Divergence of results prepomax – stress concentration in fillet

General Questions
1

Hello everyone! I hope you’re all doing well!

I’m having trouble solving a simple problem in Prepomax. I created this model mainly to compare accuracy. In this comparison, I noticed that the peak stress in a T-shaped part with fillets differs significantly from the answer.

In both cases, I am using linear mesh refinement (first-order) with 0.5 mm, a Poisson’s ratio of 0.3, and a yield stress of 200 MPa. The image below shows the results.

The results only became similar when I practically removed the effect of Poisson’s ratio in the Prepomax material, setting it to 0.01 (which seems quite strange). The images below show these results:

image
image908×854 90.5 KB

I would like to ask if this issue is common or if there is any parameter I need to adjust in EDIT CALCULIX KEYWORDS? The constraints in the model are identical in both cases.

Thanks in advance!

2

What answer ? Analytical calculation ? Does it take Poisson’s effect into account ?

3

@erosa
I’ve 2 remarks:

  1. it’s not a prepomax concern but ccx (solver)
  2. my advice: when out of plane loading is involved (typically bending): use 2nd order element (C3D20R or C3D20 if you’re using 3d elements)
  3. (you may also have use a bias in the mesh from the middle surface to the skins)

As previously said: what is your source?

4

I moved this topic to general questions since it is probably not a PrePoMax bug.