AJW2W11E

How can new bearings change a bike's performance so much? When I pulled the axle, the old ones fell on the floor. They still looked round, just dirty. After they were replaced. I stuffed the area with lube and carefully and incrementally adjusted them . Now the bike feels like its on a rail and rides faster. It's different bike. I get it, maybe the old ones were worn, but they were round . Was it the lube?

Andrew R Stewart 

Excessive hub bearing preload can be felt and if never experienced before can be hard to detect. Until one removes the wheels and starts to look deeper into their bike. I suspect most of what you feel is far better adjusted bearings. Sometimes when we do things to our bikes we also do stuff that we don't even know we did. Like when reinstalling wheels having previously slightly rubbing brakes and now don't. Or even topping off tire pressures before test riding after some work. Andy

Duragrouch

Old bearings could be spalled; Always clean and inspect with magnifying glass before reusing.

Bearing adjustment (cones) could have been too tight, harder to detect on a quick-release axle without using spacers or nuts to allow applying the quick release pressure (compressing the axle, making the bearings tighter) with the wheel off the bike, to spin the axle by hand; On the bike, the massively larger mechanical advantage of spinning the rim, you can't feel it. With a nutted axle, you can just spin the axle, with your fingers, no quick release to alter bearing fit.

Bearings needed to be lubed.

FBinNY 

Placebo effect.

Any bike whose wheels freespin a few revolutions off the ground, has negligible bearing drag in the scheme of things.

BITD we used to do downhill coasting races. We experimented with tire pressures, crappy bikes vs. top end ones, you name it.

The heavier rider tucked tighter always won.

We'd even switch bikes between heats, the same rider would win anyway.

The reality is that at anything over 15mph or so it's all about air drag and everything else becomes very secondary, and ever less meaningful as speeds get higher.

To be clear, I'm not saying these things don't matter, just that in the big picture, they matter less than people imagine.

Duragrouch

I agree, but would say, not drag, but FEEL can change. Spalled bearings or cone, pitted races, etc, can actually be felt as minute vibrations and slight noise (amplified by spokes and most especially by a composite-spoke wheel) on the bike, not really noticed until that is all corrected, and then suddenly it feels noticeably smoother and quieter.

Trakhak

In a first-hand account of a visit to the Tour de France that appeared in a mid-'70's Bicycling! magazine issue, the writer says he was allowed into the mechanics' areas one morning. He idly picked up the front wheel from Bernard Thevenet's Peugeot and spun it and was surprised to feel grinding from a too-tight bearing adjustment. He pointed it out to the mechanic, who said (or perhaps growled), "It does not slow him down."

cyccommute 

Perhaps but I find that most all OEM loose bearing hubs have bearings that are far too tight…to the point where the wheels won’t pendulum due to the weight of the tube valve. I suspect it may have something to do with machines building wheels and the need for more precise locating of the spoke hole that the machine needs. What AJW2W11E may be experiencing is something a simple as properly adjusted bearings.

FBinNY 

Ask yourself, "what does a valve weigh? Then compare that to the forces involved in propelling a bike. Even if you double it to account for the other valve, it's still negligible, and far less what anyone could feel when riding.

Trakhak

Yep. Try holding a front wheel by the axle ends and spinning it while pressing your thumbs against the bottom of the hub flange. Takes a surprising amount of friction for a perceptible reduction in the speed of the spinning. Yes, you can apply enough pressure to begin to slow the wheel, but even the worst-adjusted hub won't get anywhere near applying that kind of friction. Especially if the grease has dried out and thus is no longer in full contact with the balls.

Given the length of the effective lever arm represented by the distance from the tire surface to the bearing surfaces, plus the minimal diameters of the bearing surfaces, plus the forces involved: placebo effect. (Cars run better after they've been detailed.)

GamblerGORD53

Tight bearings make a HUGE difference. Could be 2 mph moving and 40% coasting distance. Paranoia about a hair of wiggle is nuts. Try racing with boat anchor roller brakes. LOL.
Tight pedals are awful as well, but it's mostly the seals now. 1/16" balls in some cheapos are nutso.
Balancing jumping wheels is not a totally useless activity either.
Pretty sure that ALL my new bike cones were tight and NO they will NOT loosen without wear.

cyccommute 

Come on, let’s be reasonable. The swinging of the wheel due to the tube valve is an easy indicator of how tight the bearings are. Tight bearings roll slower than properly adjusted ones and, depending on how tight the bearings are, that drag can be significant. Yes, spinning the wheel and timing the coast down would probably be a better measure but that takes time. Honestly, I can spin the bearings by hand and feel how tight they are. Some of these hubs are difficult to turn the axle by hand. That has an impact on the speed of the bike.

Andrew R Stewart 

This thread brings out an aspect of riding that I sometimes think about. "What is important to us as we ride?" Which sort of morphs to "Why do I ride a bike?"

For me much of the answers has to do with the feelings I get when riding, the sense of flying oddly mixed with the connection to the earth's undulation surface. The feelings of self reliance, most of the time in how I travel, being far away from the rest of the world whether inside my head or in lack of traffic. Way down the list of answers is an answer that has a number in it. No watts, heart rate, lactic acid count, and other so easily qualified/measured aspects that they are so abused by the selling public. Talk about air resistance or bearing grade/preload included. What i do place numerical importance on, though, are my yearly mileage and days ridden totals. But these have little to do with any specific bike or it's condition.

The reason why I take care of my bikes is to have a good feeling bike when i ride it. That the bike's functions work well, are smooth, consistent and reliable are enhancements to my pleasure and not to gain seconds over a 40K time trial.

I agree with Francis and those who argue that too tightly adjusted bearings have little to do with overall friction losses when riding (never done the math but I bet a couple of PSI difference in the tires is far more friction change then too much bearing preload creates.) But I also agree with the ability for some riders to feel the differences between well set up wheel bearings and wheels with overly preloaded bearings. Andy

70sSanO

In the last 10-15 years I will typically service new hubs to make sure they are set up properly from the get-go. It is not about speed it is about longevity. Decades ago I was less disciplined, and hubs were probably better adjusted; or so I tell myself.

A while back our daughter had an Amazon bike delivered to our house and she planned on coming over Friday evening after work to put it together. So Friday morning I decided to take a look at it and pulled it out the box. I told my self that it was a cheap bike for our 5 year old grandson and it doesn't have to be perfect. Just check it and only do what is necessary. The rear hub and and crank were pretty smooth but the front wheel would barely turn and the fork would hardly rotate. I knew our daughter would not have the patience. Needless-to-say, everything was smooth when she came by and picked it up. I left the bell off for her to put on.

John

FBinNY 

You're arguing the wrong thing. I never posited about how to determine bearing friction, and agree that the valve pendulum test is pretty sensitive. But you're arguing about inches and I'm talking about miles.

I stand by my original statement that bearing friction would have to be VERY obvious off the bike before it becomes detectable while riding. Your valve test, or the number of turns a wheel free spins off the ground are much more sensitive than necessary. To put in your terms, posit two beakers of concentrated nitric acid, add a few drops of water to one, then try to tell the difference by sticking your hands in and seeing which burns less.

Kimmo

This bit makes me think it had play before.

Duragrouch

I have found that most bikes come new with bearings that are too tight, most especially if a quick release axle. A proper adjustment there requires putting in spacers to apply the quick release while adjusting the bearings, and it's time consuming to get it exactly right. The hubs are manfactured by different companies than the wheelbuilder, and they arrive with balls installed and greased, and the cones locked, and I think they are just built into wheels, then installed on the bike, and it sold that way. My feeling is that a good bike shop should check that before sales, but sales margins are tight these days, especially with the competition of internet sales, so it doesn't happen. Even at a quality bike shop near me, highly rated for repairs, for a QR axle, they don't adjust the bearings by feel, they snug the cones down by hand to take up all slack, then back one cone off a set amount (technician indicated this is recommended amount by shimano), then torque down the locknut.

Even with an excellent adjustment, it will vary with different pressure on the QR. Thus, I now prefer nutted axles, as my perfect adjustment stays that way when installed.

Kimmo

You don't need perfect adjustment; you just need good adjustment.

There's quite a large window of acceptable preload - yes, the bearings should be getting slightly squashed; this is what ensures they're all sharing load - which is anywhere from snug to a bit tight.

It's easy to get in this window on a QR hub without faffing with spacers or whatever. You just go for the smallest amount of perceptible rattle, which disappears when the QR is tightened. As covered above, the friction from even way too much preload is negligible in the big picture; anything in the acceptable range is comparable to seal drag.

An easy way to set the preload is to tighten the locknuts onto the cones so that they're rather snug but not super tight and contacting the bearings with no rattle, and then use two cone wrenches to undo the cones against the locknuts slightly, until there's a tiny rattle. If you overshoot and get too much rattle, just tighten the locknuts against each other. This will often tighten the locknuts more, or sometimes they'll move with the cone; either is no big deal. Just don't try it if you happen to be looking at a hub with a groove along the spindle and tabbed washers between the cones and locknuts.

Duragrouch

I've found that proper preload drastically increase durability, meaning, a lot of riding daily and going years with no loosening of the bearing, a big difference from previously when I went for a tiny bit of slack in the bearing, based on dad's advice, which I later learned as an engineer, was wrong (and he was an engineer, but didn't learn this). The nut angular change for this is very small, because it's the axial difference between zero slack and a tiny preload, which is a change in pressure versus axial strain of the axle. It's small. I go until I just barely feel the notchiness of too tight, then back off until that disappears. Key is being able to spin the axle in fingers when checking this. You can easily get by without being that precise. But a great adjustment does improve durability. I have not a clue in terms of drag for racing. This is only for durability.

Aardwolf

The test I use is the number of times the valve pendulum changes direction - easy to count, and doesn't need a timer or counting on both hands.
Start with the heaviest point at the top and when the wheel stops and reverses direction that's 1.
My 4 rims all manage 4 or 5 on that test.

I'm definitely not saying that makes any difference I would notice while cycling, but it does confirm there's nothing wrong with the axles.

FBinNY 

I guess I had to learn alternate methods BITD, because my light rims were spliced and the insert weighed just about the same as the valve. So, I still watch the rate that a spinning wheel slows to a stop. No stopwatch or counting, just a "Looks OK" kind of thing.

Kimmo

I'm only concerned with durability too; as I mentioned, I consider drag to be a negligible consideration.

I imagine a small amount of preload is optimal for durability, with the optimal amount varying in proportion to the maximum load the hub is expected to regularly experience. However, I've found that even cheap cup and cone bearings can survive a surprisingly large amount of preload for a long time. It would seem likely that all the hubs I've seen destroyed by excess preload were extremely tight from the factory, which is not uncommon.

But I've come across unpitted cheap hubs which have a lot of mileage judging from brake track wear, rocking far more preload than I'd deem necessary for any normal application. Which is why I say the window is large.

Duragrouch

So... some here are arguing, that even if a pendulum test detects more bearing friction, that doesn't matter when riding, as that is small compared to the power when riding. But this does not consider that when the bike is being ridden, the bearing is under several orders of magnitude more radial load; Not only may the friction increase, but friction translates to heat. When a bearing starts to go bad, the deterioration accelerates, it feeds on itself. This is why, in so many fields, early detection of a bearing deteriorating is critical, and they've developed various high-tech tools to do so.

Andrew R Stewart 

My High Tech Tools In use in my hub bearing condition assessments are my fingers first then a couple of wrenches followed up with a cloth wipe off of the now exposed bearings and I apply my most high tech tool, my eyes.

The above is sort of tongue in cheek but is really the best way to judge bearing condition as fully as possible.

I think the main reason why in other industries bearing condition early on in the products life is motivated by the high cost of down time later. Not so much to limit further bearing damage. Andy

70sSanO

I anticipate that this thread will span dozens of pages as the pre-load debate ramps up.

While I don’t have as many years obsessing over the perfect bearing adjustment, I have noticed the impediment that comes with manufacturing tolerances.

There is a chasing the wind aspect in the inability to overcome those tolerances. I adjust cones to a barely perceptible play that goes away when mounted. But more often than not I can rotate the hub around the axle and the perception changes or s gone.

I initially thought the axle might not be straight. But then realized, nothing is straight as the ball bearings tolerances far exceed the hub tolerances; especially the freehub.

But many miles later they are definitely good enough.

John

Duragrouch

Oh I agree, absolutely. I think what I was trying to express is that bearing deterioration is not linear, once it starts to deteriorate, it accelerates, and like you said, down-time. Detecting early so can plan intervention, like on an idle shift, is a big saver. One down machine can hold up an entire production line. And also, on many machines, tearing down the machine to just inspect the bearing is as much work as just replacing the bearing (often done on a periodic basis), so detection tools are great where bearing access takes a lot of time, and, when it may feel fine, the machine spindle just spun by hand, but not under load.

Eyes are everything; Bikers should inspect the bearing balls with a magnifying glass. Unfortunately, on a caged unit bearing, that's not possible.

Oh, speaking of which, I read that with old-style bottom brackets, the best thing you can do to improve durability, is get rid of the caged bearings so you can fit in more bearing balls (all new, from the same lot). Takes a bit more skill assembling, filling the cups with grease to hold the loose balls in place, but not hard once you've done it once.