Up until now, this blog has used the term "phase" without defining it.
I should have written this much earlier, but I've been putting it off and here we are.
By "phase" in this blog, we mean the angular positional relationship between
the spoke holes on the hub flanges and the nipple holes on the rim.
Whether the phase matches up perfectly is an extremely important issue.
You might think "does a wheel with mismatched phase even exist?" but
depending on how the hub and rim are paired or assembled, it can happen,
so I'll be writing about various scenarios.
Before I do.
Imagine a wheel where the hub flange diameter equals the rim's inner diameter.
Now imagine cutting that wheel perpendicular to the direction of travel at the valve hole position,
and flattening it out.
↑When you do that, you get a state like this.
This is the basic form for thinking about phase.
First, I want to write about something I get asked occasionally: the phase misalignment
that happens when you "build a 36H hub with an 18H rim."
This applies not only to the 36H hub / 18H rim case
but to any case where "the rim hole count is half the hub hole count."
First, let me build a 36H hub with a 36H rim using radial lacing.
I've drawn the spokes on the near side as red and those on the far side as blue from my perspective.
The spoke lines in radial lacing are perpendicular to the direction of travel.
There is no phase misalignment.
Next, the hub stays the same but the rim's hole count becomes half.
For now, I've simply removed half the rim holes.
I've corrected the rim's hole spacing to match the reduced number.
This is what it looks like.
Let me try radial lacing just the far flange first from my perspective.
Looking at just this side, there's no phase misalignment.
However, on the other flange's holes, there is no rim spoke hole
at the position that would correspond to a straight phase.
I have no choice but to place spokes knowing the phase is misaligned.
When I thread it through, this is what happens.
Let's say I actually build it like this. Of course, I use spokes of the same length on both sides.
When I tension the spokes to center the rim, it stabilizes like...
...this state.
↑With an actual 18H rim and 36H hub, it's not quite this pronounced,
but the spokes in radial lacing (though calling it "radial" is a bit misleading since they're not actually radiating)
twist slightly outward like a porcupine on the near side and inward like an anti-porcupine on the far side.
It's unsettling.
You could say it's like a JIS-laced wheel with only some spokes remaining.
In reality it's subtle, so you can ride it and it's not unrideable, but it doesn't feel right.
If I do tangential lacing with this, quite significant problems occur.
If we call the normal tangential-laced spoke with correct phase F1,
and the reverse-laced spoke F2, then the combined pulling direction of F1 and F2
becomes F3 in the diagram above.
F3 is also the path the spoke would take in radial lacing.
That F3 is twisted — this is the state of misaligned phase.
If you extend the twisted F3 back toward the hub, it doesn't pass through the hub axle.
When I add F1 and F2 corresponding to this twisted F3,
↑it becomes like this. If the F1 spoke and F2 spoke were the same length,
the F1 spoke would be too short to build, and the F2 spoke would be too long.
Two different spoke lengths would be needed on one flange alone.
On a rear wheel, this means four different lengths would be needed for one wheel.
With an 18H rim, each side has 9H (an odd number) so tangential lacing isn't possible,
but when building a 32H hub with a 16H rim, this kind of thing can happen.
Next, let me consider building an 18H hub with a 36H rim.
First, the basic form.
I've removed all the holes from one flange.
In a normal hub, the flange holes have a phase pattern of left→right→left→right... at equal intervals,
so I move every other spoke hole to the opposite flange.
This is the diagram for building with "a rim with double the hub's hole count."
Let me try radial lacing right away.
↑There is no phase misalignment. However, only one side of the rim's hole pattern is being used.
If it's a rim without hole offset, there's no problem, but
when building with "a rim with double the hub's hole count,"
the realistic cases are:
12H hub / 24H rim
16H hub / 32H rim
18H hub / 36H rim
A 14H hub / 28H rim is basically impossible for hand-built wheels.
A 12H hub does exist, like on the 7700 Dura-Ace, but
in practice it's fair to say the possibilities are limited to just two:
"16H hub / 32H rim" and "18H hub / 36H rim."
Most 32H and 36H rims have spoke hole offset, so building with
"a rim with double the hub's hole count" avoids phase misalignment,
but hole offset occurs, so it still can't be called a proper wheel.
I should have written this much earlier, but I've been putting it off and here we are.
By "phase" in this blog, we mean the angular positional relationship between
the spoke holes on the hub flanges and the nipple holes on the rim.
Whether the phase matches up perfectly is an extremely important issue.
You might think "does a wheel with mismatched phase even exist?" but
depending on how the hub and rim are paired or assembled, it can happen,
so I'll be writing about various scenarios.
Before I do.
Imagine a wheel where the hub flange diameter equals the rim's inner diameter.
Now imagine cutting that wheel perpendicular to the direction of travel at the valve hole position,
and flattening it out.
↑When you do that, you get a state like this.
This is the basic form for thinking about phase.
First, I want to write about something I get asked occasionally: the phase misalignment
that happens when you "build a 36H hub with an 18H rim."
This applies not only to the 36H hub / 18H rim case
but to any case where "the rim hole count is half the hub hole count."
First, let me build a 36H hub with a 36H rim using radial lacing.
I've drawn the spokes on the near side as red and those on the far side as blue from my perspective.
The spoke lines in radial lacing are perpendicular to the direction of travel.
There is no phase misalignment.
Next, the hub stays the same but the rim's hole count becomes half.
For now, I've simply removed half the rim holes.
I've corrected the rim's hole spacing to match the reduced number.
This is what it looks like.
Let me try radial lacing just the far flange first from my perspective.
Looking at just this side, there's no phase misalignment.
However, on the other flange's holes, there is no rim spoke hole
at the position that would correspond to a straight phase.
I have no choice but to place spokes knowing the phase is misaligned.
When I thread it through, this is what happens.
Let's say I actually build it like this. Of course, I use spokes of the same length on both sides.
When I tension the spokes to center the rim, it stabilizes like...
...this state.
↑With an actual 18H rim and 36H hub, it's not quite this pronounced,
but the spokes in radial lacing (though calling it "radial" is a bit misleading since they're not actually radiating)
twist slightly outward like a porcupine on the near side and inward like an anti-porcupine on the far side.
It's unsettling.
You could say it's like a JIS-laced wheel with only some spokes remaining.
In reality it's subtle, so you can ride it and it's not unrideable, but it doesn't feel right.
If I do tangential lacing with this, quite significant problems occur.
If we call the normal tangential-laced spoke with correct phase F1,
and the reverse-laced spoke F2, then the combined pulling direction of F1 and F2
becomes F3 in the diagram above.
F3 is also the path the spoke would take in radial lacing.
That F3 is twisted — this is the state of misaligned phase.
If you extend the twisted F3 back toward the hub, it doesn't pass through the hub axle.
When I add F1 and F2 corresponding to this twisted F3,
↑it becomes like this. If the F1 spoke and F2 spoke were the same length,
the F1 spoke would be too short to build, and the F2 spoke would be too long.
Two different spoke lengths would be needed on one flange alone.
On a rear wheel, this means four different lengths would be needed for one wheel.
With an 18H rim, each side has 9H (an odd number) so tangential lacing isn't possible,
but when building a 32H hub with a 16H rim, this kind of thing can happen.
Next, let me consider building an 18H hub with a 36H rim.
First, the basic form.
I've removed all the holes from one flange.
In a normal hub, the flange holes have a phase pattern of left→right→left→right... at equal intervals,
so I move every other spoke hole to the opposite flange.
This is the diagram for building with "a rim with double the hub's hole count."
Let me try radial lacing right away.
↑There is no phase misalignment. However, only one side of the rim's hole pattern is being used.
If it's a rim without hole offset, there's no problem, but
when building with "a rim with double the hub's hole count,"
the realistic cases are:
12H hub / 24H rim
16H hub / 32H rim
18H hub / 36H rim
A 14H hub / 28H rim is basically impossible for hand-built wheels.
A 12H hub does exist, like on the 7700 Dura-Ace, but
in practice it's fair to say the possibilities are limited to just two:
"16H hub / 32H rim" and "18H hub / 36H rim."
Most 32H and 36H rims have spoke hole offset, so building with
"a rim with double the hub's hole count" avoids phase misalignment,
but hole offset occurs, so it still can't be called a proper wheel.