interpolation problem

[pp_0928]

Hi,all.

I have transient analysis that is running at a time step of 0.015

My input acceleration file has a time step of 0.01

How does OpenSees interpolate this acceleration data set??

I'm told that it's linear interpolation. However, I still want to prove it. I try to record acceleration of the point in the ground, all the data are 0. Is it a relative acceleration or resultant one? How can I record the ground motion acceleration which is used in analysis?

[silvia]

yes, it interpolates.

in the uniformExcitation pattern, you input acceleration and output is in relative form. To avoid redundancy, the program makes you keep track of the input acceleration.

in the multipleSupportExcitation pattern, you input diplsacements and output is in absolute form -- you cannot input acceleration in the current version of opensees, though.

[pp_0928]

Thank you for your prompt reply, Silvia.

Actually, I want to figure out how Opensees interpolates.

For example, the ground motion is 1 、1.3 、 1.4 、1.7、 2.0 ，the ground motion dt is 0.01.
If I want the analysis dt is equal to 0.005. Will the opensees interpolate the data of the average value between the point? Is the input acceleration 1 、 1.15 、 1.3 、 1.35 、 1.4 、 1.55、 1.7 、 1.85 、 2.0 ？

it's in the uniformExictation pattern.

[silvia]

Opensees just connects the dots.

[pp_0928]

Thanks again for your help. Silvia.

I still want to know what do you mean of connecting the dots.

The ground motion has 3000 points, the dt is 0.01, we get 3000 points of response. If my analysis dt is 0.005, the opensees will take 6000 points in calculation, we have 6000 points of response.

How does opensees create the other 3000 points?
Does it linearly interpolate? Interpolate the new point with the average between the original front and back points?

[silvia]

linear interpolation between the point before and the one after.

[pp_0928]

I got it. Many thanks, Slvia!!!