SFI-MVLEM

[pomelo]

Dear STKO Team,

I am an absolute beginner in STKO and have some problem modeling RC wall using SFI-MVLEM element.
Do you have some example files you can share on this?
q1. Do we have to assign FSAM (~physical property) to the wall or just element property(~SFI MVLEM) is enough?
q2. How can we record RC panel responses (inc. cracking angles) in STKO?

P.S. Really appreciate what you're doing!

[STKO Team]

Do you have some example files you can share on this?

Not now, but we are preparing materials for our next webinar (actually a webinar series) for modeling shear walls, and that element will be used as well.

q1. Do we have to assign FSAM (~physical property) to the wall or just element property(~SFI MVLEM) is enough?

No, just put the material into the material list in the element property

q2. How can we record RC panel responses (inc. cracking angles) in STKO?

in the MPCORecorder you can ask for a custom result:
RCPanel.ID.cracking_angles
where ID is a number from 1 to the number of micro-fibers. Make sure to use a valid number, because it is not checked inside the OpenSees's code!

[pomelo]

STKO Team,

Thank you for your response.
Good to hear that the webinar series for modeling RC shear walls is coming soon. Could you please let me know the tentative release date? Looking very much forward to it!

I tried to model the example SFI-MVLEM tcl file provided in opensees web in STKO but I keep getting the error message 'Element height is zero!'. I don't know where I messed up.
I have attached the scd file and the tcl file (from web). I'd be grateful if you could go through it.

Thank you again.

[STKO Team]

Could you please let me know the tentative release date? Looking very much forward to it!

It will be a webinar series from July to the end of August.

I tried to model the example SFI-MVLEM tcl file provided in opensees web in STKO but I keep getting the error message 'Element height is zero!'. I don't know where I messed up.

It's because you modeled a 2D model in the X-Z plane. In OpenSees it should be in X-Y plane.
So you should just rotate your geometry 90° about the +Y axis, to put it back to the X-Y plane.

Then, you will see and error due to an incompatibility between the MPCORecorder and the MVLEM element. You should switch off the section.fiber.*** results. We will fix it for the webinar.

Finally, in the disp control you should use adaptive time step because it may not converge so easily

[pomelo]

STKO Team,

Thank you for the help.
I followed your suggestion and the model runs now. However, I still got the error message saying the analysis did not converge.

I switched off the section.fiber. results.

Turned on recorders
nodal results: displacement, reaction force
element results: force, deformation

Thank you again.

[STKO Team]

Some material parameters were not matching the ones used in the examples (I changed them, for example the "nu" in FSAM).

But the main error was that you were using the Cyclic timeSeries for the loadPattern of the horizontal load.
Keep in mind that
In displacement control analysis, the load (controlled by the analysis) MUST use a linear time series!

Here's the working file

geometry.zip

(112.6 KiB) Downloaded 224 times

[pomelo]

Dear STKO Team,

Thank you so much for the solution.
I have a few more queries about post-processing the results.

1. to record the shear deformation I used "ShearDef" to the "custom Element Results" in the recorder. Could you please let me know how I can view this data? I opened the file in HDFView as well, but couldn't find ShearDef data.

2. For the cracking angles, I used
RCPanel.ID.cracking_angles
where ID is a number from 1 to the number of micro-fibers.
In this example, I have used 5 microfibers (m=5). So I added RCPanel.1.cracking_angles to the custom element results. However, I get the error saying "number of recorder input for RC panel is 0.

Thank you so much again!!

[STKO Team]

Unfortunately, this element does not follow the rules we need to make it work with the MPCORecorder. However, we talked with the MVLEM developer, and we will fix this issue just before doing the webinar series on shear walls. Stay tuned! we will let you know

[pomelo]

STKO Team,

I was modeling a RC wall using SFI MVLEM.
It converged, but I have some concerns.
1. The load-disp is not symmetrical. I concur that it can't be perfectly symmetrical. But the difference is evident, even at the low drift (0.125%)
2. Force-disp relationship kinda looks like it has shifted down (drift 1.5%, 2%, 3%). Underestimating for positive drift and overestimating for negative.

Tried to check/change the model, but couldn't solve the problem. It would be great if you can look into it.

[STKO Team]

Try this with the latest version of STKO and the OpenSees-Solvers (from prof. Tarque's website):

NSW6_29_3D.zip

(138.38 KiB) Downloaded 232 times

We recently implemented the SFI_MVLEM_3D which works directly with surfaces (like shells) and there will be a webinar about this the 24th of August.