Bike Gremlin

Roller bearings are different from ball bearings (mostly used for bicycles). They produce more drag and are better suited for heavier load applications. One roller has more surface contact with races than all the balls of a ball bearing together provide.

Googled their site:

" As a general rule, ball bearings should always have an operational clearance that is virtually zero, or there may be a slight preload. Cylindrical, spherical and CARB toroidal roller bearings, on the other hand, should always have some residual clearance - however small - in operation."

The link:
Bearing internal clearance

rodteague

I'm asking this again, since it appears it was lost in the "kerfuffle"

Racing Dan

I know what the difference is. What are you getting at?

Bike Gremlin

That the text you quoted from SKS site is not relevant for bicycle cup and cone ball bearings.

Racing Dan

Why not? Its referencing ALL bearings with special emphasis on rollerbearings ect. The second sentence does not contradict the first sentence or negate it.

SquidPuppet

Pairs of ball (like on a bike) bearings should be preloaded before being put into service.

Source AST Bearing company

It appears that AST Bearing Company agree with Jobst Brandt.

I would think the third bullet point would be of significance on bicycles considering their service environment, which includes impacts caused by riding surface imperfections and the fact that they lean in both directions.

Preloading Ball Bearings: Full AST Guide - Bearings : Blog : AST Bearings

3alarmer 

...I honestly don't know, but there was quite a back and forth over that one, too. Proof that a QR compresses the axle?

Bike Gremlin

Apart from other links and quotes provided by other participants, I can only refer you to the SKS quote I provided - that mentions ball bearings explicitly.

Have you run hubs with play (after they are mounted / QR tightened and closed) and gotten away without pitting after a few thousands of kilometers?

3alarmer 

.
...inch by inch.....step by step.

Racing Dan

Sock puppet

It appears that you one again ignore that the text you quoted says "In certain applications ... ". We already covered this. A bicycle is not a lathe or truck wheel.

Racing Dan

That is really lazy, but fine. Im guessing your argument is that this:

"" As a general rule, ball bearings should always have an operational clearance that is virtually zero, or there may be a slight preload."

contradicts this:

""Although all bearing types can run with some
preload, SKF recommends a positive operating
clearance. "

Imo, it does not. It states that some, very little or more, positive clearance is recommended as a general rule.

SquidPuppet

A lathe sits on a concrete floor and doesn't move, or lean, or go over bumps. A bicycle wheel is a perfect example of a PAIR of BALL bearings that should be preloaded.

Bicycle wheels don't spin in a vacuum.

That's the same reason SKF preloads their bottom bracket bearings.

Without preload, the balls are free to bash back and forth, up and down, over and over, repeatedly making blunt force impacts with the races.

With preload they are forced to stay where they belong and ROLL in between the races.


TOO FUNNY. SKF has design software that assists engineers in determining the proper amount of preload.

You can find it here, on their "Selecting Preload" page.




Selecting preload

SquidPuppet

An empty wheelbarrow has no load.

A wheelbarrow containing a single feather has a load. A very light load.

It's a matter of degrees.

Bicycle mechanics with sensitive fingers can achieve feather-like preloads.

That's direct from the SKF website.

prathmann

Yes, but that seems to agree well with the graph presented previously. The absolute optimum durability should be with a slight amount of preload and no clearance. But the decline is rather gradual when small amounts of clearance (i.e. play) are present. So maybe the bearing would have lasted up to 60000 km but only lasted 50000 km due to a little bit of play.

rodteague

Well, that cleared things up.....NOT! I read about half the thread and it's just as entertaining as this one. I'll have to take your position as well; I still don't know Thanks for the link BTW, I'll finish reading it later.

Best Regards

Rod

SquidPuppet

The dumb engineer resides in the compression camp.

https://www.sheldonbrown.com/brandt/...djustment.html

3alarmer 

rodteague

Although there are physics and scientific principles at bear here, your comment as well as many others, leads me to the conclusion that there are other dynamics that can't be measured or quantified. There is an element of craftsmanship and intuition that will take some time to master. I'm in the process of building two wheels for a bicycle trailer and it is the first time I have done this by myself. I'm taking all this to heart, so wish me luck! This grasshopper needs a good wheel whisperer

Best Regards

Rod

rodteague

Does he build wheels?

3alarmer 

...

SquidPuppet

Blind dudes are the best at setting preload, obviously.

sweeks

Some people can't seem to accept that steel, an elastic material, elongates (however slightly) under tensile loading and shortens (however slightly) under compressive loading even if it is in the form of a rod or a cylinder.
Steve

SquidPuppet

I believe that hollow axles compress when the QR is closed. Steel expands and contracts from mere temperature changes. How could it NOT succumb to physical forces? We can bend steel. Bending is compressing and stretching at the same time.

Also, the other theories for the elimination of play caused by closing the QR have gaping holes in them.

Bowing of the axle? No way. The cups and cones would then be misaligned and you'd never achieve a good preload setting. And things would wear in uneven ways, at an accelerated pace.

Slack on the axle between the lock nut and cone? Nope. They are locked together after all, immobilized. The fact that they don't move on the axle is clearly demonstrated when you crank down on a axle nuts of a solid axle wheel with a big wrench and the preload adjustment remains unchanged.

sweeks

You don't have to "believe" this... you either accept it as a fact or you don't!
/PreachingToChoir Mode
The rails trains run on are deformed slightly under the weight of the train, and the wheels are slightly flattened.
Also, bending does involve compression and elongation at the same time, but in different places: a cylinder (or rod) that is subject to bending forces is under tensile loading on the outside of the curve and compressive loading on the inside of the curve.

Agreed, but when the axle is loaded there is probably some bowing which is nevertheless tolerated by the bearings.
/DevilsAdvocate Mode

Exactly!
Steve

woodcraft

Hmm...