Robert P

I was in the shop of the local Trek dealer where I recently got my Trek and was asking the head tech about adjusting the cup/cone bearings comparing notes to what folks said in here regarding a thread I started on the subject where it was brought up that the QR compresses the axle slightly and it has to be considered when adjusting the bearings. I've also seen articles elsewhere that state this. When I mentioned this he completely disregarded the notion as absolute nonsense.

So I've got conflicting assertions from bike enthusiasts. It seems important enough to try to get to the bottom of it.

Any way to demonstrate that the QR does in fact compress the axle?

JanMM

Did the tech dispute that closing the QR can change the adjustment of the bearings or just the mechanism of how that happens?

79pmooney

I don't know how much aluminum skewers compress. Maybe not much. But steel skewers? Yeah, they do. BITD wa always set up cup and cone bearings with a touch of play, knowing if we didn't were were shortening the life of the bearings. That play we could feel with our fingers on the axle, wheel off the bike. Put the wheel in and tighten the skewer and that play was gone. Think about it. That play lost is because the cones are now closer together, but they were already bottomed out on the threads by the locknuts. So they went nowhere on the axle. The axle shortened in direct proportion to the amount the QR skewer lengthened (stretched). An axiom of steel is that if you apply a stress to it, it will stretch or compress in direct proportion to the stress applied. Always. Modulus of elasticity, You can look it up. It's sophomore level engineering.

How old was the head tech? The BITD I am referring to is the 1970s when all hubs were cup and cone bearings (except Phil Wood). I'm guessng he has spent his life with sealed bearings and never had to know this.

Ben

thumpism

Try what was mentioned above with a bare QR front hub that is not built into a wheel. Without the additional mass of spokes and rim and their flywheel effect you can better see the result. Adjusted perfectly with no play then clamped into the fork with the skewer, you will probably discover that the hub is now tight instead of perfectly adjusted. Repeat with the hub adjusted to have a slight amount of play and you will likely find that when clamped in it is now perfectly adjusted. A QR skewer is a mighty mechanism.

ecnewell 

It may depend on the type of QR used, too. Modern external cam skewers don't compress with the same power of internal QRs, so might not have the same effect?

brianmcg123

It depends on how this was asked.

If you stated "I heard on an internet forum..." His reflex response of "absolute nonsense" would probably be correct in 99.9% of cases. He may not have even heard what you were talking about.

sweeks

Yes. How else would you explain the disappearance of clearance between cones rigidly fixed to the hollow axle?

Steel is an elastic material which can stretch or compress according to its dimensions and the forces acting on it. When a straight steel rod bends, one side gets longer and the other side gets shorter (under tension and compression, respectively). The same thing happens with the hollow axle, which compresses slightly as the quick-release skewer stretches slightly. As explained in the Jobst Brandt article on "Wheel Bearing Adjustment" (here: https://www.sheldonbrown.com/brandt/...djustment.html),

"Wheels with quick release (QR) axles present an additional problem in that closing the QR alters bearing clearance. Closing the lever requires increasing manual force with a slight over-center feel near the end of the stroke. This lever force arises from compressing the hollow axle and stretching the skewer. The ratio of elastic length change between axle and skewer is that of their cross sectional area and active lengths."

The tech guy you quoted doesn't seem to know much about material science. Just saying...
Steve

Bill Kapaun

Just lightly tighten the QR. Just enough so the wheel doesn't fall out on its own weight.
Feel the rim (near the brake pads for example) and you should feel a little side play.
Tighten QR firmly and compare.

dsbrantjr

And their respective moduli of elasticity.

davidad

When I adjust the bearings on my shimano hubs I put the QR through a 7/16" nut on the right side of the hub and close it. I then make the final adjustments to the bearings so that there is a slight amount of drag I can feel. When I open the QR 1/4 there is a small amount of play in the bearings. I tighten up the lock nut and mount the wheel on the bike.\
This puts a small amount of preload with no play at the rim.

Rowan

I wish we had an engineer with knowledge of metal compression. But I will go with the Park website and logic and suggest that hollow axles in fact bow, not compress, thus bringing the cones closer together. The force required to compress a steel axle -- albeit follow -- would be much much greater than a QR could apply with human hands closing it.

One way to demonstrate this may well be to use a short straight-edge steel ruler and check the straightness of the axle prior to QR use, and then after the QR is closed.

hueyhoolihan

as someone stated in another post a while back, i'm going with any apparent compression is really just the slack being taken out of the threads on the two cones and two locknuts. this alone could possibly affect the proper adjustment of the bearings.

IMO, to compress the axle we would ultimately have to compress two aluminum nuts on the skewer (if made of aluminum) and possibly two aluminum or carbon fiber dropouts to a point that they resisted an additional compressive force greater than that of hardened steel, before the axle would even begin to compress. highly unlikely.

Robert P

He treated the notion of the QR changing anything after the cones have been adjusted and tightened such as compressing the axle as nonsense. Words to the effect that there's no way it's going to compress a steel axle.


The way I put it was along the lines of "it's my understanding that you have to allow for the QR compressing the axle - I didn't specify where that understanding came from.

CliffordK

I've always adjusted my cones without the QR.

However, you could probably demonstrate the effect easily enough.

Adjust your cones until they're perfect (use good quality cones/bearings). Then toss two 5mm spacers onto the ends of the axle, and clamp down the skewer as tight as you normally would do. Then feel whether the bearings are rough (too tight) or not.

Bike Gremlin

Rear axles are a bit stronger, thicker, with a larger diameter and they don't compress nearly as much as the front ones, but even they do a bit. It's easily shown by elimination of slight play once the QR is tightened.

Not many mechanics know the physics and mechanics behind the things they service.

Rowan

Out of all the solutions in this thread and the other, this just seems the simplest, except for sourcing the two 5mm spacers. Heck, just cutting off the dropouts from the rear of a disposable bike seems easy enough.

AnkleWork

And you want more conflicting assertions? Who will you trust?

I'd say just try it yourself then you would know, but people hate first-hand experience.

CliffordK

Shimano uses 5mm spacers on the left side of several of their rear hubs (sometimes 3 or 4 spacers for 135mm hubs). Someone has to have a few laying around. Ask your local shop or co-op.

You could also do the same thing with washers, but you'd need a pile of them.

altenwrencher

Decades ago at the shop I worked at, we had a set of Campagnolo dropouts. Adjust the bearings to run smoothly with no play, then put the dropouts and a QR skewer in place and tighten. Voila, tight and grainy running!

Whether its compression of the axle, slack being pressed out of the cone-locknut-axle threads interface, or the grease screaming for mercy, the effect is that the QR does make the bearing run noticeably tighter.

Bike Gremlin

What I do to set the preload:

scott967

As a practical matter, I don't think the physics or mechanics matter much. It's an engineering problem.

1. How much force is developed when the QR is closed. Best way to determine is testing with strain gauge I would think.

2. How much does the length of an axle change when subjected to compression force comparable to that found in step 1. Again, I would do this by test.

Sure, you can get a general idea by doing the calcs, but I would want the test data to confirm my assumptions (since the calcs pretty much always require simplifying assumptions). Of course, a good engineer knows when the calcs are "good enough", but I'm not a good engineer.

Lacking a test setup, the accomplished technician gets the answer by using the "bearing feel" test described in the thread. Because the technician doesn't need to know Young's Modulus, rather whether the bearings work as intended or not.

The difference between the good engineer and the better engineer, is the better engineer knows when to trust the tech, and when he should test for himself.

scott s.

Tape2012

I have been a licensed Professional Engineer for 25+ years. Steel absolutely compresses according to its modulus of elasticity. If it didn't things like your bathroom scale, the strain gauge mentioned above, all the way up to truck scales wouldn't work. Also, the cam action of a QR mechanism can produce a surprising amount of tension. While it is possible that there is some bowing involved as well, there is definitely enough force involved to compress the axle enough to change bearing adjustment. Just the fact that you can close a QR lever means the skewer is lengthening and the axle is compressing.

The amount of compression could be calculated and measured, but there is plenty of anecdotal evidence from folks on this board to make that kind of work unnecessary I would think.

maddog34

not one mention SHIFTING of the cone/jam nut on the threads.... those tiny, overloaded threads FLEX.

bowing axles.... :-D
what happens when an axle bends, again?
the axle would only bend if loaded in a VERTICAL plane... since it's supporting weight on both ends when the rider is on the bike.... what happens to the axle? it FLEXES in the middle... Inboard, and OUTBOARD of the bearing surfaces. The cone/jam increases resistance to the outboard flexing, but it would still flex a bit... not enough to take up the 1 to 2 thousandths of slop....
if an axle set remained bolted in the same position for an extended time, and the bike was ridden extensively, IF the AXLE FLEXING/BOWING was a real issue, we'd all see wear on only ONE AREA of a cone, when the axle was removed...

If the axle was bent to begin with, the axle would indeed bow when over-compressed by the sideloading of the QR...

If an axle uses nuts, instead of a QR... what happens to the central part of the axle, when the nuts are tightened? NOTHING. no compression whatsoever.... the THREADS carry the load imparted by the axle nuts.

SIDE LOADING does.... what, again?

side loads would compress the contact surface of the balls, cones, and races.... and the HUB, but just a tiny, totally un-noticeable bit.... and flex the threads... stretching the outer face, and compressing the inner face, with some deformation of the thread material,in the case of tightening the CRAP outta the QR... which stretches much more than the axle ever gets compressed.

and then, there's all those serrations on the faces of the QR mating surface, and the jam nuts... what do THEY DO, again? DEFORM the frame material. OOPS! there goes the "axle compressing force", right out the window.... especially in the case of aluminum dropouts, eh?

Makes fresh batch of popcorn... this is great entertainment... beats watching trump make a bigger idiot out of himself.

the threads flex. They are the weak point, in the case of a steel dropout. Aluminum dropouts eat up a lot of the force, when incorporated into this equation.... as do the QR pieces themselves.... once the dropouts are deformed enough, they assume the shape of the serrations, and the forces are distributed more evenly... distribute the forces more uniformly, and less deformation of the dropout mating surfaces is realized.

"carbon fiber DROPOUTS"..... tee hee.

Bike Gremlin

+1
Pretty much sums up the difference between a (good) mechanic and a (good) engineer, as it was explained to me once by a (very good) mechanic.

For all I know, the axle doesnt bow, it (slightly) compresses.

Rowan

Irrespective of what happens on tightening the QR (and I am still persuaded more by the people at Park Tools) I am grateful to the posters in both these threads for showing how simple and elegantly adjusting cup-and-cone ball bearings (and to an extent cartridge bearings) can be. I think I have a couple of 5mm spaces somewhere in a bin. Certainly beats adjusting, checking, readjusting, checking, readjusting, and checking...

We may not know the theory, or the reality, but we sure know how to get things working!