[email protected] (Jon Isaacs) wrote:
(me):
> >It's 'cause those Mavic hubs have industrial standard sealed bearing cartridges, like every other
> >kind of high quality rotating machinery on earth besides bicycles.
>
> Watches use jewels.
I guess, technically speaking, that watches are "machinery", but like dial indicators, barometers,
and other jewelled mechanisms, they are not what I was talking about.
> Actually in many applications, cartridge bearings are not used. In fact, wheel bearings in cars
> and trucks often use pressed in races and separate seals. Much more like a bicycle wheel design
> than a throw away cartridge bearing.
No, it's much more like a three-piece cartridge, configurable by assorting inner race w/rollers,
outer race, and seal separately for different applications. It may be three pieces, but it's
completely modular, replaceable, and manufactured to _much_ finer tolerances than bicycle cup 'n
cone bearings.
> And of course, those sealed bearing have seal friction so they seem smooth but they can never have
> the free rolling "smoothness" of a Superbe Pro track hub or a number of other such hubs.
2RS rubber contact sealed bearings might have more seal drag than pista parts (which don't, after
all, have seals), but running drag will be less with the bearing that has tighter tolerances, better
race alignments, and smoother surfaces. The best that Suntour or Campy can supply can't hope to
compare to the best offered by industrial bearing houses.
Besides, if you want to compare apples to apples, then the cartridge to compare against is a ZZ
double metal shielded type. Those are free-running as can be.
> I also imagine trying to stuff a sealed bearing into a cassette hub might take up some valuable
> mm's that are needed to pack everything in there.
That's the problem of a cassette hub as envisioned by Shimano, not the problem of replaceable
bearings. I could design a "cassette" that had room only for sleeve bearings, but that would
be dumb too.
> The standard heavy duty bearing for rotating machinery is the Timken roller bearing which in my
> experience most often comes in at least two pieces and with separate seals.
It meets my criteria for total replaceability, standardization, and finer, more controlled
tolerances than those offered by loose ball assemblies.
> And of course, rotating equipment like internal combustion engines and turbochargers most often
> use sleeve type bearings.
Only when they're pressure lubed. And there are plenty of IC motors
(m.f. 2-strokes) that use ball or roller cartridges, unsealed of course.
My point is that industrially, any machine that works even remotely like a bicycle from a
torque/power/RPM standpoint has cartridge bearings.
> Throwaway sealed cartridge bearings have their place. They are cheap and are a plug in design. But
> they are far from standard, especially in high load areas or where a specific design can optimize
> the specific design needs.
Those applications might call for a full complement bearing cartridge, or a roller bearing
cartridge, or an angular contact bearing cartridge, or a tapered roller bearing set (which amounts
to a cartridge).
No decent machine designer would permanently affix bearing races to a machine that was not designed
to be disposed of instead of serviced. No decent machine designer would spec bearings that required
periodic preload adjustment when a maintenance-free alternative was just as cheap to manufacture.
And no decent machine designer would use unique dedicated bearing components when industrially
standard parts would do the job better. Yet this is just what the major parts mfgrs have given us
for our bicycles.
Why? I can only speculate that the cost of manufacturing, say, Shimano hubs would make most cyclists
laugh or cry if they knew how tiny it really was. There's no other good reason to be so intransigent
about adopting well-supported, economical industrial bearing cartridges for such a component. Unless
perhaps you count planned obsolescence, though there are other ways to accomplish that.
Look at it this way: would it be acceptable to configure parts in such a way that each different
gruppo required its own unique chain? Or that when that chain wore out, you likely had to get a new
wheelset? This is precisely the effect of hub-specific cones and non-replaceable cups in Shimano
hubs. From an engineering standpoint, it's an outrageous shortcoming. Yet it is a proven maketing
success. Buying new parts must be half the fun for Shimano users.
Chalo Colina "opt-out of Brand S"