Tension only bracing

[seismosoft]

Upol,

We had a quick look at your model.

You should:
- use 'truss' rather than 'infrmFB' elements to model the braces
- introduce link elements with an adequately calibrated 'gap_hk' response curve in dof F1, so as to ensure that only tension is transmitted to the truss elements. To the remaining dofs of the link element you should associate a 'lin_sym' response curve with with infinitely high stiffness values (i.e. 250 times larger than that of adjacent elements)
- introduce such link elements between the truss and the beam-column joint, making sure that the dof 1 is aligned with the brace

You may also:
- disabled the visualisation of your non-structural nodes (their extremely large coordinates force the rendering of the model to be excessively zoomed out)
- avoid the exaggerated discretisation of beams and columns (read the Help System)

Seismosoft Support

[smhzs]

Hi,

I am trying to model a simple 2D Braced Frame (1-bay, 1-story) with tension only X-braces based on your suggestions. I have two queries please:
1. I do not know why but I could not model the tension only braces just by defining 1 (only one) Link element for each brace. In order to get no compression, I had to put 2 links at the ends of each brace element (Truss).

2. By doing so, although there is no compression in the brace elements, surprisingly, the two braces show different behavior. Note that the simple frame is symmetric. I cannot find the reason for it. Can you please help me model the tension only braces accurately?

Thanks in advance,
smhzs

[seismosoft]

Smhzs,

We looked into your model and noticed that you have forced (through the use of a 'Equal DOF' constraint) your upper link nodes to be rigidly connected to the beam ones.

As such, the link elements that you have positioned at the top of the braces will evidently not work (since their two end nodes are being constrained not to move with respect to one another).

If you wish to enforce a rigid beam behaviour, you should constrain only the beam end nodes, not the link ones as well, probably through the use of a diaphragm constraint, instead of the one you have employed.

Seismosoft Support

[amy]

Dear All,
I was searching the forum to find the solution of my problem. But I didn't get any answer. I modelled a bolted connection. Bolts are modelled using truss element. The truss element will act as a tension member only. So I connected the truss element with lin-asym link element with very high stiffness in tension. At the same location I have rigid element which will act as a compression element only. The rigid element was connected by lin-asym link element at two ends with very high stiffness in compression. The model works perfectly in pushover analysis. But when I conduct dynamic time history analysis, the model is not working. The rigid elements take both tension and compression and the truss elements do not work.

I really appreciate if anyone can let me know why this is happening or how can I overcome this. I am attaching my model so that you can look at it.

Thanks
Amy

https://www.dropbox.com/s/cki5crqdvucvhj6/push.spf?dl=0

[amy]

Hi,
I have just conducted static time history analysis subjected to cyclic load. My model also worked for this. But problem is with the dynamic time history analysis. I think Lin-Asy link elements do not work for the dynamic analysis. Am I correct or am I missing something else?

[huffte]

Hi Amy. For the benefit of others who might run across similar problems, I post that you resolved your problem and that the link type referenced is perfectly acceptable in a dynamic analysis as long as the properties are properly selected. Keep us posted on your interesting work.

[amy]

Dear Huffte,
Thanks for your comments. According to your recommendation, after doing further research, I found that using gap-hook element at one end of braces will be perfect for tension only braces . Using asymmetric link does not work as the Eigen value analysis gives different values for the same braced frame.

Thanks
Amy

[hosseini_memory]

Hi
could you please help me.
I want to model a 2D frame with only tension X brace.
How can I model Pinching behavior of brace?
is there any special specially material or element for this?
is there any example to help me?
thanks a lot

[seismosoft]

For an example of what you are trying to model, please refer to Example 7 of the verification report and the corresponding files Ch2-07_A.spf, Ch2-07_B.spf & Ch2-07_C.spf in the samples files folder.
For the pinching behaviour you can select a link element curve with the behaviour you want to simulate

Seismosoft Support