• Registered: 16-04-2013
  • Posts: 1

Topic: Simple test

Dear developers,

First of all I want to congratulate you for all your work. Just amazing! Very nice program you have here.

On the other hand I was doing a very simple test to compare the results I have from OOFEM and DIANA and to learn the basics of exporting to Paraview. I attach the example I was performing.

For the outputs, when I choose only displacements variable and stress variable it runs just fine, but when I choose more relevant primvars and vars such as the principal stresses, principal strains, total load level, internal forces (to plot a force-displacement curve at one node and compare principal stresses). I don't know if I'm doing it the correct way since I'm new at Oofem and receiving errors. Could you point out my mistakes?

Thanks in Advance! Best Regards!

BWO

Post's attachments

masonrycontrigidnolin.txt 95.58 kb, 6 downloads since 2013-04-18 

You don't have the permssions to download the attachments of this post.
  • From: Gothenburg, Sweden
  • Registered: 08-09-2009
  • Posts: 1,419

Re: Simple test

Hi BWO

There are a few things (which are mostly our faults as developers)

1.
I'm sorry to bring disappointing news, but there is unfortunately no way to visualize the load (or reaction forces) in the VTK output. The good news is that this is definitely a planned feature (I'm currently working on having all engineering models support the concept of External and Internal "forces" in order to make this general).

2.
TotalLoadLevel is not actually a primary unknown (or even a per-node-data), so it can't really be stored in the VTK file in any nice way. I think you will need to keep that in the normal output file.

In the latest version in GIT, I've actually removed a few values in UnknownType, since they weren't actual unknowns (the VTK export didn't know how to deal with them anyway, and they shouldn't have been there). The VTKXML export has just a few of them (which makes sense to export per node), like displacement, temperature, eigenvector, while other values such as "eigenvalue", "loadlevel" didn't belong there.
InternalForcesEBENorm was never actually use anywhere, and was actually intended to be the norm of the internal element forces (used when computing the relative error in the newton solver).

3.
Also, you have 5 values for "vars" but you specify only a length 4, but that aside, it seems the material model in question doesn't really like to export the principal stresses or strains. I would say this is a bug in OOFEM (without having had time to dig into in more detail yet).
StructuralMaterial fails to take into account that you only get 2 principal values when exporting plane stress (or we should pad it out with an extra zero).
I'm currently fixing this, so the GIT version will soon have this fixed (until then, you'll have to do without principal stresses and strains)

  • Registered: 16-04-2013
  • Posts: 1

Re: Simple test

Dear Mikael,

Thanks a lot for the quick response! It's good to hear this since I can point my analysis with a different material model or try a different output. You really saved me a lot of time. Will keep you posted if I can find a way to show these results. Thanks again!

BWO

  • From: Gothenburg, Sweden
  • Registered: 08-09-2009
  • Posts: 1,419

Re: Simple test

You actually only need to cut down on the primvars (just "1" = displacement vector will work in this case), and vars to avoid the damage and principal stresses (which should have worked, but there were some bugs).
Those bugs have actually been fixed in the GIT version now, so you are welcome to try that version out if you want.