Fella's
I am seeing what appears to be run speed and harmonics in the peakvue spectrum. Best I can recall, and I have only seen this one other time, this is an indication of severe looseness. The last time I saw this it was severe looseness and the looseness was present in the non enveloped spectrum. This time there are no run speed multiples in the non enveloped spectrum, only in the enveloped spectrum.
Then looking at it more it looks like a second set of peaks with side bands of run speed. I was under the impression that side bands would not be present in the PeakVue spectrum. Is this true.
Looking deeper the second set of peaks is actually lining up with BPFI on the SKF 6319 Bearing.
Is this a true bearing fault or is this run speed with several multiples or is this some thing else.
The first two spectrums are peakVue and the ones that follow are non enveloped spectrums.
If this is run speed and I I have a severe looseness issue why don't I see it in the non enveloped spectrums.
Nothing unusual in the Wave form.
What am I missing here?

1E_Pulv.doc (146 Kb, 66 downloads)

Could be the inner race on that 6319. The amplitudes are pretty low in all the data. Are you able to listen to the bearing with a stick or screwdriver? What does it sound like?

Looks like Peakvue is picking up an inner race defect out past 1000 Hz. The fundamental of the inner race for a 6319 bearing is 4.904X, and looks like you are seeing 1X sidebands around the inner race frequency rather than looseness of 1X.
IMO looseness is not here, only an early inner race defect.

quote:

I was under the impression that side bands would not be present in the PeakVue spectrum. Is this true.

Sidebands do show in a Peakvue spectrum. Especially around an inner race defect.

Take a normal spectrum out to 2000 Hz and see if you see the inner race defect develop[ing.

The 1X comes from amplitude modulation of the impacts from the inner race (you will see a similar patern for a BSF fault being modulated at cage). If you look at the autocorrelation function, the 1X will be absent (it would be present in the case of advanced looseness).

I'm with Ralph - The peakvue data sure looks like an inner race fault.
Even without autocorrelating the waveform, you should find that the two dominant peaks occurring each revolution in the waveform will match up with the BPFI. If you autocorrelate the waveform, you will see peaks at BPFI, modulating each rev.

I would be interested to see the FVE waveform autocorrelated.

Ian

Thanks Ya'll,
I just wanted to make sure that I wasn't missing anything. There is still no faults present in the nonenvelope spectrum, so I just wanted to make sure that I wasn't missing anything. It is nice to have that warm fuzzy feeling again. I will post again when there is some resolution, as in the repair or failure of this bearing.