Another wheel build today (et cetera).
But first.
GroTac pointed out to me that
measuring flange hole center-to-center with calipers
is a measurement method prone to error,
so going forward in this article
I'll call that "my method"
and I've decided to measure
flange hole diameter another way besides my method.
The calipers interfere with the freebody,
so I've disassembled the hub temporarily.
I didn't want to take apart a brand new hub though.
The nominal value for the
freeside flange hole diameter is 51mm.
I measured the flange holes
on opposite sides, outer-to-outer.
53.
4mm,
53.4mm total.
Next, I measure the flange holes inner-to-inner.
48.
6 or 7?
If it's 48.6mm,
(53.4 + 48.6) / 2 = 51.0, and
if it's 48.7mm, that's 51.05, so
51mm is fine.
Next, as the reverse of my method,
I preset the calipers to 51mm beforehand
and place one measuring point visually
where I think the center is.
This also measured 51mm by my method.
My method is supposedly prone to error,
(though I can't claim it has no error
in the 0.1mm range either),
but this time it matches the nominal value.
The non-freeside flange hole diameter nominal is
49mm, and the same applied there too.
For calculating spoke length for this rear wheel,
I'll use flange hole diameters of 51mm and 49mm,
but these are values I determined from my method,
not by referencing nominal specs.
There was one thing that bothered me
about what GroTac said.
The part about how even if flange diameter
differs by 0.5mm, there's no problem
building the wheel.
Well, that's technically true.
If the rim's precision is ideal.
As you'd often see with old ZIPP rims,
there are rims where individual variation
in rim inner diameter within a single rim
exceeds 1mm.
In the diagram above, I've drawn the red line
as the longest radius and the blue line as the shortest,
but I haven't drawn them to accurate scale.
I was pretty sloppy about that, just so you know.
If the rim's inner contour were a perfect ellipse,
the red and blue lines would be perpendicular,
but most often that's not how it works.
I've marked the center point where the rim's center
would be if it were a true circle,
but the line segments A and B that bisect the longest axis
and the line segments C and D that bisect the shortest axis
have no guarantee of being the same length.
My rim inner diameter measurement method
can't distinguish this.
If B, C, and D are nearly equal in length
while A is noticeably longer—to exaggerate,
like looking at an egg from the side—
spoke tension will vary, with higher tension
near the A phase.
When dealing with rim conditions like this,
that 0.5mm flange diameter difference
can actually become a non-negligible dimension.
Even before that,
5mm—ten times that 0.5mm—is used
by hub designers to eliminate
left-right spoke length differences
by changing flange diameter left and right.
For someone designing hubs that way
to dismiss 0.5mm seems questionable.
Regarding the left-right flange width interpretation
for butted spokes (three-cross),
the majority of hub makers use center-to-center of the flanges.
I think that's natural,
because presumably that's what you get
from presenting the hub design drawing directly.
But the information needed for hub design
doesn't have to be the same as the information
needed to calculate spoke length.
If there were some kind of "kindness"
from the wheel-building perspective,
I think it would be better to present
flange width as outer-to-outer.
When spoke length is calculated based on
flange center dimensions,
with unfavorable conditions stacking up,
you can end up with a length
where you think "I can't really sell this to a customer."
If you had the audacity to say 0.5mm flange difference is no problem,
that would be different.
So even though 99%+ of rim-brake radial front wheels
are outpoke radial builds, I build wheels using
spoke length calculated from outer-to-outer flange width—
which is, in a sense, the inpoke dimension.
There are other merits and reasons
for interpreting left-right flange width as outer-to-outer,
but I can't write about those—trade secrets.
Far as I can recall off the top of my head,
Shimano and Novatec are makers that use
the outer-to-outer interpretation, though I don't know
if that's based on the kindness I mentioned earlier.
Recently I heard from a customer who can build wheels themselves:
"I knew the spoke length was long, so as a workaround
I planned not to do the final crossing, but then I went ahead
and did it anyway, trying to use up spoke length—
but it was like pouring water on a hot stone."
They meant it as a funny story,
but hearing that, I had some deep thoughts.
I won't write about it in detail,
but the fact that I'm mentioning it here
means it relates to this flange width interpretation issue.
So, going forward,
I'm not going to change my approach of
interpreting left-right flange width as outer-to-outer
and measuring hub dimensions by my caliper method.
Please pay it no mind.
Another day, another wheel (et cetera).
I built the rear wheel for Nomu Lab Wheel #9.
GroTac X102 hub, 24H
Half CX Sprint Yonroku JIS build
With lacing (结线).
But first.
GroTac pointed out to me that
measuring flange hole center-to-center with calipers
is a measurement method prone to error,
so going forward in this article
I'll call that "my method"
and I've decided to measure
flange hole diameter another way besides my method.
The calipers interfere with the freebody,
so I've disassembled the hub temporarily.
I didn't want to take apart a brand new hub though.
The nominal value for the
freeside flange hole diameter is 51mm.
I measured the flange holes
on opposite sides, outer-to-outer.
53.
4mm,
53.4mm total.
Next, I measure the flange holes inner-to-inner.
48.
6 or 7?
If it's 48.6mm,
(53.4 + 48.6) / 2 = 51.0, and
if it's 48.7mm, that's 51.05, so
51mm is fine.
Next, as the reverse of my method,
I preset the calipers to 51mm beforehand
and place one measuring point visually
where I think the center is.
This also measured 51mm by my method.
My method is supposedly prone to error,
(though I can't claim it has no error
in the 0.1mm range either),
but this time it matches the nominal value.
The non-freeside flange hole diameter nominal is
49mm, and the same applied there too.
For calculating spoke length for this rear wheel,
I'll use flange hole diameters of 51mm and 49mm,
but these are values I determined from my method,
not by referencing nominal specs.
There was one thing that bothered me
about what GroTac said.
The part about how even if flange diameter
differs by 0.5mm, there's no problem
building the wheel.
Well, that's technically true.
If the rim's precision is ideal.
As you'd often see with old ZIPP rims,
there are rims where individual variation
in rim inner diameter within a single rim
exceeds 1mm.
In the diagram above, I've drawn the red line
as the longest radius and the blue line as the shortest,
but I haven't drawn them to accurate scale.
I was pretty sloppy about that, just so you know.
If the rim's inner contour were a perfect ellipse,
the red and blue lines would be perpendicular,
but most often that's not how it works.
I've marked the center point where the rim's center
would be if it were a true circle,
but the line segments A and B that bisect the longest axis
and the line segments C and D that bisect the shortest axis
have no guarantee of being the same length.
My rim inner diameter measurement method
can't distinguish this.
If B, C, and D are nearly equal in length
while A is noticeably longer—to exaggerate,
like looking at an egg from the side—
spoke tension will vary, with higher tension
near the A phase.
When dealing with rim conditions like this,
that 0.5mm flange diameter difference
can actually become a non-negligible dimension.
Even before that,
5mm—ten times that 0.5mm—is used
by hub designers to eliminate
left-right spoke length differences
by changing flange diameter left and right.
For someone designing hubs that way
to dismiss 0.5mm seems questionable.
Regarding the left-right flange width interpretation
for butted spokes (three-cross),
the majority of hub makers use center-to-center of the flanges.
I think that's natural,
because presumably that's what you get
from presenting the hub design drawing directly.
But the information needed for hub design
doesn't have to be the same as the information
needed to calculate spoke length.
If there were some kind of "kindness"
from the wheel-building perspective,
I think it would be better to present
flange width as outer-to-outer.
When spoke length is calculated based on
flange center dimensions,
with unfavorable conditions stacking up,
you can end up with a length
where you think "I can't really sell this to a customer."
that would be different.
So even though 99%+ of rim-brake radial front wheels
are outpoke radial builds, I build wheels using
spoke length calculated from outer-to-outer flange width—
which is, in a sense, the inpoke dimension.
There are other merits and reasons
for interpreting left-right flange width as outer-to-outer,
but I can't write about those—trade secrets.
Far as I can recall off the top of my head,
Shimano and Novatec are makers that use
the outer-to-outer interpretation, though I don't know
if that's based on the kindness I mentioned earlier.
Recently I heard from a customer who can build wheels themselves:
"I knew the spoke length was long, so as a workaround
I planned not to do the final crossing, but then I went ahead
and did it anyway, trying to use up spoke length—
but it was like pouring water on a hot stone."
They meant it as a funny story,
but hearing that, I had some deep thoughts.
I won't write about it in detail,
but the fact that I'm mentioning it here
means it relates to this flange width interpretation issue.
So, going forward,
I'm not going to change my approach of
interpreting left-right flange width as outer-to-outer
and measuring hub dimensions by my caliper method.
Please pay it no mind.
Another day, another wheel (et cetera).
I built the rear wheel for Nomu Lab Wheel #9.
GroTac X102 hub, 24H
Half CX Sprint Yonroku JIS build
With lacing (结线).