Issues with Buckling Analysis with Riks Step

[kpajunen]

I am running a Static, Riks step in Abaqus Standard. I am analyzing a 3D solid geometry. When compressing the geometry with a negative load at the top (connected to the structure using coupling constraint) and fixed at the bottom, I know that several of the members are supposed to buckle. This is why I am using a riks step.

- The stopping criterion for the Riks step is a maximum displacement of the loading point of -20mm (the structure is around 57 mm in height).
- NLGEOM is on.
- Using automatic incrementation, 1000 maximum number of increments.
- Initial, minimum, and maximum arc length described next.

Issue:
The load-displacement response is incorrect. The LPF starts out positive, but then become negative until the structure is being stretched, not compressed. Please see attachments. I have tried many combinations of initial, minimum, and maximum arc length increments, but I still can't get it to compress correctly... When I use an initial arc length of 0.1 and maximum of 1, for instance, this occurs. I find I need to increase the maximum arc length, because when I constrain it to a small value (10 or less), the maximum arc length is used every step and the LPF goes more and more negative.

Are there any suggestions for how to pick the arc length increment parameters to get the response to exhibit the correct response??

 

[StefCon]

Hi,
The peak of your LPF is your maximum buckling resistance (looks correct to me).
Is your structure really being stretched (check displacements)?

If you try to do a general static step and try the same thing it should stop before the slope of the LPF gets negative, right?

 

[Mustaine3]

Have you checked the behavior when running it displacement controlled?