Cylindrical roller thrust bearing: Inherent sliders?

[Juanda]

TL;DR Summary

Do they slide by design? It seems impossible to have proper rotation.

In ideal conditions, thrust ball bearings will be able to roll without sliding having a single contact point with the lower and upper ring.

(From a different post but applicable)

However, can cylindrical roller thrust bearings roll without sliding? I understand cylinders can be preferable when loads are very big because they spread the pressure on a bigger area. But in this case, even in the ideal scenario where the contact happens in a line, there'll still be relative movement between the surfaces.

The cylinder can't make a circle without sliding. It'd need to rotate slower near the end closer to the center of rotation of the housing washer which is impossible. I guess it'd be doable with conical geometries but that doesn't make cylindrical roller thrust bearings any less real.

How is that sliding in cylindrical roller thrust bearings justified within a design? I guess you sometimes need to take bigger loads and can't afford to insert bigger thrust ball bearings due to space constraints but having inherent sliding like that feels wrong.
Is the additional cost of tapered roller thrust bearings significant enough to justify choosing the cheaper cylindrical roller thrust bearings, even though they come with added friction and wear?

Do tapered roller thrust bearings have sliding friction in this area? It seems it's necessary to have contact there because the axial load will push the cones outwards so that shoulder limits the movement.

 

[Baluncore]

Cylindrical rollers, and needle rollers, are often used to carry axial loads, but I have never come across a cylindrical roller thrust bearing in use.

The cylinders would need to be short, more like narrow wheels, in order to carry the axial load on a flat race. I would expect that each virtual roller, would be made from shorter rollers, that are housed in the same brass cage. That would enable the inner and outer ends to rotate at different velocities.

Thrust Cylindrical Roller Bearings can carry high axial loads, but are only available from NSK with bore diameters of 35 – 320 mm. Smaller thrust bearings would use short needle rollers, and I expect, not be operated with a preload.

 

[Juanda]

The cylinders would need to be short, more like narrow wheels, in order to carry the axial load on a flat race. I would expect that each virtual roller, would be made from shorter rollers, that are housed in the same brass cage. That would enable the inner and outer ends to rotate at different velocities.

That sounds like a reasonable design to avoid some of the sliding although I haven't encountered it so far.

 

[Tom.G]

I remember that Record Players (at least the higher-end ones) used roller thrust bearings under the turntable. The cheaper players used ball bearings.

I never inspected the roller bearings to see if they were tapered though.

 

[Baluncore]

The cheaper players used ball bearings.

I remember a bronze sleeve bearing to oppose the side force of the drive wheel or belt, with a single ball on the axis of the sleeve, to carry the mass of the platter.

I am surprised bearings that might generate noise were used.

 

[Tom.G]

I remember a bronze sleeve bearing to oppose the side force of the drive wheel or belt, with a single ball on the axis of the sleeve, to carry the mass of the platter.

I am surprised bearings that might generate noise were used.

Yeah, I had forgotten about those! They seemed to be the 'First' player that the kids would get. By their teen years they either lost interest in records or wanted something a bit better.

The sleeve bearings were pressed in, making it impractical to clean out the dried oil. About all you could do is add more oil and cross your fingers.

(Are we showing our age here?)

Cheers,
Tom

 

[Baluncore]

(Are we showing our age here?)

I would call it our "broad engineering experience".

 

[chazemz]

NASA operate in less than ideal conditions and their bearing are made to roll not slide. Just type in NASA Bearing to your search and click on Ball bearing analysis with the ORBIS tool. These will be expensive and well made bearings. I would think that cheap bearings would have a short life span and tend to fail and slide quite quickly.

 

[Lnewqban]

None of these types of bearings can be subjected to any radial or angular misalignment, or to any significant radial load.

Therefore, some radial "guide" bearings should be incorporated in the design of the machine, just to keep shafts rotating concentrically and preventing any radial load (even due to unbalanced masses) to reach the thrust bearing.

A typical application of this concept happens in vertical centrifugal pumps, in which the thrust bearing is carrying the loads of the weight, plus the differential pressure (front-back) of the impellers.

 

[Juanda]

I'm still searching for books that cover angular contact bearings and the properties of the assembly depending on the installation. Out of coincidence, I found a picture showing just what you said.

The cylinders would need to be short, more like narrow wheels, in order to carry the axial load on a flat race. I would expect that each virtual roller, would be made from shorter rollers, that are housed in the same brass cage. That would enable the inner and outer ends to rotate at different velocities.

I could be misinterpreting the picture but I'd say that it shows two cylinders instead of having a long unique cylinder.
The source is Fundamentals of Machine Component Design 7th Edition by Robert C. Juvinall and Kurt M. Marshek.
Anyway, back to searching.