Lil History for ya - this 1200rpm pump was rebuilt in Sept 06 and before the end of Sept is was no longer available. The pump inboard bearing (greased & balled) was buzzing like a huge bee. Packing was leaking badly. Someone put a cover over the leaking packing to keep the water from the floor. Must have put water into the bearing... easy 1st guess, a no-brainer... but wait nothing is ever that easy.

Spike energy is at 23-24 orders. The are only 12 (2 rows - see last note in next entry) balls in that bearing & 8 bolts at the flexible coupling.

Motor is out of alignment low .027 outboard & .015 inboard.

Is it the alignment, coupling or bearing? or all three?

Where is that 23-24 order energy coming from?

See the jpg below. The vertical spike is missing & surrounded by side bands.

Maint is working the pump & I will let you know the results.

This vibration stuff is fun!

This message has been edited. Last edited by: CBM Novice,

Novice,

My guess would be that you are looking at the natural frequency of the outer race. I wouldn't be too concerned about direction in the presence of advanced wear. What are your fmax, lor, etc? Do you store timewaveform data?

Failure progression would be misalignment, leading to seal failure, leading to lubricant contamination, leading to rapid bearing degredation.

Welcome to the neighorhood. Where in VA are you?

Good Luck,

Where's the point or data taken from and you're trying to relate spike energy to a frequency spectrum and why do you say there's misalignment vertically?

Why not analyze TWF and FFT's?

Thanks for the rapid response.

What are your fmax, lor, etc? Do you store timewaveform data?
fmax = 100Kcpm
Yep but not sure how time wave forms help you. Do you have a good site/method of TWF analysis?

Where in VA are you?
Yorktown - name is Stan

Where's the point or data taken from and you're trying to relate spike energy to a frequency spectrum and why do you say there's misalignment vertically?
Data was taken from the inboard pump bearing. Alignment was measured across the coupling with some fancy roto/laser machine.

Why not analyze TWF and FFT's?
TWF - Not sure how. My training has been trial & error (more error than the other). It seems easy for me to look at the spectrum figure out the multiples and write them up for maint.

This is why this problem is a neat one. The inboard bearing has all the bad energy. But the bearing only has 12 balls (seen it this morning).

Maint - water was in the bearing (spins easy on the shaft) & they will provide me the alignment info when it's back together. Packing was healthy & gland had plently of room to be tightened up. They are going to replace some worn parts on the coupling. Ain't much more I can do now.

Take a look at this. It is the horizontal reading off the inboard pump bearing. Above was a close up of the same spectrum. What does it tell you?

This message has been edited. Last edited by: CBM Novice,

AOK - got it. There are 2 rows of 12 balls. 23-24 orders would be from the inner race passing by each ball as it is frozen in place. Water displaced the grease, bearing over heated & the balls stopped.

Ignore the rust, look at the groves. There are plenty.

IMO,Waveform needs to be in acceleration units rather than displacement, for better analysis.

"There are 2 rows of 12 balls. 23-24 orders would be from the inner race passing by each ball as it is frozen in place. Water displaced the grease, bearing over heated & the balls stopped."

Sounds possible, but I have never actually seen it in real life.

Only my opinion and I could be totally wrong.