F1 in schools bearings

[xpensive]

I don't really know what to say here Tok, is this your idea of a joke or are you just trolling us?

What you suggest is similar to argue that you can change the ratio on a bike by shortening the chain.

[rjsa]

Hopefully you will regret what you posted:
On a F1 in Schools track of 20m with 30mm diam wheels the wheels turn 212 as the car traverses the track.

Typically they have 2mm axles with a 6x2 bearing:


d1 = 2mm d2=6mm ball = 0.8mm ID of outer race = 4.8mm OD of inner race = 3.2mm

Let us approximate what happens by assuming that the balls roll without slipping on both the inner and outer races.

If the bearing is mounted in the wheel and the axle is rigid in the body the then each rev of the bearing outer shell will drive the balls π*4.8 = 15mm

If the wheel is rigidly mounted on the axle which turns the inner race of the bearing which is mounted in the body the balls will be driven π*3.2 = 10mm

The balls run 50% further when the outer race rotates.

If the balls don’t slip then in the first case each ball will have run 15 mm along the inner race taking it 1.5 revs. In the second case the balls run 10mm along the inside of the outer race taking them just 240 degrees.

I guess I was a bit provocativecondensending, I'm sorry.

You are looking at it wrong. Try a little experiment. Get two rullers and a pencil. Position the rullers with the pencil in the middle. The rullers will simulate the races and the pencil the ball.

Now carefully move the upper ruller by let's say 5cm. If nothing slides, the pencil will move by 2.5cm. Always. In the ball bearing, the balls will go by half the realative displacement of the races in ideal conditions, that means the balls do not slide. The ball will comple one turn every two turns of the races. Get a big ball bearing and mark the races and one ball. Rotate the outer ring while you follow the ball. Now do the same with the inner one. See it for yourself.

[thisisatest]

i'd say that the outer race rotating clockwise one rotation with the inner race stationary
=
the outer race stationary with the inner race making one rotation counter-clockwise.
it's a matter of perspective, therefore the bearing distance traveled is the same.

[xpensive]

Good Lord, is this the level to where this great technical forum has come down to, embarrassing, moderator please!

[tok-tokkie]

I was provoked by the post. It was quite clear to me how it works but as a general explanation had not worked I decided to give a concrete example. So I wrote that example quickly before supper - ended up a little late. After supper I wondered how the hell the inner race could have rotated more than 1 rev then I realised it is an epicyclic gearing problem. My calculations are quite correct for the case where the planet wheel (ball in our case) is held still but that is not what is happening in our case. You need to add or subtract 1 rev to the answers I gave. So outer rotating sends the balls 540 - 360 = 180° Inner rotating sends it 360 - 240 = 120°. Still the 50% difference.

I expected to be roasted.

[xpensive]

This could perhaps be useful; http://en.wikipedia.org/wiki/Planetary_gear

Scroll down to the chart called "Sketch and output speed of planetary gearings", then behold the two cases on the right on the second row from the top, speed of the driven planet-carrier are the same, no matter if sun- or ring-gear is the driving one.

[tok-tokkie]

X I don’t see Sketch and output speed of planetary gearings on that Wiki page.

OK to hopefully convince you guys I have photod a bearing. I have a 2x6 bearing but have instead used a 6306 bearing as it is easier to mark & photograph. The relative lengths of the inner & outer ball tracks are not in the ratio 10:15 so the difference in the ball travel is not as great when using a 6306.


Before rotation.


Inside race turned 1 rev.


Outside race turned 1 rev

[xpensive]

Right, let's try it this way then;

You mean that the friction of the bearing will be different if you spin the inner or the outer race, everything else equal?

[rjsa]

Would'nt you say the ball travelled the same distance in opposite directions? It looks like that to me. And that's exactly wah I'd expect.

[tok-tokkie]

Right, let's try it this way then;

You mean that the friction of the bearing will be different if you spin the inner or the outer race, everything else equal?

Yes becuse the friction losses are proportional to the distance traveled. The friction force is mu*load and the work done is the distance that force travels.

rjso: i turned the inner & outer races clockwise starting from the same initial position in each case. The difference in distance travelled is in accordance with the mechanics of epicyclic gears - as I corrected myself. Simple: repeat the test I did.

[xpensive]

When you analyze the powerloss of a roller bearing, you estimate the torque needed to turn one race relative the other at a certain speed, to say that it would make any difference how they independently turn is a misconception.

http://www.skf.com/portal/skf/home/prod ... nk=1_0_37b

If the inner race turns 500 rpm clockwise and the outer race 1000 rpm anti-clockwise, the speed of interest for calculating the powerloss is naturally 1500 rpm.

[rjsa]

This thing is starting to make noises in my head, I'll go after a ball bearing myself. I'm not sure of anything now since I keep turning these things in my had and won't have the patience to run numbers at all.

[tok-tokkie]

X, in my first post I said this (emphasis added):

Mounting the bearing in the wheel with the axle fixed in the body makes the balls run much faster compared to bearing in the body with rotating axle. 1 rev of the wheel makes the balls travel much further if the outer race turns as against the inner race. In your application you are looking for minute reductions in resistance.

I suggest that your formula is a first approximation of the truth. Bear in mind that the certainty of the data is pretty suspect so real precision is not required of the formula. I still think what I suggest is valid from a fundamental analysis point of view.

[xpensive]

...
I suggest that your formula is a first approximation of the truth. Bear in mind that the certainty of the data is pretty suspect so real precision is not required of the formula. I still think what I suggest is valid from a fundamental analysis point of view.

I recommend you to make contact with SKF to share your groundbreaking findings, I'm sure they will be most appreciative.

[Richard]

Are we agreed that the bearing is subject to relative movement of the two rings? The direction of movement relative to outside observers is irrelevant? Agreed?

So that brings us onto Tok's experiment , it is theoretically not possible. Is it repeatable? What secondary effects are going on to result in different movements?