Can you Attach your *.sldprt (or a similar mock-up exhibiting the same behavior) here?
Are you looking at an exaggerated deformation scale?
Do you have enough constraints (or a soft spring) to make the part static is space?
Hi Mr. Mather,
Thanks for your reply! I've attached a mock-up file showing the issue on a simplified part. It does exhibit the same behavior.
Those photos were indeed showing an exaggerated deformation of about 3000x.
I'm still developing my competency with FEA but I do understand the importance of proper constraints. I constrained three faces constituting three orthogonal planes to in-plane translation. That should be adequate, no? I did not apply any soft springs.
I'm thinking about just disregarding the FEA results around those constraints. A convergence test showed pretty clearly that I have a singularity there (stress goes up up up as the mesh is refined). But the results in other areas away from the singularity should still be ok, right? In my convergence study I probed stress on another face that has relatively high stress and it converged nicely, can I expect that it converged to a realistic value? Areas away from stress concentration zones show similar stress to what I get from sigma=F/A, but the F I'm using in that hand calc is the reaction force I get from the FEA results.
I'll do hand calcs to get stress in the region I'm disregarding from FEA.