I have several overhung fans with spherical roller bearings 'rumbling'. Vibration spectrum is showing dominate 3xTS or 4xTS on the driven end. I removed the bearing caps and gave a few of the bearings a visual. I can insert a .005" feeler gauge on the side of the bearing and not so much on the others. The bearing clearance was measured at .0015" to .002" across the board.

Is this a normal amount of housing clearance? The bearing are 22215 split housing tapered bore spherical bearings. I have added a few attachments.

Any help would be greatly appreciated. I zipped some photos.

This message has been edited. Last edited by: Vibe King,

EXHAUST_FAN.zip (597 Kb, 260 downloads)

From your picture, I see that this bearing has stab rings so it should be located in the fixed (non-floating) position. I can also see a red oxide from corrosion fretting (rubbing under pressure) in a zone near the split line. I suspect that the shape is oval in that area.

ISO/R286-1962 says for a stationary outer ring in a split housing that the appropriate fit class is “H7”. This calls for a fit of between 0.0000” to 0.0023” loose.

A 22215 bearing has a basic outside diameter of 130mm- with tolerance this translates to 5.1181 to 5.1174 inches. So, the acceptable measurements of the housing, without precision measurements of the actual bearing, would be 5.1181 to 5.1197 inches.

Caution: This tolerance assumes a round fit. The fit should never be oval as this will allow the outer ring to distort and concentrate all the load on just one or two rollers.

Brown grease? Double nutted bolts? Worn seals to the point of middle spline missing?

Bearing clearance between the rollers and raceways is ________________. Or, is the avg 2 mils between the housing and the outer race of the bearing.

And the OEM spec clearance is ________________?

The grease used is Castrol Tribol 860/220-2. It is a black grease which contains what they describe as lube solids.

Is double nutting a problem?

Bearing clearance between rollers and raceways is .002" on this particular bearing. Which is about correct according to spec.

How do you add attachments as a hyperlink in this post? I've got some data to share. On some bearings I'm seeing 3xTS as high as .1 in/sec. On other bearings I'm seeing 4xTS as high as .13 in/sec.

The fretting is occuring on the load side of the bearing. I'm guessing the fretted area is about .002" deep. How soon would you change this bearing?

Could imbalance cause this problem over time? If so, how much of an imbalance would it take over 5 years on average? I've only noticed this problem on exhaust fans and not supply fans. Of course, exhaust fans get paint on them from the process and supply fans do not.

This message has been edited. Last edited by: Vibe King,

Is this unit belt driven or coupled. If belt driven, what is the angle of pull from the motor sheave to fan shaft sheave?

Double nutted bolts. Depends on the reason they are double nutted.

These fans are belt driven. The belts are about 30 degree from horizontal. The motor is lower.

These fans ar double nutted from the factory. They are Northern blowers.

The belt drive with that angle could be the answer for the bearing load zone being at the split. Would it make sense that belt tension with misalignment would work the bearing in the housing causing the fretting? Possibly the sheave is eccentric or you have unbalance?

If you think that the bearing will get deformed from the wear, will not secure properly or fail when you most need the fan then the housing and bearing should probably be changed.

The double nut of the pillow blocks from the factory could be just an extra measure to secure the pillow block (locking nut) or they have had some negative experience and this was the correction.

Is the base resonating at a running speed frequency (belts, fan speed or motor speed)? Do you see multiples of any running speeds in the spectrum and is it higher H, V ,or A?

Good Luck (my 2 cents)

Vibe King,

Based on fretting corrosion observed you are likely having excessive looseness problem. But your measurement method may not be accurate. A better way to measure clearance is to utilize "plastigage" you can get in any auto parts store. Use them when bearing is not very hot.

Please post vibration data, just click on "add attachment"

David

I've considered plastigage but decided that the wear is in the load zone. It would be difficult to get an accurate information because the bearing don't look wore except in the load zone of the housing.

How much out-of-roundness can a bearing tolerate before it causes problems? As, it was mentioned earlier about loading less balls.

If you got a new fan and set it you may want to consider a complete bearing alignment as you need bearing to shaft alignment in two planes and bearing to bearing alignment. If you haven't gone through the complete alignment process, then I'd do it now. Don't depend on the factory settings performed somewhere else other than your specific environmennt.

These fans have been in service for about 5 years. As they continue to wear the groove deeper, at what point would you say the bearing housing is going to cause bearing problems? Maybe a better question would be, at what point can you let 4xTS get to before you change the bearing?

I had some Northern Blower fans come in here a few years ago with fixing rings installed on both bearings. Normally this would not be an issue except where heat is involved. We have removed the fixing rings from the non drive end bearings and the fans ran quieter. Fixing rings in both bearings can cause distortion in the shaft and wear the housings. Have you checked the NDE bearing?

Quote from the SKF Bearing Maintenance Handbook-
“The cylindricity tolerance as defined in ISO 1101-1983 should be 1 to 2 IT grades better than the prescribed dimensional tolerance, depending on requirements. For example, if a bearing seat on a shaft is machined to tolerance m6, then the accuracy of form should be IT5 or IT4.”

So, if the housing bore is H7, the accuracy of form should be IT6 or IT5. IT6 for a 130mm is 25micrometers or .0009inches, while IT5 would be 18micrometers or .0007inches.

Vendor Warning- I work for SKF Reliability Systems but will attempt to be unbiased in the following statements:

I have toured an SKF bearing manufacturing plant although I have not worked in one. At one point during the tour, I noticed a sorting machine where rollers were being grouped according to diameter after they had been honed. The rollers would pass through a measurement machine and then were placed in one of several bins.

Bearings were being built with all of its rollers being pulled from only one bin. I asked the tour guide how accurately the rollers were being sorted. He removed a roller that had been sorted from one bin and then he used a marker to write on the roller. He placed it back on the line before the measuring machine. The roller was sorted to the next bin. That’s how serious SKF is about dimensional accuracy, and I think that other leading manufacturers have similar practices. We wouldn’t do that if it did not have a significant benefit.

I think that you are asking the proverbial “How long will it last?” question. But it is really dependent on load conditions. I have seen outer races crack with little warning due to housing problems like yours. My two cents- Replace the housing and bearing as soon as possible

We are having a 3x times problem showing in a Northern fan that was recently installed in August 2007. This fan is very critical and cant be out of service very long without causing a lot of problems. We were able to take it down long enough to do a lift check on the shaft. Had appox. .006 vertical movement. Hopefully it will survive till the next outage to find the looseness(fan wheel, bearings or sheave)
vibbase, how fast is your fan turning?

This message has been edited. Last edited by: Scott Harper,

High_3xTS_in_Spectrum.pps (2,412 Kb, 22 downloads)

quote:

Originally posted by DavidM:
Quote from the SKF Bearing Maintenance Handbook-
“The cylindricity tolerance as defined in ISO 1101-1983 should be 1 to 2 IT grades better than the prescribed dimensional tolerance, depending on requirements. For example, if a bearing seat on a shaft is machined to tolerance m6, then the accuracy of form should be IT5 or IT4.”

So, if the housing bore is H7, the accuracy of form should be IT6 or IT5. IT6 for a 130mm is 25micrometers or .0009inches, while IT5 would be 18micrometers or .0007inches.

Thanks David - that is useful info for establishing limits for cylindricity. I looked around for awhile to locate those IT grades, then I realized they are available at a handy location in that SKF Bearing Installation and Maintenance Handbook table 7 on page 56.

Interesting data Scott. I will try to post what I'm seeing.

I believe I will start with the worst one, change it, and see how it goes.

The zip file contains data on two different fans. I'm seeing similar wear in both fans.

I'm not asking 'How long will it last'. Instead, how bad is it really?

The thing that bothers me is this. I have recommended repairs in the past. The customer decides not to repair. The stinking thing runs for a long time. I'm talking about obvious things like bearing defects and so forth.

I seem to stay on the edge all the time because of the closeness to failure I have to call pieces.

This message has been edited. Last edited by: Vibe King,

FAN_DATA.zip (44 Kb, 20 downloads)

Mr King,

Besides the obvious loose housing fit, there may be additional issues that cause the sound and vibrations:

1) Loose fit of inner race/ring on shaft
2) Excessive internal roller-to-race clearance
3) Bearing fault (rollers, races, or cage)
4) Lubrication (type, quantity, and interval)
5) V-belt tension and sheave alignment

Failure to identify all fault issues will lead to a partial fix that does not last very long.

Walt

I don't want to stray. But, here are some more data on another piece that worries me. IB has had the spike energy for a while. It came suddenly and hasn't gotten any worse.

The 4xTS housing loosesness on the OB has just come up. The RPM is right. Is the autocorrelation polar plot indicating some migration of the outer race in the housing?

Thanks, Walt, for your input.
Roller to race clearance is .002". I don't see anything in peakvue indicating any kind of bearing defects. We are using Castrol Molub-Alloy 860/220-2. Its a black grease with lube solids in it. Alignment and tension look good.

This message has been edited. Last edited by: Vibe King,

22322_Spherical_roller_bearing.zip (109 Kb, 27 downloads)

Here is peakvue data of the fan. I am seeing 1xTS in peakvue. But, its really not that excessively high.

The funny thing is that the 1xTS is on the OB. That bearing doesn't show any wear of the housing.

This message has been edited. Last edited by: Vibe King,

FAN_DATA_(peakvue).zip (58 Kb, 22 downloads)

Mr. King,
I suggest you change the grease you are using to a #2 Heavy Duty Lithium Complex
Grease with tackifiers

One reference brand name would be CONOCO Multiplex Red #2.