Do you think the waveform circular plot represents the load zone of this dryer can bearing with a cracked inner race?

Here is the PeakVue Spectrum

It could be. I would need to know the bearing number, rpm of the dryer and time waveform data in g's no filter to give an accurate analysis......

Mike, Is 10.32 orders bpfi? What is the spacing of the sidebands? Is the spacing in the circular plot at bpfi?

This looks more like inner race wear maybe with a bent shaft journal.

Good Luck,

Danny

quote:

Do you think the waveform circular plot represents the load zone of this dryer can bearing with a cracked inner race?

Hi Mike,

Since the crack doesn't really show much until it is in the load zone, then what you are seeing in the circled waveform plot, more than likely is the "represented" load zone.

It seems to be showing at around 270 degrees, so my question is, is the transducer approx. 270 degrees from the bottom of the dryer bearing, measured from the transducer to the bottom, in the direction of rotation? Or is it about 90 or 100 degrees from the bottom of the bearing, with the bottom being the load zone? The distance in relation to the rotation might tell us what actually is being seen, huh?

The amplitudes are extremely high for a peakvue crack aren't they, but then the speed is high also. Does it show in the velocity data?

Danny,

My bet's with Mike on this being a cracked inner race.

This message has been edited. Last edited by: Ralph Stewart,

I think Mike and Ralph are probably right about this being a cracked inner race. As for determining if the location of the load zone is 270 degrees from the sensor, only if there is a pulse tach input to the data collector from this particular roll, and there is a way for the instrument to relate the location of the sensor to the incoming tach pulse. Otherwise, there is no way to relate the location of the load zone to the location of the transducer. The order lines that are on the display of the time waveform (and the zero degrees location of the circular plot) simply begin at the begining of the time block of incoming data unless there is some tach pulse reference from the roll itself. The correct spacing or difference between the order lines on the waveform and the degree markings on the circular plot depend on how close to true speed the analyst input an RPM. This can also be derived from the spectrum. Judging from the appearance of the data it looks like Mike got the RPM about right!

Skip Hartman

More pics to follow, I expect it to be cracked and spalded. and BAD !!!

Faster fpm, less shutdowns, more chances taken, all allow this bearing to become severe. Expected to be removed March 8th unless fails first.

My thoughts were, if the crack had widened by the bearing creeping up the tapered journal, taking clearance out of the bearing, I might then see spiking in the circular plot 360*, so I am hopeful from the circular plot clearance hasnt been removed from the bearing, thus the reason for the initial question

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

getting worse !!!

Ralph,

Glad you found your way over here.

Looking at the circular plot (assuming that the impacting is occuring at 10.32 orders) does each impact represent the crack in the inner race passing over a rolling element? Would you expect to see much spalling or just a cracked race?

Can you explain represented load zone?

Thanks,

Danny

quote:

Can you explain represented load zone?

Danny,

Not really. At least not right now.
I was just wondering what the relation to the load zone is to where Mike's transducer was sitting and if the circle plot moved if he moved the transducer.
But Skip probably has the right solution or explanation as to load zone vs transducer position.
I have never tried "this". I was wondering if Mike could move his and see if the position followed. Or even if the latest data collected has moved since the posted circle.

Mike,
I was wondering about a couple of things. I see there is a significant increase in amplitude of you BPFI defects from Feb 03 to Mar 01 on the last two spectrums you posted. But there is also an increase in speed. Is the increase in amplitude a result of the increase in speed or is it further degradation of the bearing? Also, your speed is listed as FPM, is that feet per minute and if so, what is the RPM? Not drawing any conclusions here, just hoping to fully understand and learn.

Ronnie

The only time I have seen a cracked inner race there was nothing one month then the next month it was there. In PeakVue it had BPFI and harmonice with sidebands of running speed and running speed and harmonics. I started taking data every couple of days until it could be shut down for repair. As the crack edges started getting rounded off due to the rollers impacting it, the amplitude started dropping. The velocity started showing looseness with some slight 1/2 order peaks. When the bearing was pulled the tapered adapter had just started turning on the shaft.
As far as the circular plot, withiut a trigger for phase you could take the data 10 times with the transducer in the same place and the impacts would be in a different location. It is determined by how far into the revolution the impact occurs after the button is pushed to collect the data. It is just taking the regular waveform with normalizing frequency marked and wraps it in the circle.

D42 Bearing after removal on shutdown

Cracked and Spladed on inner set of rollers, appx 360*
-----------------------------------------------
What was odd on this shutdown, I turned in another dryer bearing to inspect, 4 good inner race harmonic peaks standing out nice and clear in the routine route specturm (unfiltered vel data). not many sidebands present but still, turned in to check. Peakvue, I might add, was not picking out a inner race harmonic pattern, also might add the bearing was only 4 months into run time.. found inner race was cracked as well.. no spalding, just cracked inner race.. was suprised to find that peakvue didnt pick this one out, tried different filters at different speeds as well..
What I learn from this is not every bearing defect reacts the same, whether is depends on fit, load, etc.. I dont know.. I just know that normally peakvue picks out cracked inner races like a sore thumb and on 38 this time it didnt