link element deformation vs attached element'sone

[hani.akkari]

Dear Seismosoft/ Huffte
I contacted you before for some issues in the modeling of my concentrically braced steel frame and you responded to me. Actually I finished modeling that CBF using as braces INFRM FB element having link element on its sided to mimic the behavior of real brace , I didn’t use the inelastic truss element as I can’t get benefit from the performance criteria part where only it accepts frame elements.
My problem is that after a static pushover analysis, I got very low deformation in the mimicked brace member , I was so surprised that at the last step of the pushover analysis still no buckling appears in these members , until I went to the results at the element level, for a brace and to its attached link elements , I found that the deformations in the brace is completely different from the link deformation , the link deformation is logical as values but those of braces are not logical values (3-4 mm) at the last step , meanwhile the attached link has 80 mm !!!! , sure the deformation inside the brace is wrong, but is the deformation in the link is correct? And how can I match the link with the attached brace to have the same deformation?
Attached is a snapshot of the mentioned values and the complete model, your help is highly appreciated as I’m in an advanced step in my thesis and stacking here may lead me undesirable consequences.
Thanks a lot and waiting to hear from you soon.
https://dl.dropboxusercontent.com/u/339 ... 0model.zip

[hani.akkari]

Dear Seismosoft/Huffte,

attached is a simple model 1 bay,1 floor CBF that i produced (for simplicity), showing the same problem as above , if you want to check the smaller model , feel free as it is easier to manipulate .


https://dl.dropboxusercontent.com/u/33925535/tryyy.spf

Reagrds,

Hani

[huffte]

Hi hani.akkari.

It seems to me that there is no discrepancy in your results. At least, this is my interpretation of what you have modeled.

I checked both models and I don't see any reason that the two deformations you compare should be equal, or even close to one another. The links did yield in the axial direction, which I think you would interpret as buckling. And the frame element deformation I believe, should in fact be smaller than the link deformation as reported.

I interpret the brace element deformation as the shortening or lengthening of the brace, whereas the link deformation is an absolute deformation in space. For example, if the links at each end of a particular brace happened to move 100mm in exactly the same direction, then the brace deformation would be zero - no axial deformation within the brace.

I also checked the brace force against the attached link force and they also appear to be consistent.

Further, the compression links (L2 and L4) yield (buckle) at a load factor of about 1.9. The tension links (L1 and L2) yield at a load factor of about 4.5. This should be encouraging as well I think. Perhaps you could do a hand calculation to estimate the buckling load in the compression braces and see how this compares?

So, as I understand it hani.akkari, the model is consistent. Whether or not it represents the behavior you wish to study is for you to answer, but the model seems to me to be producing reasonable results.

You may also estimate the brace deformation by PL/(AE) and see how this compares to the force reported in the model, given the deformation reported in the model, and I think you'll find reasonable agreement in the linear range.

It is a clever model, by the way. But you would be wise to take perhaps your simple model and estimate the behavior using hand calculations to see how the results match with SeismoStruct and to see if the model is indeed acting as you intend.

Best of luck with this interesting work.

[hani.akkari]

Thanks a lot Huffte.[:)]