Hi;

Equipment : Circulating Water Pump
Bearing : SKF 22217 C / may be 22217 E
RPM : 600

Nature of problem : Predominant Peak at 17100 CPM with side bands around 3375 CPM. ( Refer attachment )

Please discuss in this regard.

Please discuss in this regard

In which direction is your measurement?

If the X-axis grid spacing is 5000 cpm and the big peak and cursor are positioned above(to the right of) the 20,000 cpm grid line, then how can the cursor label be 17,100 cpm? What is the spectrum frequency resolution or number of lines?

Walt

As Walt said, it looks like the cursor is above the 20,000 cpm line and the amplitude would be about 0.185 IPS. The amplitude and frequency in the the display is 0.039 IPS at 17,100 cpm and it looks like it shows Pos -1, which is the sideband left of the cursor. Not sure how that works, but confuses me. I don't see any of the frquencies related to any of the information Prabhu listed.

Regards,
John J

I agree with John.
Did a quick lookup at SKF bearing freq and nothing seems to line up.
The peak and sidebands appear to be non-synchronous.
Are you sure you have all the bearing information?

Jim P

hai friends

Prabhu is my colligue,i had seen the comments, thanks for ur comments.
the first spectrum posted by prabhu is 4 horizontal direction & it is a Jackskaft DE.
find the attachemnt , here the strange freq is appearing on 17100 CPM.it is in 4vertical Direction.

the no of lines: 400.

the bearing info for 22217C or may be 22217 E

as per Skf Info:
ball dia:16
no of ball:19
PCD:118.2
Contact Angle:8.5

any more info, we are ready to give for this Spectrum analysis.

pls discuss with frequency

senthilkumar,

Am I seeing a black image on your last post or is something wrong with my computer?

It is not possible to do bearing analysis using a 400-line/1300 Hz spectrum. The resolution is too coarse. Many bearing defect frequencies will be very close to 1x harmonics. For instance, BSF=3.481x for a 22217E. What you would more likely see is 2xBSF which is 6.962x rpm. A 400 line spectrum will not resolve the difference between 6.962x and 7x. I would use at least 3200 lines for this analysis, and more lines if available.

You can't do accurate diagnostics without adequate data.

We have a large peak at 17100 CPM identified by the cursor as 28.6X TS with 3375 CPM sidebands.
17100/28.6= 597.9 CPM TS
3375/598= 5.64X TS
17100/3375= 5.06
Granted 5.64X does not matched the bearing faults we think should be present and the resolution could be better but what else can it be? If it looks like a duck and sounds like a duck...

It looks like a duck and sounds like a duck, but this is like hunting in a desert with a bow and arrow.

Aubrey

So if we have a dominant peak we can't identify, we just call it a bad bearing?????

Prabhu,

The vibration spectrum posted by you indicate one or more of the following:

1. Looseness
2. Wear ring rub
3. Inrusion of foreign article into the volute hitting the impeller vane.

We have multiple peaks at ~5.65X TS. Call it something, not insufficient resolution to analyze a bearing fault. What else fits the pattern?

Is it possible the 22217 bearing number is incorrect? A 22317 (same bore and OD) has a frequency of ~ 5.6X or possibly some other number other than 22217.
It is obvious there is a pattern of harmonics non-schyronous with 1x, but are WAY OFF from the frequency of a 22217.

Why a jackshaft? Is there a gearbox also? Split bearings?

Here are the diminisions associated with some 22217 and 22317 bearing housing. This might help deterinmine which you have.

http://www.skf.com/skf/product...13&prodid=5106130217

http://www.skf.com/skf/product...11&prodid=5106110317

quote:

Call it something, not insufficient resolution to analyze a bearing fault

"Call it something".... this is not politics. There is simply insufficient resolution to do ANY kind of meaningful analysis.

I was wrong. I thought it to be multiples of running speed.

I apologize!