beam and mass idealizations), which, although not ideal, enabled it to run. Even with the simplification, the analysis file sizes were MASSIVE. With the unsimplified model, they would be even more massive.