earthquake motion time step (dt) & analyze time step (dt

[go2shin]

Hi, Silvia

Thanks for the quick answering last time.
I have another one.

My question is about the time step (dt) in my earthquake simulation analysis. Let's say I define one motion (2000 earthquake data, dt = 0.02 sec) in a load pattern. Then I want to increment earthquake shaking with small time step. So in my "analyze" command, if I use "analyze 4000 0.01", is it o.k? Does OpenSees interpolate the applied earthquake motion? Some of my simulation do not have the same result...

Thanks.

HyungSuk

[silvia]

yes

the first time step you give is associated with the ground motion, typically given in the header of the ground motion, you should not change this value.

the step you give in the analyze is yours to choose. opensees interpolates in between ground-motion time steps. you definitely want this time step to be smaller than that of the ground motion, as you don't want to miss a peak. the value of half that you chose is great.

if you use time steps that are too large, you are likely to get a different answer.
A small time step helps with convergence and improves an answer, but can make the analysis run for too long.

so, whether you use 0.005 or 0.01, you should get similar answers, i wouldn't go bigger or small than those values.