dedhed

Looks like a lot of hand assembly here.

SquidPuppet

Mounted rim tape, tube, tire, inflated, done................ 27 seconds.

WizardOfBoz

SKF, no slouches when it comes to bearings, has a page on bearing retainers and talks about the bearing minimizing axial runout:

Enables several benefits

By significantly reducing any axial clearance between the bearing and gearbox housing, the bearing retainer helps optimize overall gearbox efficiency and NVH (noise, vibration and harshness). Full benefits include:

Optimal stiffness in the bearing arrangement
Enables gearbox length reductions
Reduced system costs
Improved operating efficiency
Reduced noise and vibration
More compact, lightweight designs
Optimized designs before prototyping
Reduced development time
Simplified assembly

SMB had comments on retainers

Retainers keep the balls evenly spaced around the raceway preventing ball to ball contact and allowing higher speeds. They also help to retain grease around the balls and raceways.

Full Complement (F/B)

A full complement (or full ball) bearing contains extra balls and has no retainer. It is used for its greater radial load capacity although axial load capacity is very small due to the risk of the balls fouling the filling slot during rotation. These bearings can only be used at low speeds and bearing torque is increased due to ball to ball friction. An exception is a hybrid full complement bearing (ceramic balls) which can be used for higher speeds due to the lower friction coefficient of the ball material. Improved steel and hardening techniques have increased the load capacities of bearings with cages and the full complement bearing is much less common now.

• Higher radial load capacity
• Low speed only (except with ceramic balls)
• Low axial load
• Increased bearing torque

A pretty thorough article about cage friction shows that a full ball with cage has more friction than half the number of balls with a cage. That is, more balls = more friction. But the article states that the retainer adds friction as well. Would have been cool to see the data for no retainer.

Anyway, it seems to me that a retainer does more than make assembly easier and reduce costs. It ensure centration, facilitates lubrication, and reduces one source (ball-to-ball) of friction while adding others (ball-to-retainer, guide-to-retainer). And more balls do increase load bearing, but (unless you have ceramic balls) increases friction.

Lakerat

I missed hub bearing assembly, is it on there?

spedrunr

Wow, I'd have to say that this is the most scientific (sounding?) forum to date. And the sentence structure/grammar is to be commended.

I still don't think we've come to a final conclusion(?)

I'm in the middle of rebuilding my bike and would rather not have to tear it down and put loose bearings if the benefit hugely outweighs the caged bearings.

I'm anxious to ride it so I guess I'll keep the cages and go back to loose bearings if the data shows that there is a significant benefit.

Thanks to all of you and your comprehensive research.

Metaluna

I'd say when in doubt, leave it the way it came from the manufacturer. Some of them do actually know what they're doing, and it's not like there's carnage all over the roads from peoples' retained-bearing headsets and wheels exploding. The difference, if any, will be small, especially if you keep up with maintenance.

spedrunr

Not to mention the extra time spent on acquiring the extra bearings, taking out the old ones, the extra time it takes to load up the grease making sure that the bearings don't fall all over the floor when you install the spindle etc.

Again the question to be answered, to what benefit (reduction in the friction coefficient?). I laugh because we try and reduce friction and weight at the cost of making our work out easier. I could see if I were trying to break a world record or win race. Most of my riding is for the benefit of health or picking up small items from the store.

AnkleWork

Another video shows loose bearing hub assembly. They actually pour the balls into place; takes about one second. Bearing cages can't speed that up no matter how many deified sources say so.

Also note in the above video from 2:50 the mounting of a skinny tire takes less than ten seconds.

Further evidence that most of the hand-wringing here is just silly emotion.

spedrunr

15 second chain installation. Hire these people for TDF wrenchers

79pmooney

A couple of thoughts. Headsets - the only issue that matters is axial loading. More balls rule. Hubs - funny how that one of the world's leaders in components, including very high end components used to professional athletes uses loose balls for their best hubs. Are they stupid? Locked in a tradition that just won't die? Or maybe they have access to testing and knowledge we don't.

Ben

rydabent

Both the headset and the BB put high loads on the bearings and races. So the best thing to do is to use a heavy bearing grease. The heavy grease lets you stick the loose balls on the race for easy assy. And yes with the loose ball allow for a little extra space, but more balls will spread the load better.

dedhed

I find the most time is spent going to the fridge for another beer.

Bill in VA

Unfortunately in industry today, it is not the engineers who make the final decisions. The engineer may have a certain design, but bean-counters and marketing tell them to 'control' costs.

spedrunr

frogbiscuit

I will throw my $.02 in. As one who worked in manufacturing (steel, aluminum, plastics, composites) for 35 years, in the grand scheme of things, bicycle bearings move very slowly. The argument for or against retainers is moot in a bearing moving that slow, that why you can get away with greasing it every few years or so. Most high speed bearings used have the inner and outer races attached, so the bearing is thrown in as a unit.
If you drilled a hole in the frame and installed a grease fitting, you would never have to replace the bearing, but that brings up other issues.

And....as an industrial engineer, the difference between taking a few minutes to install the balls and a few seconds dropping in the retainer are huge. That kind of lost time will kill an assembly line.

Mad Honk 

Wiz,
My experience with the ceramic bearings is that no lubrication is necessary for bicycle applications, as well as in outer space. I know a few folks who have replaced pedal bearings with free range ceramic bearings and are quite happy with the no lube part of the equation. I am tempted to try a full bike with free range ceramics but am bewildered by the almost $600 cost of the bearings. Help a stoopid brother out here... Smiles, MH

Brian_Kawasaki

Yes. I think proven because retainers don't normally wear out. If something goes wrong the retainer made from softer metal or plastic is destroyed in no time. But even plastic retainers can be in good shape in older bikes with lots of miles. There has to be almost no friction between balls (and at a very small contact point) when no retainer is used. But also the reason a retainer isn't a complete deal killer—very little friction between balls and the retainer. The whole context of the NASA article is high speed industrial and jet aircraft engine bearings working under heavy load. Plastic isn't one of their choices for retainer material. Sadly as I sit here I'm realizing Friday nights are not what they used to be, lol.

sweeks

As there's (supposed to be) a little extra room in the circle of balls, and the loading of the balls is at right angles to a circular line connecting the potential contact points, I like to imagine that the balls are spaced out and rarely, if ever, touch each other.

Hahaha... you nailed it. Keep the faith; more normal times are coming. Stay safe!

JohnDThompson 

IMO, rollers make more sense than balls, for headsets. And it's only the lower race that carries load. The upper race only provides alignment.

rydabent

When I was 13 I re-greased my ashtabula crank with loose balls. At that time I used heavy thick grease, and continue to use for BB and head sets.

rydabent

I totally agree. Get rid of the cage and put in as many more balls as you can. BBs and especially head sets take a lot of abuse. The more balls to spread the stress over, the better off you are.

cjenrick

you would probably need a very expensive test system to gauge the difference between loose vs caged bearings, the load resistance to the chain would have to be very precise as well as the friction sensor on the bottom bracket.

not all bearing cages are created equal. this Suntour Superbe 68mm Bb cage is not stainless, not that a regular steel cage is gonna rust with grease all over it, unless of course you ride in the rain or lots of miles.

the bearings fell out of the cage in the pic as soon as it was removed, the flanges wee bent up pretty good,

we installed a Sugino non chrome molly BB in place of the Superbe, it had no cages and seems to roll pretty good.

wonder if we could mage a grease guard spindle out of the Sugino since it would be abled to be drilled easier than cro-mo?

OT: does anybody know how to waterproof leather toe strap? we are gonna ty sno seal first.

Geepig

When I am rebuilding Velosteel coaster brakes I generally replace the sprocket caged bearing with just balls. For most of the damaged coaster brakes I have seen the problem starts with the collapse of that bearing, mostly due to a lack (a total lack) of maintenance. So I figure the more support I can give the sprocket, the longer before the brake breaks.

cbrstar

The problem with caged bearings is often they are not packed with grease properly. Take a look at a cartridge bearing, they are basically a caged bearing but they are sealed and packed with grease properly.

Geepig

Yes, very true!

From my experience of cartridge or other sealed bearings the cage is either like a strip of metal on either side of the balls with some kind of riveting between the balls or like a solid brass/plastic cage in which the balls sit. The caged bearings we classically use tend to have open structures that hold most of the grease in suspension between the balls, and which only collect the grime and surface losses from the bearings. Loose or caged, there is very little to control where the grease goes or how much is present. The cages also wear, especially when there is a lack of maintenance, a process which continues until the cage self destructs and the pieces can then find their way to places where chunks of metal are unwanted.