Shaker Tables - frequency precision for the inverse DFT

The shaker itself is just an actuator, driven by an amplifier. The input waveform is typically calculated in advance, so can easily be an hour long. If you are trying to create a time history to match a PSD profile then it is often an iterative process to find a 'good' signal - simply IFFTing the PSD does not guarantee you'll create a practical waveform, in particular the crest factor needs to be limited.

BT>=1 is what you are talking about, and it isn't 'inefficient' it is very fundamental maths.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

Thanks Greg for you quick response.

Can you tell me any rough estimate for the frequency precision (aka frequency bin size), df, that shaker tables normally use in order to discretize the input PSD when they are trying to convert an input PSD to a time-history (is it in the order of df = 1 Hz, 0.1 Hz, 0.01 Hz, 0.001 Hz, etc.)?

Even a rough number or range would be helpful. Thanks!

Cheers,
B

Hmm, not really, I don't run durability tests. I do generate the time histories collected from the car on the track, but I know they are reprocessed before being used to do a rig test, otherwise it would only be possible to test in real time.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376