Dear All;

GMF, BPFO,BPFI.... Every thing normally comes with it family of harmonics, My question is why it is appearing in FFT!

I learnt th side bands are the result of amplitude modulation of the impacts during motion inside and out of the load zone. so what is causing harmonics to really occur.

It's look a basic question to me and hence I wish to calrify it immediately

Some pairs of spectra / TWF's

A - Spectrum has only a single frequency <==> TWF is a sinusoid

B - Spectrum has a series of harmonics <==> TWF is periodic (period is inverse of fundamental frequency)

C - Spectrum has Non-harmonically-related peaks <==> TWF is not strictly periodic.

Note the arrows go both ways. If you observe something on one side of the arrow is true, you can conclude the other side of the arrow is true. If you observe something on one side of the arrow is not true, you can conclude the thing on the other side of the arrow is not true.

A is pretty simple. A single peak in the spectrum corresponds to a sinusoid. If what you see in the TWF doesn't look like a sinusoid, you shouldn't expect to see a single peak in the spectrum.

B is the somewhat surprising result that was discovered by Fourier (I think). Any periodic signal can be decomposed into a sum of sinsuoids at frequencies which are harmonics (multiples) of the fundamental frequency.

C is the expected conclusion when the conditions of B are not met.

Amplitude modulation ... I'll let someone else tackle that.

A single sinusoidal wave is rare in practical world vibration signature. It is normally a combination of so many sinusoids which by FFT can be broken into discrete frequency vs. amplitude signals. In certain cases, such as misalignment, mechanical looseness and many others, the motion of the shaft is not sisusoidal since it has not got equal freedom to move in all directions. for example, in case of misalignment, shaft motion is curtailed/clipped by bearings at extremes. These clipped portions of the wave have energy. they are also picked up by the analyser and they appear somewhere else in the waveform which in spectrum represent harmonics.

Hi guys,

Then why there is no harmonics for a resonant frequency?

Have a nice day!!

This message has been edited. Last edited by: Jenish,