Seismosoft | Earthquake Engineering Solutions

Dear seismosoft/huffte
I participated many times before, i have a concentrically braced steel frames of 6-9 & 15 storeys. I finished the part related to pushover analysis by using the link element to mimic the real behavior of braces( strength reduction in compression) . Now i'm at the stage of performing time history dynamic analysis and my main concern now is the Damping, actually i want through the help and the related discussion in the forum but i still have some doubts and have some questions :
1- in the damping settings if we are using the stiffness proportional damping, the fundamental period that we have to input is the elastic value or the inelastic one ?( ie from the eigenvalue analysis we obtain the elastic period, so shall we transform it to inelastic one and input it in the damping settings?)
2- for such steel braced frames what are the best damping value to use( 0.5% ,1%, 2%?)
3- i'm bit confused about the damping at the element level, so as it is mentioned in the help that the damping is taken into account in the formulation of the frame elements and links , and the other part is considered in the project settings /damping..... , so what is the function of damping at element level? And when shall i use it?

Best Regards and thanks,
Hani

Hi Hani. For nonlinear response history analysis, the overall current belief is that tangent-stiffness-based damping is more accurate than initial-stiffness-based damping.

There is an excellent discussion on this in the SeismoStruct Help section, Project Settings - Damping. Perhaps you have already read this.

I am far from being an expert on this, but it seems that using the initial period to compute the required constant, specifying a small amount - say 0.5% - of damping and selecting initial stiffness as opposed to tangent stiffness would be one way to mimic tangent-stiffness damping with a more stable solution.

If tangent-stiffness based damping is attempted, I believe that the period used to compute the coefficient should still be the initial period, but I am not certain.

I'll look forward to hearing other responses to this important question as well.

Very interesting issue indeed. There is a related article that probably gives explanation in details of this subject (although i have only read its abstract..., because i have no access on this source).
http://ascelibrary.org/doi/abs/10.1061/ ... 4%28581%29
If someone has any link on this or on other related articles, or generally if someone more experienced on this subject could post his option (prof. R. Pinho maybe..?) it will be realy helpful.