seismic analysis tips

[asad93]

hi few days ago i have posted one question and wondered about answer. i am worried about either my question was not well posed or otherwise. let me frame my question from other perspective as i am beginner in LSDYNA so can anybody expert if FEM and LSDYNA also give me some tips of seismic analysis in LSDYNA.
Thanking in anticipation

 

[deepquest06]

Hi,

Presently I am working with ANSYS/LS-Dyna. My problems relates to seismic analysis of steel joints. Like asad93 I am also a beginner with LS-Dyna.

I tried a some standard problems like a SDOF system, a single bay portal and continous column. I applied the seismic loads as acceleration time histories (Using AX DOF) to the ground nodes. The ground nodes are fixed. I am not sure if I am doing it correctly because the results do not make any sense. The resultant accleration at top of the structure should amplify and the structure should act as a filter, i.e., the resultant accleration history at top nodes should have frequency content close to structural natural frequency. But it is not happening, instead I am getting rigid body modes.

Can anybody in the forum throw some light on how to analyse a structure for a seismic ground motion using ANSYS/LSDyna or LS-Dyna.

Thank you all.

 

[asad93]

Dear deepquest06 according to your description of problem analysis i tried to solve simple frame but by applying damping also to frame. results have some sense but i am not sure as at the end of seismic event structure shows eventual displacement of large magnitude. Although i tried material elastic material so no residual deformations within structure members but as a whole structure displaced.
i would be grateful to share your experience.

 

[deepquest06]

Hi asad93,

I think the software gives us the absolute displacements. We need to subtract the ground displacements at each time instant to get the relative displacements. But I am still not confident of the analysis. You are right if its an elastic analysis there should be no residual deformation.

Thanks for sharing your experience. looking forward to your reply.

deep

 

[EdR]

Just a couple of comments that may shed some light on the problems.....First deep said he applied motion history to "ground nodes" and then said "ground nodes" were fixed...I'm not sure how Dyna or Ansys handle this but usually you do not fix nodes that have an input motion...Check the input definitions for the code you are using.....

Second check the displacement that results from the defined acceleration time history (integrate the time history twice) I believe the large displacements you are seeing result from the base motion displacement....which is probably what it should be since earthquakes usually don't return to where they started from .........If you want to remove the rigid body displacements you should filter and detrend the input motion...

Ed.R.