Moment Curvature using zeroLengthSection

[Sylwester]

Dear STKO

I want to reproduce in STKO an example of moment-curvature from the opensees wiki page, but performing it on my own fiber cross section.
[https://opensees.berkeley.edu/wiki/inde ... re_Example]
I saw this example replicated using opensees navigator on youtube
https://www.youtube.com/watch?v=M14Mr8GpVRU.
An additional change I'm making to my model is that I'm not putting a concrete cover in, and I'm introducing a min/max model for steel to get a plot like the one below:

Moment Curvature Figure 2.jpg (20.8 KiB) Viewed 2367 times

I tried two ways:
(a) I inserted 2 nodes in the same place 1 [0,0,0] and 2 [0,0,0], then I connected them with the zerolenghtsection command, but that didn't work very well

b) I built a geometry with 2 nodes 1 [0,0,0] and 2 [1,0,0] unit length as in the video :
#07 Learning Opensees: Nonlinear Modeling of RC Frames (Lumped Plasticity)

I do not know why the zeroLengthSection element is created twice in my model (we can see it in the output elements),

elements.jpg (14.53 KiB) Viewed 2367 times

and why in example #07 the moment-curvature is obtained from the moment-rotation results.
I thought that I should export the M-Kz result and not M-Rz

In addition, the beamsectionProperty and displacement beam column are created in example #07.
My second uncertainty is the value of the target displacement (dk) in the integrator setting of the displacement cotrol.

In the opensees wiki example, this value is calculated.

openseewiki.jpg (10.85 KiB) Viewed 2367 times

In my example, I calculated
1) epsy = fy/E = 0.0025
2) Ky = epsy/0.7d = 0.0025/0.7*0.36 (h-cover)=0.00992
3) dk=ky/numInc = 0.00992/100 = 0.0000992
Is this correct?

my analysis.jpg (10.69 KiB) Viewed 2367 times

The program performs the calculation but if I do the M-KZ plot (with gauss plot) the moment I have only 1 kNm.

my gaussplot .jpg (13.27 KiB) Viewed 2367 times

What should I do to recreate the example from opensees wiki in STKO?
I want to perform the calculation:
a) the bending moment itself (considering that the section is a beam)
b) the effect of the axial force (considering that the section is a column)
I would like to know exactly at what force the bar will rupture or crush the concrete

crush.jpg (30.63 KiB) Viewed 2367 times

The same way it was done in opensees navigator. I am posting my file from STKO [rar file].
I will appreciate your help.

[STKO Team]

An additional change I'm making to my model is that I'm not putting a concrete cover in, and I'm introducing a min/max model for steel to get a plot like the one below:

It's ok, but double check your minmax material, your limit strain is exactly equal to the yield strain, so the steel response is elastic-brittle. If this is what you want, it's fine.

I do not know why the zeroLengthSection element is created twice in my model (we can see it in the output elements),

It's because you created 2 overlapping lines into a polyline. In the modified file I corrected it with a single line:

momcurv_1.png (92.81 KiB) Viewed 2340 times

and why in example #07 the moment-curvature is obtained from the moment-rotation results.
I thought that I should export the M-Kz result and not M-Rz

It's because, in both examples, the geometry has a unit length, so the rotation corresponds to the curvature (i.e. they have the same value).

My second uncertainty is the value of the target displacement (dk) in the integrator setting of the displacement cotrol.

In STKO you have to put the Target Displacement, i.e. the final value you want to obtain. Not the increment as you did in opensees. Why? because in STKO we support the concept of "Adaptive Time Step" to overcome the non-convergence issues. With the adaptive time-step you do not know in advance what would be the time-step, simply because it may change.


Furthermore, I found some mistakes in your file. Mainly you used all the necessary tools, but their order in the AnalysisSteps is not correct! keep in mind that the order you use in the Analysis Step is the order it will be executed in OpenSees.
This is the correct order

1. apply the constraints
2. add axial load
3. load control analysis for axial load - don't forget to keep the load const and reset the time to 0!!!!
4. add moment
5. Disp-control

In the modified file I also changed the file to a 3D model (instead of a 2D one), and I also added the monitor to track the moment curvature in real-time. I also had to increse the target rotation to see the complete failure

MOMENT CURVATURE.zip

(129.33 KiB) Downloaded 199 times

[Sylwester]

Thank you very much,
I just have one additional question:
why do I get a higher bending moment (load capacity) for a section with a lower height. In the revised file, the section under consideration is of lower height (0.4x0.2 m), I changed the height with the width to get 0.2x0.4 m . Shouldn't it be in the opposite way?
I attach the results below

Positive.jpg (125 KiB) Viewed 2281 times

negative.jpg (128 KiB) Viewed 2281 times

[STKO Team]

In the file I sent you (moment capacity of about 70) the bending moment is acting on the strong direction:

MZ.png (179.58 KiB) Viewed 2267 times