Seismosoft | Earthquake Engineering Solutions

Dear Seismo

I modeled one 3D bare frame with inelastic displacement based element for columns and elastic element for beam as I knoe the beam element would remain in elastic range, also the beams are hinged ended beam ,I analyzed the system with nonlinear dynamic time history analysis with 12 s earthquake record but I have the unexpected base shear versus displacement curve it is a bit illogical ,more I divide the column in 2 parts I consider the length of the first part is 2 times of the height of cross section , and i control the behavior of column after analysis with the frame hysteresis option and the moment - rotation curve which indicates that the moment - rotation is in the linear range or not actually I control the damage with the value of rotation ) ,is it true? further I send my model with mail , pleas guide me and correct my errors .

What do you mean illogical curve? Note that in dynamic analysis the curve is not smooth like in pushover analysis or static time-history analysis.
The second part of your question is not very clear. Can it be that you apply a small earthquake load and the elements are still in the elastic range?

SeismoSoft Support

Dear khajehdehia,

I would suggest you to start a simpler version of the model, without the presence of very small elements (to model plastic hinges) and without end-releases.

Check if things run ok with such simpler model, and then gradually introduce the new modelling featuring (end-releases, etc) and see if and which of them causes the numerical inconsistencies that you seem to be observing.

Rui

PS: use a tight displacement-based convergence criterion of at least 10^-4 m/rad.

Dear Rui

I send my model to you pleas control my model
thanks alot for your guide

khajehdehia,

I repeat what I have said before; simplify your model (if you remove e.g. the end-releases, the higher-modes in the base shear response disappear), re-run the analysis, compare the results, and start understanding what is going on.

Rui