I am hoping to get some help with a problem on an over-hung, belt driven fan. The size is 4-75, 15650 CFM. I do not know the blade count or type. The fan bearings are Dodge pillow block 520, 3 7/16". Motor speed is 1192 rpm, and fan speed is 1613 rpm. The fan is on the 3rd floor of a steel platform with concrete base under fan/motor.
Velocity data and peakvue data show impacting at fan speed on fan end bearing. Peakvue data shows a BPFO at drive end bearing, but amplitudes are low. The highest amplitude in peakvue is 3x fan speed. A "clicking" sound can be heard between fan end bearing and blades with, or without headphones. The fan bearings were inspected a couple of weeks ago, nothing was observed except old looking grease. The bearings were cleaned and repacked, in place. The sheaves were realigned and the belts replaced. Initial readings after work was done, decreased, but have steadily increased. The primary frequencies seen on the motor are fan speed and 2x fan speed. Any help or suggestions would be greatly apprieciated!!!!!

HVLC_FAN-_211.doc (220 KB, 106 downloads)

Brickbat,

What version of RBMware are you running? Maybe you could email some of us the data where we can have a closer look and see what we can see. If you have version 5.2, I can't help, but some others might can who have 5.2 or higher.

My first impression from the post would be outboard bearing a little loose, possibly on the shaft, but a little hard to tell with a couple of data sets.

Email ralph_stewart@alertanalytical.com

I do have version 5.2

FOV peakvue time waveform amplitude is very high. Autocorrelate this and see what the major peaks are. Looks like looseness and/or bearing fault. If you see inner / outer race peaks change the bearing. Don't expect the millwrights to find anything on an inspection. It's very difficult to see a scored race until the bearing is pulled-out and you spin the balls / rollers 90 degrees out of the way. What were the clearances in the fan bearings when inspected?

I agree with Ralph. Check the bearing to shaft fits. There may be some looseness there. Also, check the shaft with dial indicators looking for runout or eccentricity.

Gary B
General All Around Flunky

I would agree that it is likely a Mechanical Looseness. You say that they found nothing of concern during their bearing inspection, but do you know how well they inspected the housings? Looseness within the housing is the most common issue that I see on this type of machine. Typically though this is pretty easy to spot when you run finger across the bearing fit of the top pillow block.

Even if the shaft nut lock nut is tight I'd try to observe if a feature on the inner race and shaft stay aligned with a strobe synched to 1X fan speed.
Unfortunately, if I understood the 500 series bearing correctly, everything is hidden within the pillow block.

Then I'd scribe a mark on the shaft or use Dykem or paint to put a mark across the race/shaft joint, and inspect it after a few hours.

Sorry it has been so long replying, I've been off for a few days.
PCVIBE, the major peaks on the correlated peakvue wave form are at fan turning speed. The highest peak is at 3X fan speed in the spectrum.

Michel, One of our better mechanics did the inspection and repack. He called me to the job when the caps were off. I/we did not see any signs of of bearing creeping or spinning in the housing. If I recall ( mechanic is off today) there was .002" clearance in the bearing. Not sure if he was able to measure/see any clearance on shaft to inner race fit.

Dan, due to guarding, can not scribe shaft.

I collected data again today, and it is still climbing. I was asked if it would last till June. I told them I would check my magic 8 ball and get back to them. I will be watching it closely for now. I'll post anything we find. Thanks for the input!!!

Has this waveform changed in your lastest data?

This could be a rub in the fan.... had one a few years back that looked like classic looseness.... growing harmonics, 1/2 harmonics... turned out to be the fan hub rubbing on the housing (wheel had moved on the shaft). You'd think it would be audible, but there was no excessive noise.

Ralph, the waveform and spectrum from the outboard vertical fan bearing has increased. There is a little over 11Gs in the waveform. I am not seeing fractional harmonics in the data. I had wanted mechanics to check the rotor on thier last inspection, but that would have required a more extensive lock-out and use of S.C.B.A.s. Limited time frame did not allow for it. They plan on replacing the bearings in June, and hopefully they will check the rotor to housing clearances and rotor to shaft fit.
We also think the wrong hubs may have been used when the sheaves were changed out in 9/2008. The hub is pulled up all the way on the fan sheave, and there is about 3/8" gap on the motor end sheave. There is only 1/16" difference in bore diameter. However, I do not think that is the main problem.

HVLC_Fan_Too_doc.doc (46 KB, 29 downloads)

I also see rubs with harmonics that do NOT have fractional harmonics. You might take some acceleration data on the fan housing. I would expect to see broadband noise in the spectrum, probably from 1kHz - 3kHz, with a significantly raised noise floor.

Rusty,
It's been kind of hectic around here the end of last week. I finally got a chance to post the acceleration data I collected. Not sure if this is indicating a rub or not. I took these readings at various different places on the housing. Some seem to show the bundles of energy between 1KHz and 3KHz, and some don't. What do you think???

HVLC_Fan_acceleration_data.doc (229 KB, 34 downloads)

You did not specify whether or not this is an overhung fan. If that is the case we have several around our plant that will act this same way (many harmonics that look like looseness with impacting in the waveform). Usually we see the highest vibration with the most harmonics on the outboard bearing vertically. The periodic publication from the vibe institute had an article on this as well.
We find that sand blasting the fan reduces this vibration.

There may be rub or looseness but the amplitude is not so high worth the headache. Max. amplitude is around 4 mm/sec. In 1-3 KHZ range some energy bundle is there but in acceleration mode, it should not be considered high.Also, 1X hits may indicate BPFI modulation when the bearing moves in and out of the load zone.
Regards