Turning Outer Bearing Ring - Yes or No?

[smckennz]

Hi there

a process plant where I do some work has a number of large ball mills; fitted with fairly large rolling element bearings. The loads are predominantly radial, speeds are low, the inner races rotate and the outer races are stationary.
From time to time an outer race defect would occur in the loaded zone. When vibration limit thresholds were approached, we would rotate the outer race 180 degrees to present an unworn surface to the loaded zone. This seemed to work OK.
However now we are told it is better not to rotate the outer race, especially when the bearings are oil lubricated.
I can see pros and cons for both practices, but am unclear when to rotate and when to leave as is.

My feeling is to rotate the race if the rollers are showing no signs of uneven wear, and leave as is if roller wear looks like it would result in reduced contact area if the race was rotated.

Can anyone provide me with a bit more information on this subject?

Thanks

Steve

 

[Hippo41]

Sounds like a consultant's answer to me.

I've often rotated large sphericals for exactly the same reason you say with no problems whatsoever.

Speaking as a "bearing man" I am sure that there are some "sales engineers" out there who would advise you to change bearings simply to get the sale - and some because they don't know the answer.

Assuming that at next 'shut down' you woudl prbably replace the bearings and further that you have not had premature failures of the bearings in service 2nd time round, I would advise you to carry on.

Some simple condition monitoring techniques could be applied to 'predict' failure.

It all comes down to economics.

Lester Milton
Telford, Shropshire, UK

 

[Tmoose]

Can the person who said NO offer an explanation as to "WHY?"

 

[smckennz]

Hi there

It is a reputable vibration monitoring firm saying NO.
It was another reputable vibration monitoring firm saying YES. There is no apparent incentive for the NO guys; they do not say replace; just continue monitoring.
If I understand correctly NO guys claim the outer race and rollers will have worn as a matched set - like running in.
So the race to individual roller contact "line" will be nice and long - apart from a gap where the outer race pit is.
By turning the outer race, this match is lost and the rollers will be subjected to higher local stresses.

Sort of like replacing journal bearing shells without linishing the journal.

Cheers

Steve

 

[Tmoose]

Have you seen if the spalls typically appear mid roller, or at the edges?

I appreciate the Naysayers' concerns, and understand their argument, but think it may be based on kind of simplistic assumptions. The stress profile of a purely cylindrical roller in heavy contact with a race shows very high edge stresses even with perfect alignment. IT's worse with any misalignment. I think the typical solution is to make Cylindrical roller bearings tapered a few microns to reduce the brutal stress concentration, for a large overall boost in load capacity. Running the rollers over a hole brings on similar stress concentrations, and might be a big part of why a spall continues to grow.

If the rollers and races "wear" in significantly, as opposed to becoming slightly polished, it implies to me the lube viscosity is too low or there are contamination issues.

If this has happened frequently you may already have data to see about how long it takes for the rotated bearing to hit the vibration limit threshold again. And could compare that to what happens to bearings that are left alone.

 

[diamondjim]

Depending on your wear pattern, you
might even consider turning them 120
degrees and get three times the life.
I am assuming most of the wear will
take place on the inner race.