I’ve been working towards creating the NAFEMS Benchmark Test: LE10
@chris has been helping with some bug-fixes & explanations over on the issue tracker (example 1, example 2)
It’s not quite there yet, but I thought I’d share my current WIP. Note that I don’t have the geometry matching the benchmark specification yet, I’ll update this post once I’ve got the true geometry matched. Also, no guarantee on materials – mostly I was just concerning myself with figuring out how to build geometry using the basic capabilities in Coreform Prep (see this issue).
This PDF is from a PowerPoint tutorial I’ve been putting together. If a user follows the tutorial step-by-step they should be able to repeat this simulation.
EDIT Fixed a few mistakes I had in the problem setup, leftovers from learning Coreform Prep’s selection modes.
OK. Here are updated results for what I would consider a “coarse” mesh (NAFEMS states to test with a coarse and fine mesh). I believe I’ve got the geometry accurately modeled, but wouldn’t mind a peer review.
Coarse mesh has a 4x4 planar layout and has 2 elements through the thickness
32 Total Elements
Elements are P2-C1
The interfaces in the Z=0 plane are C0
Results
Direct stress (sigma_yy) at the location specified in the NAFEMS benchmark computed as ~5.55 MPa. Scientific error to the target value specified is +3% which is surprisingly close given that I’m still only about 15% confident in my Coreform Prep & Analyze aptitude.
Future Work
I would like to do two things in the near future:
I’d like to run this in Abaqus, preferably the “V&V” input files that ship with Abaqus
I’d like to conduct some mesh refinement studies in Coreform Prep, but it seems to crash & corrupt my workspace when I try to add refinement cards via the “feature tree” tab. Guidance here would be most welcome.
