Truss results for frame building

Hello! In this section we'll talk about the post processing module and the data visualization.
Post Reply
Larisa GR
Posts: 34
Joined: Thu May 06, 2021 2:20 pm

Truss results for frame building

Post by Larisa GR » Tue Jun 14, 2022 2:52 pm

Hello stko team,

I'm working on a model for an experimental campaign and I need to design plastic braces to break during the test. The structure is a 3-storey steel frame building (4 columns + a rigid steel plate, no beam) to which I add the braces.
In order to calibrate the model (without braces) I run an eigen value analysis and compared the results with the modal identification from an analytical model.

After the calibration, I added the braces which are made of plastic and only working in tension. Therefore I implemented them as trusses (with special_purpose.truss), however, I do not manage to see any results on them (I can't see axial force (N) or strain/stress results). I don't know if that is due to the element definition or I missing sth.

I also know that merging the model or wiring the model is different and in the case of wire the nodes are not coupled in translation and rotation, rigth? In the model, I cannot wire since I have the steel plates as shells, so is merged, could this be a possible reason?

I've also checked the webinar of Francesca of week·#19 where the bracings are defined as beams with zero-length elements to remove the moment transmissibility, is this better than a truss? I did it but still, I cannot get any information from the bracings.

In addition, in the new stko version in the post-process there are results such as [E1] and [E2] forces with 10 and 24 components...where can I find the definition of each component??

I attached both models, one with trusses and one with beams+zerolength element.

Thank you very much for your help!

Larisa
Attachments
Numerical Analysis.rar
(284.22 KiB) Downloaded 17 times

STKO Team
Posts: 1346
Joined: Tue Oct 29, 2019 8:45 am

Re: Truss results for frame building

Post by STKO Team » Tue Jun 14, 2022 4:17 pm

The truss model is fine.

You just don't need to use a truss with a fiber section. The stress will be constant anyway for all fibers. So you can just use 1 material + Area.
In this way, you can ask for material.stress. You can view them in the surface plot, or in the gauss plot if you need to extract data:
TRUSS.png
TRUSS.png (461.83 KiB) Viewed 680 times
Pay attention to the brace material. You made an elastic multilinear. Is that what you really want? Even if it is non-linear, it is still elastic, i.e. it does not generate plastic deformation, but the unloading will follow the loading path.
Also, it is behaving linear in compression, while you said it should not work in compression.
Double-check it.
Frame_shell_brace_truss.zip
(180.69 KiB) Downloaded 18 times

Larisa GR
Posts: 34
Joined: Thu May 06, 2021 2:20 pm

Re: Truss results for frame building

Post by Larisa GR » Wed Jun 15, 2022 7:59 am

Hi,

thank you for the replay!

You are right about the compression behaviour, I should change the description of the material so the truss does not work in compression.
so two more issues/questions:
-If I don't make the fiber section of the truss I cannot extract the information in my results, the material.stress that you show in the image is zero.
-if I want the braces to break then I cannot use this multilinear material, I should use a DamageTC1D and assign it to my truss element right?

Thank you again!
Attachments
DamageTC1D.png
DamageTC1D.png (299.81 KiB) Viewed 666 times

STKO Team
Posts: 1346
Joined: Tue Oct 29, 2019 8:45 am

Re: Truss results for frame building

Post by STKO Team » Wed Jun 15, 2022 10:05 am

-If I don't make the fiber section of the truss I cannot extract the information in my results, the material.stress that you show in the image is zero.
Why? I see values there. In the image a posted are small, but just because I made a screenshot of the first steps. Are you sure?
if I want the braces to break then I cannot use this multilinear material, I should use a DamageTC1D and assign it to my truss element right?
It really depends on what you want to do.
For example:
Is it steel? Then you need a steel material, like the ElasticPP
If you don't want to make it work in compression, what do you mean? Are you trying to model compression failure due to buckling?, So if it reaches a negative deformation it should break? Then you can put the steel material inside a MinMax material.
Or should the material be able to resist tensile forces even if before if was in compression?

Larisa GR
Posts: 34
Joined: Thu May 06, 2021 2:20 pm

Re: Truss results for frame building

Post by Larisa GR » Wed Jun 15, 2022 12:09 pm

STKO Team wrote:
Wed Jun 15, 2022 10:05 am

Why? I see values there. In the image a posted are small, but just because I made a screenshot of the first steps. Are you sure?
I don't know but in the truss elements I can't see anything (the picture is of the results with the truss without fiber section)
no_result.PNG
no_result.PNG (251.63 KiB) Viewed 662 times
Regarding the material of the brace: my frame is steel but the braces are plastic and indeed they don't work in compression (or have small strength) they only work in tension (E=1650Mpa, σ=48 MPa, εelatic=2%, εbreak=18% as provided by the manufacturer).
The idea is to break the braces in tension when subjected to earthquake excitation, so as to design the proper cross-section of the brace to ensure the damage during the test.

STKO Team
Posts: 1346
Joined: Tue Oct 29, 2019 8:45 am

Re: Truss results for frame building

Post by STKO Team » Tue Jun 21, 2022 8:38 am

I don't know but in the truss elements I can't see anything (the picture is of the results with the truss without fiber section)
Ok so probably it's because you are using an old version of OpenSees when the truss elements only supported the "-material" keyword. Now they support both "-material" and "material".
Try to download the latest version of the OpenSees-Solvers.

Regarding the material of the brace: my frame is steel but the braces are plastic and indeed they don't work in compression (or have small strength) they only work in tension (E=1650Mpa, σ=48 MPa, εelatic=2%, εbreak=18% as provided by the manufacturer).
The idea is to break the braces in tension when subjected to earthquake excitation, so as to design the proper cross-section of the brace to ensure the damage during the test.
You can use a Hysteretic material:
Hysteretic.png
Hysteretic.png (102.01 KiB) Viewed 608 times

Larisa GR
Posts: 34
Joined: Thu May 06, 2021 2:20 pm

Re: Truss results for frame building

Post by Larisa GR » Wed Jul 06, 2022 1:46 pm

Thank you very much.

As suggested I downloaded the latest version of opensees, but still I'm not able to see the stress and strain in the truss element as a truss section not fiber. I don't know if I'm missing some step when installing opensees.

In addition, I change the material for the hysteretic one shown in the post before and I have a problem, the response of the truss brace (to which I assign the material) is not following the material, the response is linear and working in compression when it was defined not to work in compression. I attached a picture showing the response of the brace which stands -50MPa when it was designed not to, and also does not evidence signs of plasticity as it should.
error material.PNG
error material.PNG (308.99 KiB) Viewed 563 times
Is there any chance that this is connected to the fact that the truss is defined with a fiber section?

I also attached the model.
Thank you!
Attachments
Frame_braces_45_1.66_x2.rar
(140.03 KiB) Downloaded 16 times

STKO Team
Posts: 1346
Joined: Tue Oct 29, 2019 8:45 am

Re: Truss results for frame building

Post by STKO Team » Mon Jul 11, 2022 4:02 pm

As suggested I downloaded the latest version of opensees, but still I'm not able to see the stress and strain in the truss element as a truss section not fiber. I don't know if I'm missing some step when installing opensees.
I do not understand your problem.
In the attached file you are using fiber section for the truss, so you are visualizing the section.fiber.stress output. And it is visible.
What is the issue?

By the way, the truss don't have a curvature, so you don't really need a section. You can use the standard truss (not trussSection) using a material and an area (equivalent to a fiber section with only one fiber).
In that case you need to look at material.stress
In addition, I change the material for the hysteretic one shown in the post before and I have a problem, the response of the truss brace (to which I assign the material) is not following the material, the response is linear and working in compression when it was defined not to work in compression. I attached a picture showing the response of the brace which stands -50MPa when it was designed not to, and also does not evidence signs of plasticity as it should.
The problem is another... It's more than 50 MPa just because you selected many fibers and you used the SUM operation. So it's summing up the contribution of many fibers.
Try to select only 1...

Post Reply