sd5782 

Working on cones on another vintage wheel, so posting the question here rather than the bicycle mechanics sub forum. Nearly half of the vintage wheels I disassemble have pitting on the cones. It of course seems to always be localized to a certain area. Minor pitting is never just evenly scattered around the whole circumference of the cone. In general, there is grease there and not always dried hard.

So, does it form from wheels sitting without turning over a period of 5-10 years? Does moisture condense and sit and and then pit a small area? Do misadjusted bearings put a high load on an area? Do cones just basically have crappy metal? It can’t be grit as pits are never evenly spaced. Even seemingly well cared for bikes with seemingly well adjusted wheels fall prey to this.

I can’t tolerate even moderate and fine pits. Gladly, the cups are almost always fine. I know folks may live with it and maybe orient the slightly pitted area upwards or try to polish the pits out, but my fastidious mechanical mind won’t let me go there. Oh well, back to cones, spacers, washers, seals, locknuts, etc.

base2 

Any imperfection magnifies the load at the imperfection. The worse the imperfection the greater the multiplication. The imperfection then multiplies in severity.

The balls ought to wear out first. Round is a hard shape to make.
Then the cones are made of a slightly harder material. The idea is they are less disposable than the balls & harder to manufacture square up, thread, heat-treat, etc...
The races pressed into the hub are the hardest & most durable on account of they define the serviceability of the wheel assembly.

Pitting occurs because of a failure of the lubricant. The reasons are many, but being insufficient for the load however (properly or not) it is applied is the root cause.

oneclick 

No always, and especially for vintage parts that were not the highest quality to start with. Pits and flaws-which-would-quickly-lead to pits were not uncommon during the bike boom. I wonder if the majority of his hubs were Normandy/Atom.

sd5782 

No, not always Normandy at all, although they have a special place in cone quality. One recent example was a vintage Campagnolo high flange. I won’t go into the heliocomatic hub, because that is a different case with smaller bearings and it’s French too. I guess I shouldn’t complain too much as cones and spacers are easier to replace than cups.

Interestingly, many offenders upon closer examination after disassembly have bearings that turn copper colored even though there is still grease present. I assume that is a sign of overheating. I have been using a slightly thinner wheel bearing grease lately under the theory that it can migrate around more freely than the thicker “high temp” grease. New grade 25 bearings too of course.

base2 

I was speaking as a person that ran a fuel oil & lubricating oil quality management laboratory for a number of years professionally.

Starting with garbage certainly can't help. The lubricant failure may well be from the added stress from poor surface texture. A more extreme pressure rated lubricant & lighter loading may provide longer service before the eventual hockey stick rise in localized deviation from nominal.

What the lubricant can't make up for is the parts fitness for duty. Hardness & area to support the load with a minimum of deformation can certainly be a factor. Too few of too small balls that compress too much or bearing races that are too soft will always be a factor. This is a metal fatigue issue & poor engineering design issue. I've seen bearings balls & rollers that flake like a pastry.

But the root of the failure mode of pitting of races is the balls and the races physically touch such that the surface of each is deformed. Hence, pitting.

smd4

How does using ceramic balls affect this equation?

dddd

Cone quality is a huge variable, as some hubs are much more resistant to pitting than others.

Rust is a common source of the initiation of pitting.

Given that so many bikes even into the eighties had lower quality, un-sealed bearings, the biggest factor in the bearing's life will be to what degree that the bearings were over-tensioned at the factory, with perhaps no play whatsoever in the bearing's adjustment prior to the quick-release lever being closed.
Those traditional "internal cam" quick release levers applied many hundreds of pounds of compression along the bike's axles, making the bearing surfaces flex like an asphalt roadway with big rigs rolling over it all day! So only very good bearings will survive much of a service interval before said steel surfaces start flexing themselves into a micro version of a tile floor, with subsequent loss of surface metal.

So firstly one must test their combined bearing and QR lever adjustments by adding thick washers or loose dropouts to both sides of their axle, then flipping the QR lever closed and marveling at the unbelieveably rough rotation of the axle spun by hand.
This is how I was taught at the first shop that I worked at some 40+ years ago (and was also a sales aid to show off the turning smoothness of the better Fuji's cartridge-bearing hub axles, which for some reason always got properly adjusted at the factory).
I've mentioned it here before, that I automatically discount my "buy" price on any (typically Schwinn or Trek) bike from the 1980's having Atom/Maillard hubs, whose severely over-tensioned bearing cones typically started pitting during what I would call these bike's break-in miles.

One trick I use to keep wheels with pitted and hard-to-source cones in service is to firstly rebuild them and then to mark with a file the spot on the exposed surface of the cone the location of any significant pitting, thus allowing me to position the pitting defect top-side (opposite the more loaded side of the cone) while installing the wheel.

SirMike1983 

You are correct to replace pitted cones when possible. The situation only worsens if left unchecked, accelerates damage to the bearing balls, which may in turn eventually then damage the cup. It also is more work in that you need more frequent checks, ball replacement, and re-greasing.

Very shallow pitting can be polished out in some cases by turning the cone in a lathe (or even a drill chuck) and evenly polishing. This is not a substitute for a new cone, but it will prolong the life of the existing cone a bit if you need to do that. It will not work for very deep or extensive pitting.

Some cones certainly seem to last longer than others, a measure of quality and proper care.

sd5782 

@dddd, I have read many of your maintenance posts and hopefully learned from them. That would make sense that they were over tightened at the factory, or somewhere along the way. It sorta exonerates said mechanic at point of sale as no one can complain of a loose feeling wheel. An overtightened wheel with all its mass will still feel smooth with a casual spin test. That also can explain overheated looking bearings even though grease is present. For the most part, I just have to dispel rust pitting on anything without also having a solid rust chain and other steel parts.

base2 

Ah, good catch. I meant that the surfaces touch. The softer one will deform & will accumulate damage. How is this so? The balls have much more surface area & there are many more of them. Yes they are softer, but by the sheer amount of surface area they will degrade at a much lower rate than a finely localized pit that gets hit 8, 9, 10 times (or however many) per revolution.

Ceramic balls are the harder ball so they accumulate damage even slower than steel. So have an even longer degradation curve. They are no less dependent on the surface adhesion & intermediate layers resistance to shear/compression of the lubricant. But when failure starts, the cones & races will suffer the same or at an accelerated rate as the balls yield even less load than another ball of another type. In short, longer service life but when failure happens it will be more sudden & severe.

branko_76 

I bought several pair of NOS Normandy High Flange hubs from a bike shop that went out of business, every one of them grind when the axles are spun by hand. I suppose they were over-tightened at the factory and if used as-is, will form pits in the cones rapidly.

steelbikeguy

My understanding is that cheap hubs tend to pit quickly because they are ... cheap.
Of course, I've also helped with a friends new Trek 400 and was shocked at how tight the hub bearings were adjusted! Not sure why that was.. maybe like MKS pedals which are adjusted tight so they will "wear in" quickly? Seems ludicrous, and I always adjust my Sylvan pedals correctly when I get them.
The last cheap hub that I bought was a Sovos fixed gear hub. The cones wore out in a few thousand miles, and it wasn't really a surprise. Fortunately, a buddy gave me his old track wheel with a Campy Record track hub!

Still, a cone that is made of cheaper steel, is barely hardened, and barely ground to create a smooth bearing path is not going to last too long. I still recall when I transitioned from cheap Sunshine hubs to Campy Records! Just spinning the axle tells you that the Record hubs had some very high quality cups and cones.
A quick photo from an overhaul of my Olmo's Campy Record hubs....




A good cartridge bearing will have smoother surfaces for the bearings, but these are pretty good for old technology.

Steve in Peoria

branko_76 

I suspect that those with pitted cones had been hastily assembled at the factory and when used year after year without maintenance, they brake down and pitting occurs.

I say this because I have rebuilt a couple of dozen Normandy hubs and I'd say less than half have had pitted cones. These are the "cheap" High Flange Hubs commonly found attached to low end bikes with steel rims from the 1970's. For the most part, those bikes were not well maintained so if used regularly, of course bearing surfaces will start to wear.

I am currently using a pair of these hubs on my commuter, and checking before I type this, the front still runs smooth but the rear has a very slight thump. I don't recall when the last time was that I repacked them, but it's been several years or more. To satisfy my curiosity, I disassembled the rear. The Phil grease was still green, the races are smooth but both cones have what look to be factory defects, not pits. I may have overlooked this when I repacked them since at the time, I only had a limited supply of parts. Now that I am better stocked, I'll replace the two cones with fresh ones and check them in a couple of years.

steelbikeguy

That doesn't look too bad... definitely better than some of the cheap stuff I've seen and used.

At this point of the hobby, where you can't just go out and buy a new Normandy hub, we have to hope that the previous owners have cared for the hubs.. or maybe just ignored the bike after a couple hundred miles?
I've never owned Normandy hubs, instead starting out with some Sunshine hubs and then buying a pair of Shimano 600 (circa 1976) for $18. They were decent mid-level hubs, but still wore out the cones after a while.
I did pick up a pair of Campy Record low flange hubs in 1978, and I'm still using them! No need to replace the cones yet. All things considered, I'd say they were a better value than the Shimano 600 hubs.

Steve in Peoria

Chuckk

I remember back in the 80's the wisdom was to stay away from the super hard balls.
"Which would you rather have wear out, your cones and races, or the balls?"

Unrelated wisdom was from a shop-owner about a decade ago.
A lot of their new bikes came in with wheels set up so tight that they would not spin smoothly.
Wheel bearing readjustment was part of their new bike setup.

branko_76 

.

It's been a while since I opened any of my Campy hubs, those are things of beauty !

JohnDThompson 

Pitting is usually a result of lubrication failure, leading to fretting damage. Lack of adequate lubrication allows microscopic welds to form between the balls and the race, which are torn out as the bearing rotates. Localization to only part of the circumference of the bearing track is a result of misalignment between the bearing race and the rotational axis of the wheel, either because the axle is bent (fairly common on freewheel hubs, especially those supporting freewheels with seven or more sprockets), or dropout misalignment, or both. Cones exhibit this more than cups because the circumference of the bearing track is smaller on the cone than on the cup, which means the wear is more concentrated on the cone than on the cup.

1simplexnut

At one stage in my working life I had a lot to do with bearings and failure analysis .
I have seen bearings in large electric motors that have sat unused show corrosion and spalling from moisture sitting between the ball and raceway .
They will fail prematurely because of this .
The other example we saw BITD was the boat trailer wheel bearings that sat around outside for six months or so then got used without being checked .
Pretty normal for them to fail and show typical moisture related damage to raceways .

All good fun !

.

daka

I would hazard that the opposite may be true. Bearings like a little pre-load, more balls sharing the load, less pressure on individual balls and their point-contact on the race. Free play may allow as few as two balls at-a-time doing all the support. Clearance also allows the balls to act like little ball-peen hammers on the races with relative motion between cup and cone.

But what no-one seems to have mentioned, at least to this point in the thread, is spalling. In a part like a cone, which needs to have threads cut on the ID, you can't start with hardened material. The part is machined close to shape and then the surfaces are hardened afterwards. The final precise, shape of the race then has to be ground. However the hardening is more of a surface effect it doesn't necessarily go all the way through the cone. What happens then is a lot like the formation of a pothole. The surface material is hard and brittle, the material below is more ductile, just like asphalt paving is harder (especially in the cold of winter) than the road-bed below it. The more brittle material will develop cracks as the base material flexes under load .Eventually the cracks propagate to a point where a small piece can detach, exit the scene and, voila!, you have a pit (or a pothole).

In a less expensive hub the hardening process on the cones may be more abbreviated to limit cost and speed production thus hurrying the progression.

oneclick 

The purpose of pre-load is to attempt to get as many balls sharing the load as possible, on average as the bearing rotates.
Were the balls EXACTLY round and EXACTLY matching and the cup and cone ground EXACTLY to their curves and the adjustment EXACTLY correct, all the balls would be in contact all the time.
This does not happen.

If you adjust the bearing TOO tight you will have the situation where at some point the space between the cup and the cone, into which at least one ball is forced, is less than the diameter of that ball - and you get non-elastic deformation of the intersecting surfaces of the cone, cup, and ball.

Pre-load needs to be balanced between ensuring multiple ball contact and avoiding damage to the surfaces.
That is the advantage of finer manufacturing tolerances; the less the variation from true, the easier it is to get a high percentage of ball-contact while staying in the common elastic region.

As for cone-hardening - I suspect that pretty much all ground cones are through-hardened.
There'd be no point in only hardening their surfaces, you're going to grind them afterwards.
The sort of spalling you see will happen (and easily) on case-hardened cones.

sd5782 

Lots of varying opinions on this topic. Kinda what I suspected actually. I often give the axle a roll on a flat surface to confirm it is not bent. I really never have bent ones. I don’t have anything really to check dropout alignment except crudely eyeballing it it seems funny though that there always seems to be adequate grease in offending hubs. I do like the theory of bearings being adjusted too tight.

I guess this whole situation gives a nod toward cartridge bearings of which I have none. Those, it would seem could not really be adjusted too tight so as to put pressure on the bearing races. The races I would guess may have a larger inner diameter too. Lastly, those aren’t something we disassemble, so maybe out of sight, out of mind.

branko_76 

I have read some posts where "pre-load" was mentioned. I have also read that a small amount of play is advisable because the quick-release assembly, when locked down, will tighten up the slack.

daka

I think that is correct when thinking of, say, a Campagnolo hub which, out of the box, feels so silky smooth with no clearance. When put on the bike and the quick-release is closed that probably applies something close to the perfect amount of preload.

It is a different story with nutted axles and bottom brackets though, adjusting just a teeny bit tighter than what feels perfect may be a good idea.

branko_76 

I have some hub axle cones with uneven wear, do you suppose this is due to loose fitting cones?

daka

I think the photos in this thread support your point. Photos of the Campagnolo axle assemblies show a beautifully ground and shiny surface for the for the race. Other cones show a black race with a shiny stripe where the balls have worked the surface. I would argue that those are likely to be case-hardened parts that received no post-hardening grinding and, consequently, are not exactly the right shape to run smoothly through 360 degrees without interference.