h a l f b a k e r y
There's no money in it.

idea: add, search, annotate, link, view, overview, recent, by name, random

meta:

account: browse anonymously, or get an account and write.

 user: pass:
register,

# Bi-Layer Bearings

Two rows of rollers can actually work
 (-1) [vote for, against]

See the linked "Tri-Layer Bearings" Idea for the rationale behind that one --and this one. Basically, the goal is to eliminate the "cage" that holds bearing-balls or bearing-rollers apart from each other, inside the overall bearing assembly.

You may notice that in most of the sketches that were linked to that other Idea, the diameters of the balls or rollers were pretty much the same. So, consider this ASCII sketch:
_O_O_O_O (ignore the underscores)
O_O_O_O

if that is a side view of rollers in-between the "inner race" and the "outer race" of an overall bearing assembly, then it can't work. Motion of, say, the inner race will cause the adjacent rollers to rotate, which causes the next layer of rollers to rotate --which forces the outer race to rotate, also, leaving no independent motion between the inner and outer race. Not good!

However, a few hours ago I happened to wonder what the situation would be if the two sets of rollers had different diameters, something like this:
_O_O
o_o_o

In this case the small rollers would not rotate at the same speed as the larger rollers, and so one bearing-race might be able to hold still while the other rotates.

Furthermore, I had an opportunity to try to assemble a mock-up, using a number of different-sized cylinders of metal that happened to be available for playing-with, at my workplace. It took a while to find some sizes that could actually fit together fairly tightly, but I succeeded.

AND ... when I rotated the "inner race" part, the intermediary rollers did indeed move as hoped, while the "outer race" part held still! Hooray!

On the other hand, the whole assembly almost immediately also experienced a "bunching-up" problem (as described in the other linked Idea), which means that that issue absolutely must be solved before any bi-layer (to say nothing of tri-layer!) bearings could be marketed.

But for the HalfBakery, this is enough for now.

 — Vernon, Jan 01 2012

Tri-Layer Bearings Tri-Layer_20Bearings
As mentioned in the main text. [Vernon, Jan 01 2012]

 I can't see this working as described [Vernon]. It would function if all diameters remained the same with spacing between them. Smaller diameter bearings will have greater rpm's than larger ones and will have to slip in order to keep pace.

Pending actually having read this correctly of course.
 — 2 fries shy of a happy meal, Jan 01 2012

 [2 fries], it actually does work (as the mock-up proved); that's why I posted it. And there need be no slip for it to work, either. Slipping of the rollers can (did!) occur for other reasons, such as the weight of the inner race, but as far as pure rotations are concerned, it does indeed work perfectly.

And, I still DON'T expect it to work if all the rollers are the same diameter, in-between the two bearing races.
 — Vernon, Jan 01 2012

o_o
 — simonj, Jan 01 2012

 I'm not sure I see the application here; is this an attempt to build more compact bearings that will support extra weight, or just to make the bearings 'slicker'?

 Problems I foresee:

 More parts = higher manufacture cost (and thus higher end-cost), more W&T, more parts to fail unexpectedly

 More moving parts = more friction, more frequent lubrication required

And since the rollers of a bearing must have a tiny amount of wash (typically measuring >.001") inside the races, additional races = more overall wash in the bearing = more eccentric motion in the component mounted on it. This is where your 'bunching up' comes from, most likely.
 — Alterother, Jan 01 2012

 [Alterother], the goal is to eliminate the cage, which does more than prevent bunching. The cage also prevents adjacent rollers (or balls) from rubbing each other the wrong way, as described in the "Tri-Layer Bearings" Idea. A side-effect of having a cage is that lubrication is required; most bearings die when their cages wear out.

 If rollers can be kept apart by some other means than a cage --such as the separate layer(s) of rollers described in these Ideas-- then because all these parts simply roll without rubbing anything, lubrication would not be necessary.

 You might note that my March 10, 2010 anno in the other Idea talks about rollers that aren't perfectly round, to prevent the kind of slipping that leads to bunching. I'm thinking they could be made in quantity by an extrusion process.

One general type of shape that CERTAINLY would work, to prevent the roller-layers bunching up, is the gear shape. Unfortunately, gears experience some rubbing which would require lubrication. So I was considering the idea of something in-between perfectly round and perfectly gear-shaped, that could roll without rubbing, but also not slip out of position like gears won't. I still tend to think that notion could be workable.
 — Vernon, Jan 01 2012

 [annotate]

back: main index