Another day with wheels (and so on).
Continuing from yesterday.
I'm rebuilding the rear wheel on my Gokiso.
The original rim is made by Equinox,
but it's not an off-the-shelf rim with a sticker slapped on—
it's a custom-ordered item with specified requirements.
It's built on a Sapim 15-spoke base in black aero SB,
forty-spoke pattern with the freewheel side tensioned super tight.
The rim uses internal nipples,
but you probably couldn't achieve the same tension with generic nipples.
Gokiso apparently chose narrow flanges to equalize spoke tension on both sides,
yet paradoxically they use radial lacing on the non-freewheel side,
which creates the worst possible spoke tension imbalance—
it's genuinely baffling.
While the freewheel side is under high tension with the custom rim
at levels I wouldn't normally use,
the non-freewheel side is pretty slack—not completely loose,
but noticeably softer.
The centering is spot-on though,
so within the limits of this spoke gauge and lacing pattern,
this represents maximum effort.
In other words, you can't do better than this—
this is the crystallized collective knowledge of Gokiso (laugh).
When I received it, the sprocket was still installed,
so in the image above I've temporarily removed
the cup-shaped dropout on the right side.
With it attached, the lockring tool for the sprocket
won't engage with the splines.
As for the freebody, there are nine splines total—
eight wide ones and one narrow one—and
two of them, positioned roughly opposite to each other,
are replaced with a stainless steel (probably) anti-bite plate
to prevent sprocket binding.
It's a bit hard to see, but they're at both ends in the image above.
Yet despite this, binding did occur,
and while the low-end spider arm cluster came out smoothly,
the section ahead of it with sprockets and spacers alternating
initially wouldn't pull out easily.
Looking carefully at this image and the previous one,
the bite marks are actually more pronounced on the anti-bite plate itself.
What's going on here?
Actually, maybe I'm mistaken about what that plate is—
perhaps it's deliberately designed to bind
and dampen sprocket rocking?
I must not doubt Gokiso's deep strategic thinking.
Cartridge bearing hubs normally have the bearing outer race pressed into the hub shell
while the bearing inner race slides loosely onto the hub axle,
but Gokiso reverses this.
Their hub has significant clearance between the shell and bearing outer race—
when you insert the bearing end into the hub and push with your finger,
it slides in smoothly and steadily,
like inserting the lid of a tea canister or diploma tube.
Meanwhile, the hub axle and bearing inner race are press-fit together,
so on a front hub, the axle and both bearings become
one barbell-like assembly.
I think this is a design consideration to prevent bearing rotation from being compromised
under axle stress,
but because the bearing inner race is press-fit to the axle
while the bearing outer race is loose against the shell,
with extreme long-term use the hub shell where the bearing sits
will wear.
This applies to aluminum hub shells against stainless outer races.
The barbell-like bearing-mounted axle develops noticeable play relative to the hub shell,
and lateral play that can't be eliminated by purely side-to-side adjustment
develops—(even though the quick-release is properly tightened,
precisely because it is properly tight, the play that occurs beyond that
becomes more apparent).
I've seen two examples where this was sent to the manufacturer's bearing replacement service
but the issue wasn't resolved.
I suspect this only happens with extreme long-term use.
In the examples I've seen, it wasn't that the bearings rusted
and the rust transferred and wore the hub shell—
both cases had perfectly clean hub shell seating surfaces.
On a somewhat different scale,
Shimano's Hollowtech II crank axles have
stainless shafts on higher-end models and aluminum on lower-end ones.
Both have a 24mm shaft outer diameter,
and while the BB bearing inner diameter is 25mm,
they use a nylon shim to avoid metal-to-metal contact
and achieve contact at 24mm.
Campagnolo's Ultra-Torque uses a 25mm steel shaft
(titanium on Super Record) with 25mm inner diameter bearings
in direct metal-to-metal contact.
The BB30 standard derives its name from the 30mm crankshaft diameter,
with direct metal-to-metal contact between the 30mm inner diameter bearing
and the aluminum crankshaft.
Aluminum losing in that pairing is what happened here.
So I don't think an aluminum hub shell against
a bearing without press-fit contact wears absolutely zero.
Gokiso originally didn't seem interested in or concerned with making light hubs,
but the fact that they offer a titanium hub model that's heavier than aluminum
when the shell volume is identical suggests
it's a specification born from reluctant awareness—
even if unannounced and never recalled—
of hub shell wear cases.
That's what I privately suspect anyway.
Built.
Gokiso Climber Hub 24H
Black half-comp forty-spoke lacing with wire pattern.
This is something I can't quite articulate well,
but intuitively it seems that when pedaling smoothly at low intensity,
the freewheel-side spoke stiffness is what matters for the ride feel,
but when pushing hard with aggressive high-intensity pedaling and standing,
how closely the non-freewheel-side spoke deflection matches the freewheel side becomes critical.
I think that's why equal-spoke-count radial lacing on the non-freewheel side
produces that sensation of poor engagement or sluggishness
at moderate-to-high pedaling intensity.
It has no practical issues whatsoever,
but the vertical span of the brake track processing varies slightly depending on the phase,
which made it look like radial run-out while I was building the wheel—
that was annoying to deal with.
Continuing from yesterday.
I'm rebuilding the rear wheel on my Gokiso.
The original rim is made by Equinox,
but it's not an off-the-shelf rim with a sticker slapped on—
it's a custom-ordered item with specified requirements.
It's built on a Sapim 15-spoke base in black aero SB,
forty-spoke pattern with the freewheel side tensioned super tight.
The rim uses internal nipples,
but you probably couldn't achieve the same tension with generic nipples.
Gokiso apparently chose narrow flanges to equalize spoke tension on both sides,
yet paradoxically they use radial lacing on the non-freewheel side,
which creates the worst possible spoke tension imbalance—
it's genuinely baffling.
While the freewheel side is under high tension with the custom rim
at levels I wouldn't normally use,
the non-freewheel side is pretty slack—not completely loose,
but noticeably softer.
The centering is spot-on though,
so within the limits of this spoke gauge and lacing pattern,
this represents maximum effort.
In other words, you can't do better than this—
this is the crystallized collective knowledge of Gokiso (laugh).
When I received it, the sprocket was still installed,
so in the image above I've temporarily removed
the cup-shaped dropout on the right side.
With it attached, the lockring tool for the sprocket
won't engage with the splines.
As for the freebody, there are nine splines total—
eight wide ones and one narrow one—and
two of them, positioned roughly opposite to each other,
are replaced with a stainless steel (probably) anti-bite plate
to prevent sprocket binding.
It's a bit hard to see, but they're at both ends in the image above.
Yet despite this, binding did occur,
and while the low-end spider arm cluster came out smoothly,
the section ahead of it with sprockets and spacers alternating
initially wouldn't pull out easily.
Looking carefully at this image and the previous one,
the bite marks are actually more pronounced on the anti-bite plate itself.
What's going on here?
Actually, maybe I'm mistaken about what that plate is—
perhaps it's deliberately designed to bind
and dampen sprocket rocking?
I must not doubt Gokiso's deep strategic thinking.
Cartridge bearing hubs normally have the bearing outer race pressed into the hub shell
while the bearing inner race slides loosely onto the hub axle,
but Gokiso reverses this.
Their hub has significant clearance between the shell and bearing outer race—
when you insert the bearing end into the hub and push with your finger,
it slides in smoothly and steadily,
like inserting the lid of a tea canister or diploma tube.
Meanwhile, the hub axle and bearing inner race are press-fit together,
so on a front hub, the axle and both bearings become
one barbell-like assembly.
I think this is a design consideration to prevent bearing rotation from being compromised
under axle stress,
but because the bearing inner race is press-fit to the axle
while the bearing outer race is loose against the shell,
with extreme long-term use the hub shell where the bearing sits
will wear.
This applies to aluminum hub shells against stainless outer races.
The barbell-like bearing-mounted axle develops noticeable play relative to the hub shell,
and lateral play that can't be eliminated by purely side-to-side adjustment
develops—(even though the quick-release is properly tightened,
precisely because it is properly tight, the play that occurs beyond that
becomes more apparent).
I've seen two examples where this was sent to the manufacturer's bearing replacement service
but the issue wasn't resolved.
I suspect this only happens with extreme long-term use.
In the examples I've seen, it wasn't that the bearings rusted
and the rust transferred and wore the hub shell—
both cases had perfectly clean hub shell seating surfaces.
On a somewhat different scale,
Shimano's Hollowtech II crank axles have
stainless shafts on higher-end models and aluminum on lower-end ones.
Both have a 24mm shaft outer diameter,
and while the BB bearing inner diameter is 25mm,
they use a nylon shim to avoid metal-to-metal contact
and achieve contact at 24mm.
Campagnolo's Ultra-Torque uses a 25mm steel shaft
(titanium on Super Record) with 25mm inner diameter bearings
in direct metal-to-metal contact.
The BB30 standard derives its name from the 30mm crankshaft diameter,
with direct metal-to-metal contact between the 30mm inner diameter bearing
and the aluminum crankshaft.
Aluminum losing in that pairing is what happened here.
So I don't think an aluminum hub shell against
a bearing without press-fit contact wears absolutely zero.
Gokiso originally didn't seem interested in or concerned with making light hubs,
but the fact that they offer a titanium hub model that's heavier than aluminum
when the shell volume is identical suggests
it's a specification born from reluctant awareness—
even if unannounced and never recalled—
of hub shell wear cases.
That's what I privately suspect anyway.
Built.
Gokiso Climber Hub 24H
Black half-comp forty-spoke lacing with wire pattern.
This is something I can't quite articulate well,
but intuitively it seems that when pedaling smoothly at low intensity,
the freewheel-side spoke stiffness is what matters for the ride feel,
but when pushing hard with aggressive high-intensity pedaling and standing,
how closely the non-freewheel-side spoke deflection matches the freewheel side becomes critical.
I think that's why equal-spoke-count radial lacing on the non-freewheel side
produces that sensation of poor engagement or sluggishness
at moderate-to-high pedaling intensity.
It has no practical issues whatsoever,
but the vertical span of the brake track processing varies slightly depending on the phase,
which made it look like radial run-out while I was building the wheel—
that was annoying to deal with.