Bearing Failure

Almost certainly tensile stress. Could be related to improper fit on shaft, sever out of balance service or large abrupt change in load.

Was the fan blade balanced?

It's hard to tell from the picture but it looks like your lubrication is inadequate.. If you could use some fine grit sand cloth and sand a spot and give an extreme closeup comparing the cleaned and the existing surfaces side by side that might help....This is extremely cold service, is there any moisture present? ...there could be some ice formation with embedded grit...

http://www.timken.com/en-us/products/maintdiag/Documents/5892-Timken-Bearing-Damage-Analysis-with-Lubrication-Reference-Guide.pdf

I would also check the load rating for that bearing and make sure you are operating within the spec....

http://www.skf.com/group/products/bearings-units-housings/roller-bearings/spherical-roller-bearings/cylindrical-and-tapered-bore/index.html?designation=21312%20EK

I would check the bearing speed, to make sure you have the proper bearing type....

http://www.timken.com/en-us/products/maintdiag/Documents/5892-Timken-Bearing-Damage-Analysis-with-Lubrication-Reference-Guide.pdf

This looks like stray current induced damage to me, but I not a whirled class bearing guru.

May one ask if the fan has been left at rest for a period? Do the lines correspond wth the pitch of the rolling elements?

Hello,

Bearing failure is sometimes caused by misalignment between the driven load and the motor.

I suppose that misalignment is unlikely if it's a direct-mounted fan?

I would agree that this looks like it was stored for a period of time in a location which has vibration and the shaft or bearing was not periodically rotated to prevent this type of failure. Do the defects go around the full 360 degrees of the race or are they only on one side? If they are on one side or are much more prevalent on one side than the other, this would indicate the load side of the bearing when it was in storage (assuming it was stored as part of the fan assembly).

I had a similar failure on a computer at my home. It sat off for several years and we live near the train tracks, so the fan bearings got defects where the bearing elements were positioned on the race. They all failed within a few months of turning it back on.

The parallel lines in the direction of rotation along the inner race of the bearing do look like ESD, but it's really hard to differentiate from normal wear without a closer look. If it was ESD, I would have expected similar defects on the outter race though.

FYI, this type of damage is called false brinelling.

@ Phoenix : the fan blade was dynamic balanced.

@ SolarEagle : It is possible that there is moisture in the surrounding. The fan is used to cool vegetables. In the cooling tunnel, 4 motors are working side by side to blow the cold air over the vegetables.

@ Geoffrey : according my opinion the pitch of the lines is equal to the pitch of the rolling elements.

@ Nuccore : The fan is directly mounted on the motor shaft, no belt or intermediate shaft is used.

We are thinking that the customer has adopted his vegetable - machine, and that there is something that vibrates and causes the false brinell effect. (Vibrating trough conveyor ? ) More investigation is going on.

In the past the vegetables - machines did work without problems.

I have my doubts about the ESD, because one frequency drive is feeding 4 motors simultaneously and the chopper frequency is normally low ( 3-4 kHz) on drives in the 90 kW power range.

The 21312 bearing is a self adjusting, double row spherical-roller-bearing

Is it realistic that the cold environment (cold conduction through the shaft) cools down the motor shaft and the inner bearing ring that much, that the rolling elements are gliding and not rolling in the inner raceway?

The outer raceway is kept warm through efficiency losses in the motor.

If the bearing elements were gliding, you would get a different pattern on the raceways.

There are two likely causes to the false brinelling:

1. The bearing/fan assembly was stored/transported improperly and it was already that way when installed. Idle equipment frequently fail in this way when restarted.

2. The bearing was installed improperly (such as with a hammer), but based on the pattern of the brinelling, I would not suspect this cause. You might see galling or one row of elements worse than the other if the bearing was hammered into position.