Oil film between rollers in full complement bearing 2

[fleafry]

Would anyone be able to direct me to references on the topic of calculating the lubrication characteristics in full complement roller bearings, with particular reference to the oil film thickness between the counter-rotating surfaces at rolling element contact.

The application involves a planet bearing in which the rolling elements experience varying accelerations up to 130 g, particularly due to the centrifugal and coriolis effects. This leads to tangential forces - i.e. between adjacent rollers.

While some bearing manufacturers recommend full-complement bearings in planetary systems, there are also recommendations to avoid them in cases where significan roller-to-roller loads can be generated, and in high speed applications.

Basically I would like to be able to verify the concerns with some calculations. Papers with specifics on the first principle calcs seem to be hard to come by.

Any pointers would be appreciated.

With thanks
Francis

 

[israelkk]

Rolling Bearing Analysis by Tedric A. Harris (2000)
There is also a previous version.

 

[fleafry]

Thanks for that - I'll see if I can dig it up.

Anything online specific to this problem?

With thanks
Francis

 

[diamondjim]

One of the major concerns under
heavy loads and high speeds is
that the adjacent rollers surfaces
turn in opposite directions with
respect to each other thus magnifying
the problems. That is why they
use cages in these bearings or at
least rocommend them.

 

[fleafry]

Thank you - that is precisely the point I hope to be able to illustrate by means of some analysis from first principles. I would like to be able to do the hydrodynamic calculations to derive the expected oil film thickness for the particular application with our given the roller speeds, loads, lubricant properties etc.

Anybody got the equations somewhere for reference? Have there been any papers on this application (full-complement roller bearings with loads between rolling elements due to the centrifugal and coriolis accelerations introduced by the planet motion)?

Francis

 

[israelkk]

fleafry

You can buy the book here for a starting price of $130

http://www.bookfinder.com/search/?ac=sl&st=sl&qi=HIs8AAgao4ql9kF744.QGBe7zm0_1813580629_1:10:14

 

[EnglishMuffin]

One point worth mentioning here, since a planetary was mentioned, is that if you use caged bearings in the planets, because of centrifugal effects, they must have a land riding cage, not the more common roller riding type. That is why many planetary designers avoid the problem by using full complement bearings. And on balance, you may have fewer problems with those than you would with land riding cage types - depending on the speed.

 

[fleafry]

Thanks English Muffin. True - the cage has been the area of concern. I would like to determine what the critical speed and loading conditions would be for the full complement option.

 

[EnglishMuffin]

I could be wrong about this, but I think the main reason that full complement roller bearings have a lower speed rating than caged rollers has more to do with roller skewing than lubrication. I think the type of lubrication that occurs between the rollers should be primarily hydrodynamic, rather than the elasto-hydrodynamic behavior that occurs at the rolling contacts. So if anything, if it were a matter of lubrication alone, it would seem to me that the lack of a cage should actually promote better lube because of the higher relative sliding velocity.

 

[fleafry]

On the basis of some discussion with manufacturers, it appears that for signicant acceleration fields as in our design, an outer-ring guided cage would be more in our interests. It was argued that the caged arranged would provide more effective entrapment of the oil between sliding contacts, whereas the high relative velocity in opposite directions at the full-complement roller interface would tend to discourage oil entrapment.

 

[EnglishMuffin]

Well, if they are right I don't really deserve a star for my comments. If you apply either elasto-hydrodynamic or plain hydrodynamic theory, assuming enough oil gets to the contact zone, it would appear to me that the higher the entrainment velocity, the thicker the oil film should be (eg see "Ball Bearing Lubrication" - Hamrock & Dowson). You only need a tiny amount of oil if the surface finish is good. I would also think that the type of lubrication supply would be a factor - sometimes in a planetary oil is supplied through a radial feed hole directly into the bearing, and in such a case I find it hard to believe that the inter-roller contacts would become starved, regardless of rotational velocity, although you may not be using this method of course. However, there is no doubt that a caged bearing is generally superior if you can put up with the lower capacity (less rollers). But if it's a planetary, just make sure it's not a roller riding cage!

 

[fleafry]

Unfortunately I don't have Hamrock & Dowson handy, however the argument goes along these lines:

Whatever the relative speed between the two objects you are trying to build a film between you need a route to get oil into that contact. A classic journal (plain bearing) gives that (like in some areas of the cage contacts) in the form of the characteristic wedge.

In the case of a full complement bearing the rollers touching one another have a surface velocity in opposite directions, which does not entrain oil. In fact quite the opposite - oil is swept out of the contact. Despite giving a high relative velocity, good film formation will not occur.

 

[sreid]

Another thread took me to the SK web site. It seem to have been redone and is more user friendly. Below a link to bearings for planetary gears. They also mention that their CARB bearngs have advantages in this application.

http://www.skf.com/portal/skf/home/solutions?contentId=0.000336.000550.000909.000975.004102.000976