Road Fan

I don't know how thick a case-hardened layer is. He said (I think) he's removing 0.01", or 10 thousandths of an inch, which equals ¼ of a millimeter. I'd love to think Campagnolo practiced through-hardening, but I just don't know. Calling Bulgie?

Mad Honk 

@Hilander,
Out of all of the bearing races I have polished or re-conditioned I have yet to see a failure of the race surface. I am probably correct in my thinking but the parts were through hardened at the factory as it would be a bit hard to just case harden the parts. The steel used for most of the axles was black steel which is what we currently use for impact sockets, and what most garage mechanics have in the tool box. Those did not hold up well for the freewheel hubs, bit I think it was the pressured exerted on the axle in a high stress area. That same black steel can be chrome plated as is done for a number of high cost wrenches in the toolbox including mine. HTH, Smiles, MH

repechage

‘when super record headsets arrived I was hopeful that the small races would end up through hardened as the parts were small, possibly resulting in longer service life.
‘unfortunately they paired that with a very petit crown cone race. Too delicate.

Road Fan

Back when I worked for Ford and was rebuilding my Masi, I brought in the headset to show my boss and talk about re-machining the races, a mechanical engineer by training and by passion, and an inveterate DiY. I showed him my dimpled HS races, and he did not think there was much he could do about it, between the two fo us. Two major issues: Setting up a precision lathe to cut or clean up races depended on centering the piece to be worked - to find its center axis. He did not see this as feasible.

Another issue was the hardened layer. We did not know if the parts are through hardened, and we did not know how thick or thin the layer might be if it was case hardened. All we actualy knew was that the as-manufactured nature of the headset parts was adequate for the "useful life," whatever that meant.

What I finally did was to use the department laboratory to clean and view the races, and we found the crown race was trashed, and the lower pressed cup has very light visible wear. I was able to buy a crown race in much better condition, and put the bike back together. It's been ridden a few hundred miles since then, still feeling right.

oneclick 

If the part has surfaces such that you can hold it in whatever machine tool you are using it should not be difficult.
You just put it in the chuck and use an indicator to get it right.
Ideally you hold it by the surface that is held by the frame (you are assuming that the framebuilder got those bits right).
However, what he really might have been saying was that after seeing it he knew it be MUCH more work to get the races as good as they were new, and yes, worthless to even try if they were not through hardened.

There is a reason bearing manufacturers have specialized grinding equipment.
The minimum home-machine-shop equivalent is probably one of the Quorn units, they're a few steps above a rubber drum in a hand-drill.

Kabuki12

I have only done bearing race clean up once . It was on a BB spindle that had been over tightened at some point in its life. After indicating it in and forming a wheel to the correct radius , I was able to clean up the surface with .002” on each side of the spindle. I then polished the surface lightly with very fine sand paper (800). The spindle is still going strong years later but I don’t know if I actually saved any money. I certainly didn’t make “shop rate” at my grind shop!

JohnDThompson 

Yeah, we had a lot of those split on the assembly line at Trek.

The Super Record crown race was made much thinner than the Record crown race to keep the overall stack height reasonable. The Super Record aluminum cups had to be much thicker than the steel cups of the Record unit, and would have resulted in a stack height of over 46mm, which would have severely limited the Super Record headset's aftermarket sales.

IMO, roller bearings make much more sense for the lower race of a headset.

jdawginsc 

I was telling Mad Honk that there were a hybrid set which had lower roller and upper ball bearings. Maybe YST or something.

JohnDThompson 

IRD "RollerDrive:"

https://www.interlocracing.com/shop/...lerdrive#attr=

Bad Lag

Is the jury still out on longevity? Centerless grinding into a thru hardened part sounds like a good solution to dimpling. Grinding into a case hardened part does not.

I've always thought headsets were trouble. No matter how carefully I adjust mine, they always end up dimpled.

When you sit on the bike, the bottom gets heavily loaded and the top unloads. If I start with a preload on both (to ensure both bear some load when seated) I increase the load on the bottom. As a result, I consider them a consumable, like tires.

repechage

Way back I thought the roller bearing headsets curious, as the rollers are of uniform diameter, if they were tapered, I would have considered them superior, but obviously way more money to mfg.
‘true, they don’t really rotate much.

repechage

and you had the tools.

JohnDThompson 

While the cylindrical rollers do result in some sliding contact along some parts of the roller, they are not only less expensive than tapered rollers, but also much less fussy about alignment than tapered rollers. As you point out, unlike hubs, bottom brackets, or pedals, they are not in constant rotation when in use, so the slight additional drag from the siding contact is not significant. And my experience is they are far more durable than balls in a parabolic race.

jdawginsc 

I thought I remembered right.

https://www.ebay.com/itm/37457150462...3Avlp_homepage

JohnDThompson 

Ok, but 1-1/8" excludes it from most C&V bikes.

jdawginsc 

Absolutely, was just just wondering if my memory was faulty!