> I don't ever recall seeing a problem on old bikes that used much larger
> ball bearings!
Finally we get down to what counts! My understanding is that the load capacity increases linearly with the number of balls but geometrically with the diameter of the balls. So "every thing being equal" the bearing made of half the number of balls, but twice the ball diameter, will have twice the load capacity then the one that has twice the number of balls but each is half the diameter.
> 1. Correction, I suggested that files had a hardness of 55Rc. Well that
> probably applied to common hardware store files. That number came from the
> crude "Mohs scale of mineral hardness" based on the idea that a harder
> material will scratch a softer one.
>
> Modern jewelers files are rated at about 72RC which would indicate that
> they've probably been nitrided to get the hardness over the maximum
> 66Rc-67Rc for most tool steels.
>
> Even so, the scratch test should tell you something. It's very difficult
> to make any marks on the surface of a 58Rc - 62Rc part. If you can easily
> scratch the surface of a crown race in the bearing track area then it's
> below the optimum hardness.
>
>
> 2. Headset alignment. I recently prepped a 1985 Gitane Team frame (a real
> one). Everything on that frame was very accurate except the head tube
> faces (the fork crown was good). They had been faced at one point but I
> still had to remove a little material to get them parallel. Most frames
> I've worked on or prepped have needed the head tube and fork crown
> refaced.
>
> Refacing the fork crown and head tube faces should be done anytime a
> headset is removed or installed. Even if the frame had been well prepped,
> you have no idea what kind of hammer mechanic removed the previous
> headset, especially on parted out frames from eBay (the case with the
> Gitane Team frame).
>
> One other thing, I've frequently had to readjust a headset after the first
> or second ride as the components seem to settle in a little no matter how
> carefully they were pressed in.
>
>
> 3. Lights went off when I first heard Jobst describe his theory of headset
> brinelling. I don't want to get into a discussion of that at this time.
> But as I mentioned before about the parable of the blind men and the
> elephant, there are many causes for brinelling in a headset (and
> brinelling is a word).
>
> Back in the day working on entry level Bike Boom bikes, they frequently
> came in with loose headsets that were severely brinelled. I made what
> seemed like a logical conclusion that the indentations had been pounded in
> from repeated impact from the loose headset. These bikes had cheap case
> hardened races plus most of them were on French bikes with the difficult
> to adjust toothed washers.
>
> I don't think that there is one simple cause for the brinelling problems
> with Campy headsets. I have well over a dozen bikes with them including a
> 1965 Swiss Tigra that's still smooth as glass. Inconsistent hardness could
> be a major factor. I've never seen a fretting or brinelling problem in
> Campy hub cones.
>
>
> 4. Finally, headsets are thrust bearings. Like other rotary bearings they
> permit rotation between parts but they were designed to support high axial
> loads. Ball thrust bearings are meant for low thrust applications where
> there's little radial load. Because of the head tube and fork angle there
> is considerable radial thrust applied to the front and rear of the crown
> race on lower headset cup. There are extremely high pressures applied to a
> very small contact area.
>
> I don't ever recall seeing a problem on old bikes that used much larger
> ball bearings!
>
>
> Chas. Colerich
> Oakland, CA USA
>
>
Andrew R Stewart
Rochester, NY