Modified Ibarra-Medina-Krawinkler convergence problem

[RobertoPhd]

Hello
I’m modelling a 2D simple frame (2 columns and one beam), fixed at the bases. In order to perform a pushover analysis I put two plastic hinges at the beam ends (30 cm from the end) and a plastic hinge at the column base (20 cm from the base). The hinges are given as follows:
- Beam’s hinge: finite length non-linear link - material: Pinching4 (Moment-Rotation);
- Columns Hinge: zerolength element – material: Modified Ibarra-Medina-Krawinkler deterioration model with Bilinear hysteretic response (Moment-Rotation).
On this model, I want to perform a static cyclic analysis, in displacement control, and for the beginning, I set all the deterioration parameters, both for the beam and columns hinges, to zero. I use “NewtonLineSearch” as algorithm, and the DisplacementControl integrator.
The model works perfectly until the columns stay in the elastic range.
My problem is this: once the displacement of the top of the column goes beyond the columns yielding limit, and the imposed displacement starts to decrease (increasing in the other way), the structure does not obey to the displacement history that I give, but keeps on going in the same way for some more centimeters and than fails to converge.
About this problem I carried the following checks (without succeeding):
- I made a model of the bare column (precisely copy and paste of the code) imposing a cyclic displacement on the top of it, and it works perfectly;
- I made a model of the bare beam, fixed at one end as a cantilever (copy and paste of the code), imposing a cyclic displacement on the free end, and it works perfectly;
- On the frame model, I changed the material of the columns to the same of the beam (Pinching4) and the model works perfectly;
- On the frame model, I changed the material of the columns to the Steel01 material, so to give exactly the same Moment-Rotation law that I give with the IMK material (which hasn’t got any degradation), and the model works perfectly;
- On the frame model, I tried giving a smaller step size where the column starts going in the plastic range, just before unloading, and here also the structure does not obey to the displacement history that I give but basically sends in few steps the plastic hinges of the columns to a zero resistance and a very advanced rotation (that brings the structure to a very advanced displacement). In this case I give a bigger step-size until the columns are about to go in the plastic range, and just before I reduce the step size. I attach some images (in the links below) to explain better this case. Note that in these images the Displacement history should not have such a snap (blue line) and the columns hinge should keep on staying plastic for a much bigger rotation.
From these checks I understand that the problem stays in the IMK material, even if I don’t get any problem with the bare column model, but only in the assembled structure model.
Can anybody help me?
Thanks in advance

Images:
https://ibb.co/x5xxYbR
https://ibb.co/S6kz77h
https://ibb.co/f4jWdY3