kmokstad
commented
1 month ago There is already some inconsistency in your model TestFile_localCoord.fmm in the attached zip-file:
- The model file says the two spiders are
Rigid - The ftl-file it uses (
TestFile_localCoord_RDB/link_DB/FEM_60x60x6_Square_noSpiders.ftl) contains two WAVGM elements instead, which correspond toFlexiblespiders.
When you run this model, it works, but it uses the flexible spiders and not rigid ones as you may think then looking at the Triad properties. When you the try to switch to Flexible spiders from Rigid, it will try to delete the existing Rigid spiders first, which are not there, and then add the flexible spiders. It does not detect that the FE model already have flexible spiders since they are not registered in the model file, and therefore you end up with a double set of them. I guess the reducer does not handle this very well.
So the key question is how you ended up with that inconsistent model file, some manual editing perhaps? It points to the file FEM_60x60x6_Square_noSpiders.ftl as the ORIGINAL_FE_FILE (and not the nastran file). The name is confusing since indeed it does contains some spiders.
If I try to do this from scratch, import the Nastran file, then 1) add a rigid spider at one end, 2) switch to flexible, 3) switch back to rigid, and saving the model between each step, everything looks as expected.
Starting with an empty FE part, adding spiders manually, gives the option to switch between RBE2 and RBE3 spiders in the menu field of the triad that is attached to the spider joint.
However if you toggle from Rigid to Flexible the I'm not able to get the "flexible" setting to work. Solver crashes. Toggle back to Rigid and it will solve, but the rotational-stiffness of the spider seem to be gone? At least my beam just rotates after gravity.
The .ftl file in the reducer RDB directory will get double entries of spiders after some selecting back and forth: Starting out fine: WAVGM{2781 2801 2213 2214 2215 2216 2639 2640 2641 2642 2781 2782 2783 2784 2785 2786 2787 2788 2793 2794 2796 2797} BUSH{2782 2802 2801} CMASS{2783 2801} WAVGM{2784 2803 2209 2210 2211 2212 2493 2494 2495 2496 2635 2636 2637 2638 2789 2790 2791 2792 2795 2798 2799 2800} BUSH{2785 2804 2803} CMASS{2786 2803}
But ending like this: WAVGM{2781 2801 2213 2214 2215 2216 2639 2640 2641 2642 2781 2782 2783 2784 2785 2786 2787 2788 2793 2794 2796 2797} BUSH{2782 2802 2801} CMASS{2783 2801} WAVGM{2784 2803 2209 2210 2211 2212 2493 2494 2495 2496 2635 2636 2637 2638 2789 2790 2791 2792 2795 2798 2799 2800} BUSH{2785 2804 2803} CMASS{2786 2803} RGD{2793 2809 2213 2214 2215 2216 2639 2640 2641 2642 2781 2782 2783 2784 2785 2786 2787 2788 2793 2794 2796 2797} RGD{2794 2810 2209 2210 2211 2212 2493 2494 2495 2496 2635 2636 2637 2638 2789 2790 2791 2792 2795 2798 2799 2800}
Then model starts to behave strange.
Adding model, I even observed that it got 4x WAVGM elements when re-selecting "Flexible" again, but then solver is not able to start at all.
Adding model TestFile_localCoord.fmm file is working with only 2 WAVGM elements in total, TestFile_localCoord_02.fmm gets duplicated spider elements.
To re-create issue from scratch, make new model and use attached .nas file.
01_ShellElementTest.zip