Problem of bidirectional earthquake loading

[Rain]

Hi everyone,

I want to simulate 3D centrifuge experiments with different inputs specified for two directions. Examples on the OpenSees website are for the vertical and one of the horizontal directions(1 and 3). However, I want to impose two horizontal directions perpendicular to each other(1 and 2).

I have tried the Multiple-Support and Uniform Excitation load patterns, but seismic responses are much smaller than the unidirectional loading. Are there some examples for this kind of situation?

[fmk]

play with a square elastic column with same motion in 2 orthog direction to ensure you have the commands corect.

as for centrifuge:
http://opensees.berkeley.edu/OpenSees/m ... l/4244.htm

[Rain]

Hi fmk,

Thanks for your reply.

My model is a simplified square elastic column indeed.
With same motion imposed in 2 horizontal directions, the join motion will still be unidirectional. The direction will be between those two directions, and the amplitude is their root-mean-square. However, with different motions imposed in 2 horizontal directions, the seismic response is abnormal. I cannot find an appropriate solution...

[fmk]

what do you mean the results are abnormal when you apply 2 different ground motions? did you test the 2 same motions to ensure the results were as they were supposed to be?

[Rain]

Hi fmk,

The two motions I imposed are actually sinusoidal accelerations with 90 phase difference. The results, such as the settlement and the maximum pore water pressure, are lower than that with single input motion.