Servicing Wheel Bearings

[Jameshow]

That isn't my experience. You're talking about the ability of the axle nuts to squeeze the axle, shortening it so the bearings are too tight.



The lock nuts lock the cone nuts. No? They do on my bikes. Then the axle nuts squeeze the dropouts up to the lock nut which locks the cone nut. It works for me.

Yeap the ability of the skewer to compress the axle will proportional to the cross sectional area of the skewer Vs the cross sectional area of the hollow axle which I would guess is 1-5?

Also the amount of elongation for steel in tension Vs compression needs to be factored in, which I cannot remember since learning it almost 30 years ago!!

 

[Ajax Bay]

I once replaced the bearings with cheap bearings and they didn't last 2mins now I either buy hideously priced Shimano bearings or buy them from a nearby industrial bearing supplier who do a nice line in bike bits.

I've no idea about the quality of the ball bearings I bought. They came from a place that sells that kind of stuff. It seemed a reputable company. I figured it out that eighteen ball bearings per wheel, cost a little under $4. That included the shipping cost when I bought fifty 1/4" and fifty 3/16" balls.

Charley - be aware that Tom is talking about cartridge bearings, not (ball) bearings.

Bicycle bearings: everything you need to know​

 

[presta]

understand you, perfectly, but I disagree with the concept. I've had no problems with the way I do things. Why should I change my ways just because someone else says it needs to be done differently?

The tension on the skewer compresses the axle, the cones get closer together, and if you didn't leave enough clearance when you set them they will bind. It's no use whatsoever feeling for binding with the wheel in the dropout, because the force you apply at the periphery has 100:1 mechanical advantage at the axle radius, which will totally mask small amounts of binding. What you need is a pair of dummy dropouts so that you can tension the skewer with the wheel out, and still twiddle the axle in your fingers to feel for binding.

In order to set the bearing quickly and easily without needing trial and error after the first time:
Tighten the cone gently with your fingers to take all the clearance out, then back it off, and note the angle that you turned it using the cone spanner and the spokes as a protractor. If you keep track of the angle during the trial & error process, then when you've finally got the right adjustment you can use the same angle next time to get it right first time without all the faffing. On my Deore hubs backing the cones off about 100 degrees gives about the right clearance for me, but it depends on how tight you like your skewers.

 

[CharleyFarley]

Charley - be aware that Tom is talking about cartridge bearings, not (ball) bearings.

Bicycle bearings: everything you need to know​

I already found that website. I also found a website that tells you how to tell if bearings are too loose or too tight. It said to turn the bike upside-down and spin the wheel. If it spins freely, it's good, which mine is. And if it soon slows down it's too tight, which mine isn't. Evidently, I did it right but according to someone else, you can't do it that way. So who am I to believe? That's why I usually do things my way. And if I'm not having any problems with my mechanical work, how can it be wrong? My fat bike recently turned 12,000 miles, and I'm the only one who has serviced it.

And if you look at my O.P. you'll see I was talking about how often bearings should be serviced, not how they should be serviced.

 

[Ajax Bay]

Yeap the ability of the skewer to compress the axle will proportional to the cross sectional area of the skewer Vs the cross sectional area of the hollow axle which I would guess is 1-5?
Also the amount of elongation for steel in tension Vs compression needs to be factored in, which I cannot remember since learning it almost 30 years ago!!

Though irrelevant to the force which a QR can exert I thought I'd just check that. The actual stretching/compressing distance is tiny but that is what makes the difference to the angular cone/cup bearings: with compression they're just right (see @presta post for detail).
Cross sectional area of a steel QR skewer (diameter 4.8mm) is 18mm^2. The compression force when closed securely will compress the hollow steel axle (the aluminium hub can be disregarded). The axle is 9mm dia with a 5.2mm 'hole' so cross sectional effective area 42mm^2. So calculated ratio (as opposed to 1:5 guess) is 1:2.3.
Young's Modulus of steel is 200GPa btw.
HTH

I was hoping to learn how many miles it should be between services . . . [and] I now know that 3,000 miles is too many, even just riding dry pavement. [In] my O.P. . . . I was talking about how often bearings should be serviced

A reasonable interval between servicing bearings depends on a variety of factors and you have a heavy electric bike and a no-name hubbed wheel on which you've done approx 1000 miles a year. I should think the rear hub bearings would appreciate a service at a shorter interval than the front hub (the former taking much greater load).

 

[CharleyFarley]

...and you have a heavy electric bike and a no-name hubbed wheel on which you've done approx 1000 miles a year...

I don't have an electric bike, and I never even mentioned electric bikes.

 

[Petrichorwheels]

Ball bearing hubs DO tighten every so slightly when locked into a wheel. I was told by someone I consider an expert to follow the just slack when out and just right when in the wheel. My practical fettling has shown me that this is indeed the case.

 

[Ajax Bay]

Sorry OP - false assumption about a bike called 'Electra' - calling her 'heavy' was also unfair: 'well-built' rather.
https://electra.trekbikes.com/gb/en...cruiser-lux/cruiser-lux-7d-step-over/p/17040/
"Simply elegant. Casually cool. As a beach cruiser should be. You can feel the Electra difference when you ride it – our patented Flat Foot Technology® adds a whole new level of comfort and control. With a range of 7 gears and linear pull brakes, this ride is ready to roll [once the rider has added grease to the wheel hub bearings]

The bike is an Electra - Lux 7D. It has no QRs, just axle nuts.

 

[CharleyFarley]

Sorry OP - false assumption about a bike called 'Electra' - calling her 'heavy' was also unfair: 'well-built' rather.
https://electra.trekbikes.com/gb/en...cruiser-lux/cruiser-lux-7d-step-over/p/17040/
"Simply elegant. Casually cool. As a beach cruiser should be. You can feel the Electra difference when you ride it – our patented Flat Foot Technology® adds a whole new level of comfort and control. With a range of 7 gears and linear pull brakes, this ride is ready to roll [once the rider has added grease to the wheel hub bearings]

My first cruiser was an aluminum 'La Jolla' I bought from a neighbor. I soon realized that I don't care for single-speed bikes with a coaster brake. I sold it and then found a bike shop (I was a new resident of Florida.) I bought a 'Sun' cruiser from the bike shop. Steel-framed and seven-speed. It was better but still not entirely happy with it, so I went back to the bike shop and traded it for a 'Sun' fat bike. It was heavy and had a hub gear. I noticed a lug on the chainstay and asked the shop if they could convert it to a seven-speed. They said they could and they did it before I took it home. It was a problem from the start because when I went to collect it, I found the first gear was the equivalent of fourth. They assumed that the chainring designed for the hub gear was okay for the derailleur. It took six weeks to get the right chainring. I also contacted 'Sun' because the chain rubbed the tire in the lowest gear, and they said even though it has the lug, it was not designed for a derailleur. They didn't even test ride it after they converted it. That was one of seven problems I had with that bike shop before I quit using them.

Still wanting a cruiser I found another bike shop and bought the Electra from them. It really is a good bike. I modified it with a Cannondale memory foam saddle, different handlebars, adjustable stem to bring the handlebars closer, and swapped the twist shifter for a lever shifter. I also added Electra fenders to it. Then at 600 miles the wheels didn't want to turn; the bearings were crunching pretty badly. Lack of grease. I put new ball bearings in and did another 3,000 miles, and that's when I wondered about how often they should be serviced.

 

[gbb]

If its any consolation Charley, I never really understood WHY compressing a QR squewer then compresses the hub bearings/cup/ cone. I mean exactly how does that happen ?, you're working with a steel axle, the cones are fixed on its length...I always had trouble visualising it.
I know I could research it but can't really be bothered. All I do know is early on, when I've adjusted them what seemed just right off the bike, they're then too tight in the frame? Its a generally accepted and documented thing.

Regardless, there are a couple other things that can mess up adjustment pdq....
I've had the cone locknut strip its threads without me knowing it, I seemingly tightened the cones/nut, only to have it come out of adjustment very soon after, I couldn't figure it out at first, same can apply to the axle of course, damaged threads.
Broken axle isn't unknown either.

Just musing of course, hope you get it sorted.

 

[Ajax Bay]

I never really understood WHY compressing a QR squewer then compresses the hub bearings/cup/ cone. I mean exactly how does that happen ?, you're working with a steel axle, the cones are fixed on its length

Pushing lever over on the QR exerts a considerable force on both dropouts (and for Charley, a good tighten on the axle nuts does the same).
That force is exerted on the locknuts (both sides) which lock the cones in place on the axle, and fixed on that axle exert two equal but opposite forces inwards.
The axle resists that compressive force but the tiny reduction in length of the middle section of the axle (and therefore the distance between cones) as a result of that force gets the angular bearings 'mummy bear' when the wheel's inserted and secured in the dropouts.
So setting the hub up just a bit slack achieves that happy 'just right' bowl of porridge.
HTH
Yes - I've had a broken axle too. With a QR this is not immediately obvious, until one side falls out, and then it is.
Off topic, I kept one end of the broken axle and use it for making sure the hangers on my various bikes are vertical in both planes (eg after an off, or on a new-to-me bike). The thread is the same (not like this one ), and I use a spirit level to get it as good as I can.

 

[JonBuoy]

Pushing lever over on the QR exerts a considerable force on both dropouts (and for Charley, a good tighten on the axle nuts does the same).
That force is exerted on the locknuts (both sides) which lock the cones in place on the axle, and fixed on that axle exert two equal but opposite forces inwards.
The axle resists that compressive force but the tiny reduction in length of the middle section of the axle (and therefore the distance between cones) as a result of that force gets the angular bearings 'mummy bear' when the wheel's inserted and secured in the dropouts.
So setting the hub up just a bit slack achieves that happy 'just right' bowl of porridge.
HTH
Yes - I've had a broken axle too. With a QR this is not immediately obvious, until one side falls out, and then it is.
Off topic, I kept one end of the broken axle and use it for making sure the hangers on my various bikes are vertical in both planes (eg after an off, or on a new-to-me bike). The thread is the same (not like this one ), and I use a spirit level to get it as good as I can.

I agree that tightening a QR will compress the axle and reduce the clearance in the bearings but I'm not convinced about the bit that I have emboldened. Surely tightening up the nut on the end of the axle puts the dropout into compression and the (very short) bit of the axle passing through the dropout into tension. There are also frictional and torsional things going on that I will conveniently ignore. You do this to both axle nuts independently but that has no effect on the length of axle between the dropouts and hence no effect on the clearance in the bearings. Or have I misunderstood something?

 

[Ajax Bay]

I agree that tightening a QR will compress the axle and reduce the clearance in the bearings but I'm not convinced about [a good tighten on the axle nuts does the same]. Surely tightening up the nut on the end of the axle puts the dropout into compression and the (very short) bit of the axle passing through the dropout into tension. There are also frictional and torsional things going on that I will conveniently ignore. You do this to both axle nuts independently but that has no effect on the length of axle between the dropouts and hence no effect on the clearance in the bearings. Or have I misunderstood something?

Yes to all the first part. I have emboldened the sections which errs, imho. Consider the locknut inside each dropout:
a force is exerted on it/them by the dropout (equal and opposite (both inwards)). Since the locknut/cone combo is locked to the axle those two forces create axle compression. That compression reduces the length of the central section of the axle (rear 130mm long, front 100mm long) between the dropouts.

Each nut on the outside of the dropout needs to be tightened to exert sufficient force so that the securing force is 'enough' (only 'really' important on a disc braked front wheel). I suggest that wrenching the two axle nuts securely ensures that these two opposing forces are 'the same' as securing a hub with a QR.
The difference between nuts and a QR is that, in addition to the compressive force on the axle (nutted axle) and the consequential tiny (but important to angular wheel bearing adjustment) reduction in the length of the central section of the axle:
on a QR hub the (steel) QR skewer is under the same tension as the hollow (steel, area 64mm^2) axle is under compression so it will stretch. The skewer will stretch 2.3 times as much* (ratio of the cross sectional areas of the hollow axle v the skewer, Hooke's Law). But the force needed is the same. Hope that makes a modicum of sense.

* Actually slightly more as the 'stretched' steel skewer is about 12mm longer than the axle length compressed between the dropouts (thickness of a dropout x 2).
If you take the rule of thumb @presta uses of getting the cone finger tight and then backing off 100 degrees and locking that then we can guesstimate very roughly how much compression reduction in the axle length results: it's about 300 microns (pitch of an axle is 1mm).

 

[JonBuoy]

Yes to all the first part. I have emboldened the sections which errs, imho. Consider the locknut inside each dropout:
a force is exerted on it/them by the dropout (equal and opposite (both inwards)). Since the locknut/cone combo is locked to the axle those two forces create axle compression. That compression reduces the length of the central section of the axle (rear 130mm long, front 100mm long) between the dropouts.

Each nut on the outside of the dropout needs to be tightened to exert sufficient force so that the securing force is 'enough' (only 'really' important on a disc braked front wheel). I suggest that wrenching the two axle nuts securely ensures that these two opposing forces are 'the same' as securing a hub with a QR.
The difference between nuts and a QR is that, in addition to the compressive force on the axle (nutted axle) and the consequential tiny (but important to angular wheel bearing adjustment) reduction in the length of the central section of the axle:
on a QR hub the (steel) QR skewer is under the same tension as the hollow (steel, area 64mm^2) axle is under compression so it will stretch. The skewer will stretch 2.3 times as much* (ratio of the cross sectional areas of the hollow axle v the skewer). But the force needed is the same. Hope that makes a modicum of sense.

View attachment 670914
* Actually slightly more as the 'stretched' steel skewer is about 12mm longer than the axle length compressed between the dropouts (thickness of a dropout x 2).
If you take the rule of thumb @presta uses of getting the cone finger tight and then backing off 100 degrees and locking that then we can guesstimate very roughly how much compression reduction in the axle length results: it's about 300 microns (pitch of an axle is 1mm).

Yes to all the first part. I have emboldened the sections which errs, imho. Consider the locknut inside each dropout:
a force is exerted on it/them by the dropout (equal and opposite (both inwards)). Since the locknut/cone combo is locked to the axle those two forces create axle compression. That compression reduces the length of the central section of the axle (rear 130mm long, front 100mm long) between the dropouts.

Each nut on the outside of the dropout needs to be tightened to exert sufficient force so that the securing force is 'enough' (only 'really' important on a disc braked front wheel). I suggest that wrenching the two axle nuts securely ensures that these two opposing forces are 'the same' as securing a hub with a QR.
The difference between nuts and a QR is that, in addition to the compressive force on the axle (nutted axle) and the consequential tiny (but important to angular wheel bearing adjustment) reduction in the length of the central section of the axle:
on a QR hub the (steel) QR skewer is under the same tension as the hollow (steel, area 64mm^2) axle is under compression so it will stretch. The skewer will stretch 2.3 times as much* (ratio of the cross sectional areas of the hollow axle v the skewer). But the force needed is the same. Hope that makes a modicum of sense.

View attachment 670914
* Actually slightly more as the 'stretched' steel skewer is about 12mm longer than the axle length compressed between the dropouts (thickness of a dropout x 2).
If you take the rule of thumb @presta uses of getting the cone finger tight and then backing off 100 degrees and locking that then we can guesstimate very roughly how much compression reduction in the axle length results: it's about 300 microns (pitch of an axle is 1mm).

As a thought experiment...take a 130mm OLN nutted axle and fit it into a frame with 135mm between the dropouts. Tighten the left hand side first by holding the locknut on the inside of the left hand dropout with one spanner while you tighten down the left hand axle nut with another one. Note it may be physically difficult to get a spanner in there but as this is a thought experiment it doesn't matter! Hopefully you will agree that there is no tension/compression in the axle at this stage and hence no change in the bearing clearance. Now we move to the right hand side. Again we will hold the locknut on the inside of the dropout with a spanner while we tighten down the axle nut on the outside of the dropout to close up the 5mm gap. In reality we don't have to close up the full 5mm gap because the axle will stretch a little (probably a couple of microns dependent on the frame:axle stiffness ratio). The axle is now under (a tiny bit of) tension and the bearing clearances have increased. If this was a 125mm frame the opposite would have occurred and the axle would have been under compression. The key thing is that on a nutted axle the change in bearing clearance when fitting the wheel is controlled by the frame stiffness (and OLN discrepancy) and is generally negligible whereas on a QR wheel the change in bearing clearance is controlled by the QR tension and QR:axle stiffness ratio and needs to be considered.

All of course IMHO.

 

[Ajax Bay]

Your 130mm hubbed wheel would fall out of the 135mm dropouts, given the chance.
How can the central part of the axle be in tension yet sufficient friction be exerted to stop that? Friction from the outside of each dropout (by the axle nuts) is not enough to secure the wheel. Friction from the inside (the cone lock nut) is needed for security.
Think about that, while you try to get to sleep tonight, 'as a thought experiment'.