natural frequency

[sadraddin]

Hi there
Does anyone know how to calculate natural frequency and mode shape from acceleration time history?

Thanks

[huffte]

I am not sure I completely understand your question sadraddin. Natural frequencies and mode shapes are characteristics of a particular structure and depend upon the mass and stiffness distribution of that structure. An acceleration time history is peculiar to a certain recording of ground motion at a particular station during a particular earthquake. So, you see, the two issues are completely independent. I might recommend a beginning text on structural dynamics, perhaps "Elements of Structural Dynamics" by Berg or something similar. Press on and best of luck sadraddin.

[sadraddin]

Thanks huffte I really appreciate your help. I can calculate it by mass and stiffness but this is theoretical method. In experimental method we have to use Fourier Transformation but I don't know how to calculate it from that.

[ruipinho]

sadraddin, hi.

You may indeed use SeismoSignal to estimate the dominating response periods/frequencies of a structure subjected to seismic action (by computing the Fourier spectrum of its acceleration response time-history (typically, that of the top storey)).

However, in order to obtain a more accurate characterisation of the modal periods/frequencies and, especially, of the modal shapes of the structure, you should make use of more specialised software or algorithms, typically employed for system identification in lab or in-situ tests (you can find many options by googling "system identification software" or similar).

Best,

Rui