Bearing Arrangement 3

[feaeng1]

Dear group, here is my question. Is that common that four bearings with different types are packed together?

It’s for a grinder spindle with a 200HP motor. The first bearing up front is 6222, with one N222 behind that. Then two 7222 are followed up. On the other side of spindle, there is one pair of 6220.

Thank you.

 

[diskullman]

common? no, but may be needed for this application. I'd guess the grinder end has the big bearing cluster. I'd love to see a picture of that.

Russell Giuliano

 

[Tmoose]

Belt driven spindle, with the N222 at the belt drive end?
Or motorized spindle ?

Are you rebuilding it? How long did it last since last rebuild?

I think you may have found somebody's work-a-round. That's a lot of -22 bearings to expect to play nice together

 

[feaeng1]

Thank you for your responses, Russell and Tmoose.
It’s a motorized spindle. The motor is connected with a coupler at the end having one pair of 6220. This grinder really has some problems. I’m doing the failure analysis on it. It’s failed about every 6-month. The bearing contamination is its first problem due to the seal failure at the end having four bearings. Most people here think that the large shaft deflection causes the seal opening up. The broken shaft is its second problem. The shaft was broken between the first bearing and second bearing in that four-bearing group. I think it’s caused by the poor bearing arrangement, which generated artificial high load on the shaft. What do you think?

 

[Tmoose]

The condition of the shaft near the break may be eagerly trying to tell you what went wrong. Multiple failures often drive all manner of compromised repairs like shaft welding, re-stubbing, and epoxy restored journals, as can simply lack of understanding of proper bearing arrangements.
That assortment of bearings on one end has a certain Dr Seuss quality to it.

http://www.danheller.com/images/UnitedStates/Florida/Orlando/Universal/DrSeuss/seuss-signs-big.jpg

http://www.moonbattery.com/archives/20811cs.jpg

http://he.wikipedia.org/wiki/תמונה:Dr_seuss_house.jpg

I'm still not clear about the spindle design. What does the coupling drive? "Motorized spindle" usually means the spindle shaft has the motor rotor shrunk on to it, with appropriate bearings at each end.

http://www.mmsonline.com/hsm/hsmevent/mag_images/diemold_spin2_a.jpg

I've seen a ball bearing (small radial clearance) sometimes beside a roller bearing (larger radial clearance) in an electric motor so during light side loads the roller is disengaged to avoid roller skidding. I'm not sure how successful that really is. Which bearings are intended to be the thrust bearings?

 

[xnicke]

A sketch would be really nice.

 

[feaeng1]

Sorry for the confusing. Yes, one picture is better than one-thousand words. I do have a sketch but I have trouble to post here. I don't have any website to link. Maybe I can email to someone here if you can provide your email address.

The problem area is that four-bearing group. Originally, there were only two bearings there. Later people here thought that large shaft deflection opened up the seal causing the bearing contamination. The seal is the labyrinth seal. Then they tried to pile up all types of different bearings to increase the system stiffness. It leads to broken-shaft, which I believe.

For the contamination issue, I think that labyrinth seal along just doesn’t work no matter if the shaft deflection is large or small. The grinder operates in a very dirty environment. The debris and cooling water are directly pouring on the seal area. It needs a different seal or a shield on the top.

For the bearing arrangement, I think it’s not really necessary to use four bearings and better to use the same type of bearing.
Those are just my opinions.

 

[JJPellin]

I was always taught that you can never have more that two bearings on one shaft without risk of shaft fracture. For the sake of that rule, two back-to-back thrust bearings would count as a single thrust bearing assembly. I agree with your assessment. I would go back to the original two bearings and convert to a better shaft seal. The options that come to mind for me are:
* Lip seal - very bad idea in a dirty environement. It will score the shaft and fail.
* Magnetic Face Seal - Only if you pick a design where the magnets are isolated from the outside environment. Otherwise they will just pick up debris and plug up.
* Floating Segmented Carbon Ring - Either a single ring, or even better a double ring with a gas purge in between. We use this a lot on blowers in dirty service.
* Purged Laby Seal - You could stick with the laby seal but add a gas purge into one of the outer grooves of the laby. This has worked well for us in gearbox laby seals with problems with contaimination getting in or oil leaking out.
* Grease Lubricated Packing - This might seem like a big step back in technology. But in some cases, the simplest answer is the best. If it not a problem to have very frequent re-greasing, this could work well.

 

[feaeng1]

Yes, JJPellin, you are absolutely right. It’s my first time seeing that four bearings are put together, furthermore with different types.

Thank you for all you suggestions on seals. I’m seriously considering your third and fourth recommendations.

 

[Tmoose]

There are good labyrinths and bad labyrinths. Sometimes the difference is pretty subtle. A good labyrinth and air purge can be quite rugged and effective and take up very little space and handle normal shaft deflections easily.

Is this spindle horizontal or vertical?

From where I sit the spindle design needs review, including thoughtful analysis why the shaft broke and the working and induced loads the spindle is subjected to. I'd generally expect a few of the clustered bearings to have a very short life rather than a busted 4.3 inch diameter shaft shaft.

 

[feaeng1]

Tmoose, the spindle is horizontal.

For the normal working load, the radial reaction force at four-bearing assembly is not over 2000lb with 200HP and 2144 working rpm and without considering impact load. The axial load can be ignored. As regarding the induced load, it’s a tough part. Probably it’s caused by the different bearing clearances. Is there any special method to figure this out? or just using beam theory.

 

[Mendit]

Beware of off the shelf product when mounting bearings in tandem or multiples. You will need to use specail bearings with the high point of eccentic marked on them. Thus will allow you to assemble multiple bearing on a common shaft.

Precautions.
High speed big problem with heat
Grease lube must be into the center of the bearing group
Grease frequencies must be doubled with 2 bearing and doubled again with 4 bearings.so could be problem.

Check this site for usefull info
.

http://www.mrcbearingservices.com/product_detail.asp?product_id=21

 

[Tmoose]

"For the normal working load, the radial reaction force at four-bearing assembly is not over 2000lb ".
- How far off the spindle nose and first bearing is the load applied?

- The fatigue limit (infinite life with excellent cleanliness) for a garden variety 6222 is under 1000 lbs (SKF)or 3000 lbs (FAG). The SKF N222 is good for over 9000 lbs, So as far as life is concerned, you don't need all those bearings.

"As regarding the induced load, it’s a tough part. Probably it’s caused by the different bearing clearances. Is there any special method to figure this out? or just using beam theory?"
- The induced load results as soon as you have a staticially indeterminate problem, and results from stuff like bearing stiffness, bearing line of action, bearing clearance and bearing, shaft, and housing concentricity and eccentricity, which are tough enough to think about, and danged near impossible to really quantify. I'd start by measuring shaft deflection at the spindle nose at 100/200/300/400/500 lb radial load

With a short tool Shaft stiffness is hard to add with more bearings. An approximate weighting is spindle stiffness is 70% shaft/30% bearing stiffness. With an overhung or long tool the shaft stiffness (diameter) is all that will save me.

 

[feaeng1]

I really don’t how far off the load is since we don’t have any testing data. The load estimation is my calculation based on horsepower and the lower end of the working rpm.

All our bearings are from NSK. Yes, I agree. Two bearings are good enough.