Uniaxial tensile test: modeling of localized section loss

General Questions
1

Is it possible to model the uniaxial tensile test in PrePoMax in such a way that the necking of the steel bar is reproduced as in the image?
imagen

2

Yes, check this CalculiX example by Martin Kraska: CalculiX-Examples/Drahtbiegen/Zug at master Ā· calculix/CalculiX-Examples Ā· GitHub

1 Like
3

Thank you so much. I will study it and try to implement it.

4

Could it be that this example has changed? I remember seen a very little deformation on the mesh to force the necking (even was parametrized), but now I canĀ“t find any reference.

5

Perhaps you are talking about this example: CalculiX-Examples/NonLinear/TensileTest at master Ā· calculix/CalculiX-Examples Ā· GitHub

6

Exactly! Amazing memory. Now, why the first example needs an imperfection to create the necking and the second one donĀ“t?

7

Without the imperfection, necking may not occur or may occur in unrealistic locations and itā€™s very sensitive to various analysis conditions. Even in that example: Elasto-plastic bending - elements, accuracy - #20 by Disla - CalculiX

8

According to my last tests , the necking initiation and final PEEQ value is equally sensitive to the time step as much as the imperfection size. In case one use imperfections, I would introduce them as small as possible not to alter the result. In fact, in Mr.Kraska study case, imperfection is not essential if one choose the right time step. I encourage you to experience it by yourself.

Another interesting finding is that the result depends on the loading pattern. How much and where the plasticization zone appears depends on the shape of the ā€œAmplitudeā€ function.
I think it has to do with the fact that PEEQ is a cumulative value. I have seen that too long time steps in the wrong stage lead to incorrect results regardless of the mesh size.