Riks method cannot progress to the designated axial displacement after buckling occurred

[KyleSong]

Dear All users for Riks:
It is such a gospel to have Riks as a numerical tool to model structure buckling behaviour.

However, there is a puzzle for me, when I model a column with solid square cross-section under compression. I used displacement control as loading method.

the model can nicely capture the ultimate critical buckling load, but after the peak load, the displacement applied cannot progress, as I extract the data to make load- compression displacement curve.
To trouble shooting, I also drawn the load - lateral displacement curve of the center height of the specimen. From the curve, I can clearly see the large lateral displacement owing to the buckling.

As shown in attached illustration.

I found that after the specimen reached critical buckling load, the load kept same level and slightly decreased when the lateral deflection kept increasing. This make sense, but the axial displacement only grows very minimal after the critical buckling load occurred, THE ACTUAL AXIAL DISPLACEMENT CANNOT PROGRESS TO THE VALUE I ASSIGNED AT THE BOUNDARY CONDITION SETTING AS DISPLACEMENT CONTROLLED LOADING .

I understand that for riks there are three criterion as modelling stop criterion: Maximum load proportionality factor, Maximum displacement, and the maximum number of increments.But finally it's the arc length that determines the progressing of the modelling calculation. so finally, I should say why the arc length cannot proceed to the designated vale.

As a result, the post peak load decreasing happened just within very limited axial deformation, which results in a non-ductile behaviour.


Can anyone give me some idea why this is happening, how I can make the designated displacement fully reached during the modelling process.
Thanks everyone!


Regards,
Mengli

 

[Mustaine3]

Why do you use Riks when running the analysis displacement-controlled?

 

[KyleSong]

I just imitate the situation in experimental testing, which use displacement control.

Can you inform me what is the normal way to exert loading scheme?

By the way, the model give reasonable prediction for the modelling process before peak load.

Regards,
Mengli

 

[Mustaine3]

The loading used in the simulation depends on the situation in reality.

And again, when you run the analysis displacement controlled, then it is really rare that Riks is needed. Plot the load-displacement curve. If the displacement is always increasing, then you can use the normal static method.

 

[KyleSong]

If not displacement control, can you suggest me an alternative way to be the load scheme?

I want to nominate the compression load at the very beginning, but how to determinate the magnitude of this applied load, so that the program can break it down and carry out a series of increments?

Regards,
Kyle