Wheel "adjustment" with cartridge bearings

capt_slog

I've been used to cups and cones. Cartridge bearings aren't a mystery as such, I've fitted plenty to various things, machinery to motorbikes, but I'm uncertain as to how they should be treated on a bike wheel as I still have the double nut arrangement as on the cup and cone wheel.

Normally on systems like this, there's a spacer between the two bearings and no matter how tight you make the nuts, you can't force the bearings towards each other, because you tighten onto the spacer.

Is there similar in these hubs? How tight are the nuts supposed to be?

Mavic Aksiums btw.

Thanks.


EDIT.
I did a bit of google (again, I couldn't find it when I looked before) and see that axles have a flange on them. Well, that's a surprise, every day is a school day . I now know how it's done, but would still like to know how tight the nuts should be against the bearing.

hambini

I'll assume you have 6903 bearings

Bearing clearance for a CN (Normal clearance bearing - usually unmarked) is 6 to 20um (microns). To take up the preload, the torque specified is actually too small to measure accurately with the average torque wrench. The guidance is to get it the point where this is no lateral play in the axle.

From a practical standpoint, if you have the same bearings in either side of the wheel, you'll find one takes up all of the axial movement whilst the other one is in pure radial, once the bearing has "worn in" you'll need to readjust them - usually tighten them a bit more - 25km of bedding in is the recommended interval.

Hope that helps.

Hambini

keezx

Cartridge bearings are not designed to take preload so don't put it on.
Just take out play > job done.

hambini

For the OP

If you go here http://www.hambini.com/download/ntnbearingbook.pdf

Download the book and then go to page A-58 it will give you the loading you require. Ref table is 8.1 and 8.3.

If I've calculated correct it's about 2N. If there is no preload on the bearings, the balls will skid as opposed to rotate on the races and you will kill them quickly.

capt_slog

hambini wrote:

I'll assume you have 6903 bearings

Bearing clearance for a CN (Normal clearance bearing - usually unmarked) is 6 to 20um (microns). To take up the preload, the torque specified is actually too small to measure accurately with the average torque wrench. The guidance is to get it the point where this is no lateral play in the axle.

From a practical standpoint, if you have the same bearings in either side of the wheel, you'll find one takes up all of the axial movement whilst the other one is in pure radial, once the bearing has "worn in" you'll need to readjust them - usually tighten them a bit more - 25km of bedding in is the recommended interval.

Hope that helps.

Hambini

Thanks, yes it does help.

It's a new bike, not ridden it yet, but I was having a look around it and took out the front wheel just for a look. I noticed that the feel of it was rather tight, basically there seemed to be a lot of drag on the bearings, and if I span the wheel, it would try to turn the axle in my fingers. Not much, but enough to cause me concern and to look into it.

I got out the cone spanners and undid the nuts, and then tightened them a just a little. I found any amount of preload will start to cause drag, I'm guessing that they will need to run in?

Thanks for taking the time to do the look-up too. Good guide that, I never realised there was so much to it.

hambini

Have a look on the side of the wheel bearings if you can get to them and see what brand they are. If you have no name brand bearings, the tolerances may not be so good. You get a tolerance stackup on all of the parts and this might cause the drag.

It certainly should not be trying to rotate your fingers. That indicates too much drag - or the grease is extremely stiff (unlikely but it's possible). you could try heating with a hairdryer then you'll know if it's the grease.

cycleclinic

You want just no play with the wheel clamped i the dropouts. That is how much preload you apply. None. Preload leads to early wear. Just check after the first ride and adjust if necessary.

hambini

Some more literature with regards preload written by SKF and NSK (two large bearing manufacturers). Both calculate out the preload based on bearing clearance.

http://www.skf.com/uk/products/bearings ... index.html

http://www.jp.nsk.com/app01/en/ctrg/ind ... at_e728g_6 page 157

And here's the SKF rule of thumb for the amount of preload if you don't want to do the heavy maths

http://www.skf.com/uk/products/bearings ... index.html

keezx

Does not apply for wheelbearings of bicycles, which are pretty slow rotating in relation to the rpm's they are designed for.

hambini

Keezx wrote:

Does not apply for wheelbearings of bicycles, which are pretty slow rotating in relation to the rpm's they are designed for.

Can you point me to a link so I can update the guidance on my website.

I'm slightly curious on this one as bottom bracket bearings which spin even slower than wheel bearings are preloaded. Either by a preload collar, wave washer (SRAM) or the preload star nut (Shimano)

keezx

SRAM GXP is not preloaded (in a threaded BB) and Shimano is just poor design to prevent the cyclist getting drunk by the movements of his feet.
The ONLY bearing in a bike that needs preload is the headset.