October 17, 2015 3 Comments
Part 4 closes our ball bearing series with a visit into how these little mechanisms impact our wheel building.
Bearings make wheel building possible. However, their effect is exaggerated. What happens at the hub is magnified ten-fold at the rim (where we work). This is owed to wheel geometry and must be appreciated throughout the build process. Bearing play has a disproportionate effect on trueness but it’s there for an important reason.
Play is good
For anything to run freely there must be play, a bit of extra space between moving parts so they never suffer self-inflicted friction and wear. The loads of cycling are quite enough. Internal, avoidable loads are unacceptable. Play can be small if rolling parts are made and aligned with great precision. While grease prevents easy feel of this play, it is present none the less. If a hub is built too tight, without such play, wear will create it. Running in a tight hub makes needless wear that eventually creates the missing play.
A wheel builder wants a hub with no play so the rim is not subject to movements generated by bearings rather than spokes. Trouble is, 0.002” of hub play (less than the thickness of paper) creates 10X the rim play. Most riders consider 0.02” rim play excessive. If a rim jumps around on bearing looseness, truing becomes a moving target. In the past, most hubs could be tightened to excess (bad for riding) so they are free from play (good for building), then loosened for use. Today’s hubs are more often equipped with cartridge bearings that can’t be adjusted (magneto bearings as used in Odom hubs, an exception). Many riders perceive hubs that have no play to be superior but more often the absence of play is evidence of friction under load.
A builder can do little to fix a hub’s play/friction situation. We need to notice if untrueness seems to change from rotation to rotation. If the pattern of rim wobble is not exactly the same on each wheel rotation, you can be sure something in the bearings is responsible. You might worry you’ll be blamed for untrueness owed to bearing play but rarely will you. While truing, keep a finger on one side of the rim, applying a small side load to dampen bearing play. This way you can see actual wobble more clearly and true in spite of play.
It’s important you appreciate bearing play because it is a guarantee of smoother running and longer lasting mechanisms. All small cartridge bearings, made for modest (not engine) loads contain play. If handled individually, the play is easy to detect. Manufacturers specify the exact amount for each product. Hub bearings are used in pairs so play is nearly cancelled, but the cost is microscopic friction. Cartridge hubs that claim to be adjustable without using the historic cup-cone (as with magneto) approach, are simply side loading. Play disappears but only thanks to friction.
Play is unavoidable
Up to 0.008” of play (measured at the rim) is generally acceptable in a bicycle wheel. Such a small amount cannot contribute to handling trouble, generate high speed wobbles, or cause premature hub wear. To measure, put a dial indicator on the rim and push and pull using only ultra-low force, just a feather touch, or you’ll be watching the wheel flex. Good proof that bearing play is not a factor compared with regular flexing of the spoked structure.
Bearing play (and untrueness) impacts dishing (centering) measurement. A dishing tool applied to both sides of a wheel is very accurate. Small untrueness and bearing play will reverse dishing measurements if they are taken several times around the wheel. No worry, just don’t be a slave to dishing.
To work with modern hubs and their cartridge bearings, bear a couple of principles in mind:
(1) Spoke tension is a huge force on hubs. Most hubs expand with spoke tension so cartridge bearings no longer fit snug in the hub shell. The solution is to make hubs too tight for bearings but, after spoke tension expansion, the bearing is held with appropriate light force. This is a universal feature among hubs but hard to navigate. Who knows exactly how much spoke tension a hub will see? How can you compensate for an unknown load? As an industry, this problem is handled well. Just be aware that it underlies bearing play issues. No free running cartridge bearing would ever be made to endure the press fits of bike hubs. The expansion of hub shells with spoke tension is supposed to reduce friction-causing compression. The bearing press you use to service the hub would be a bearing engineer's nightmare if she didn't know the fit will become looser with spoke tension.
(2) Individual cartridge bearing hubs are built with play both in radial and angular movement of the axle. Since bearings are generally used in pairs in bicycles and electric motors, that movement tends to cancel out between pairs. Hubs that seem free of play actually have excess grease, hiding the play from detection, or friction .Thankfully, such low friction and premature wear are small factors and most hubs give good longevity in spite of it. In practice, more hubs die from contamination and corrosion than pure wear.
Going to extremes
(3) To have cartridge bearings run perfectly requires unreal attention to alignment. One Japanese hub maker has gone the distance on this topic: Gokiso. Take a look at their elaborate designs. Here is a hub designed by bearings! Not inexpensive but legendary. While 700g and $5000USD may seem high for hubs, these have a strong following. Some models are guaranteed for 29 years and you can watch wear tests of 100,000km at 300kph! Takeaway: all the hubs on the market are all compromises for economy and weight. We use them to take terrific rides but judge accordingly.
(4) Resist the temptation to spin your wheel fast in the truing stand. If you do, no worry, but realize it slows you down and makes trueness harder to isolate and address. Short, half turns are mostly what is needed. Adjust an area but don’t throw everything back and begin again from zero. I did this for years. You can become quite quick in spite of such a style but, in the end, it will limit your speed. Contact indicators will bounce, confusing the picture and your eye, while it perceives a pattern, is handicapped for devising corrections. Few can retain the 3D mental image of a full wheel with precision enough to act effectively. It's just a recess. Spin fast for relaxation but not to nail the target.
(5) Avoid building around a hub with blown bearings. Not only is the task made harder but once new bearings are installed, the wheel may show untrueness or asymmetry. The builder gets the blame or at least the unpaid task of retruing. We build wheels around bearings. Minute changes that arise from bearing replacement will cause effects at the rim. Of course, field conditions may give you no option.
Wheel builders, don't be overly critical about components like hubs. They're an easy target. Be a guru so the wheels you create balance cost and performance in your riders' best interests. That wisdom is too rare. Don’t let marketing define your work. Today’s great products will soon enough be obsolete. The “no free lunch” rule is in play: a perfect running hub is heavy and expensive. Be grateful for the hubs we have and work with the consequences!
November 02, 2021
Track hubs loosen as they are tightened typically as road hubs tighten considerably depending on quality and its most obvious on cartridge.
I thought it might be good to mention popping the seal and relubing cartridge bearings. Many a old sealed hub sitting in the rafters just needed a couple drops triflow to soften or grease thumbed in.
Great interest of mine. I have a little different opinion- but thats all it is. I found, with loose balls, running a little play cause the most wear. The reason…only the balls on bottom took loads mostly. The Tick Tick Tick of oiled Campy Track hubs on the Velodrome sure sounded fast, but after one night of racing without seals and oil and minimally discernable play it left burned races that turned brown. I realized oil isn’t the way to go because the stuff didn’t hang around for more than a few minutes.
Later I found white lithium with a few drops of triflow did about as good as any. Ive yet to feel a sealed bearing that can match the Campy cup and cone of the 70s and Dura Ace of mid 80s. They are also super vague on the hubs that have adjustable preload (old specialized, sansin, suntour.. I also found snugging up locknuts or very tight made a big difference in how much looseness developed with quick release tightness. Horizontal dropouts can make some stop turning.
Strangest hub of all award goes to Sansin in the form of Suntour Superbe Pro Track. It had half Cup and Cone and half Cartridge in one bearing, or so it seemed. They were strange with shallow angle race. Sure rolled well.
Also isn’t it better to have cartridge bearings with the Least difference between ID and OD? I read how Big bearings are best but big spaces equal a Lot of sliding and I think this is a factor, just as that cage sliding on the bearings to keep them in place so the thing doesn’t explode…which I have had happen at 40mph….under power.
Oh, would it be insane to run graphite powder only? I read about BMW using a dry transmission with some powdered lubricant…No pump…I don’t have the interest to sacrifice.
And thank you very much for the great articles and man its sad that so much of the fun side of bicycle racing is dying. I never forget when I saw a pro who couldn’t glue a tire….that pro was Lance.
Thanks. Zac
November 02, 2021
Enjoyed this post. I had a shimano dura ace track hub, new in the box, that was a little sticky. I thought it was odd at the time, but after building the wheel the bearings were no longer sticky (perfectly adjusted in fact) and I gained some respect for the folks that assemble those hubs at the factory!
Comments will be approved before showing up.
Tom Wray
November 02, 2021
Ric – I really enjoyed this series. Can’t wait to see what’s next.
Thanks for Wheel Fanatyk and for the two sets of Wheelsmith wheels I purchased 25 years ago and still use (from the Hamilton St. location).
P.S. $5k is a ridiculous amount to pay forn those hubs but if I had $5k I would buy those hubs.