I received Nomu Lab Wheel #1's rear wheel from a customer.
It's a rear wheel built with Molanbon lacing.
Originally it was built normally,
but the customer requested it be rebuilt.
The current lacing pattern has been like this since August 2017 (→here),
so it's been over 5 years.
This is apparently the wheel the customer uses most for practice and such.
It's a Rolf (Leaf hub) 24H Campagnolo/Compey 46-hole Molanbon lacing.
It's laced with different diameters on left and right.
The customer brought it in for freewheel body bearing replacement, but
separately from that,
the twisted section where the spokes thread is slipping progressively,
and even after truing it goes out of true quickly.
↑This area
The thing is, at the thin diameter portion where it's butted,
plastic deformation (which I call "unyo-n") is occurring.
Normally, at 1.8mm diameter (15-gauge) there's almost no unyo-n,
but with Molanbon lacing, the yielding forces concentrate on the outer circumference
beyond where the rim-side threading ends,
and unyo-n can occur even with Campagnolo spokes.
With Revolutions where the butted section is 1.5mm diameter (17-gauge),
unyo-n starts occurring probabilistically around 100 kgf
even on ordinary wheels,
but when you do Molanbon lacing, by the time you finish building the wheel
nearly all spokes exhibit unyo-n after the thread end,
making wheel building itself difficult—or rather impossible
(I've actually tested this).
With Campagnolo, there's almost no unyo-n at the wheel-building stage,
but I'd never seen it occur this dramatically during years of use.
The customer marked one spot with tape, but
to varying degrees, there are similar traces across the entire thread-end section
on the non-freewheel side.
This is a separate matter, but here's an example of a spoke breaking with Molanbon lacing.
It was built on November 10, 2018 (→here),
and was repaired on November 18 of this year, so exactly 4 years have passed.
It fractured at the thread-end section.
This wheel has Stronger (フリー側) on the freewheel side and Campagnolo on the non-freewheel side,
and the Stronger on the freewheel side is 13-gauge at the head (2.3mm diameter)
and 14-gauge (2.0mm diameter) after that—a single-butted design,
so the diameter doesn't change beyond the thread end.
Also, there were no spokes showing unyo-n on either side.
What this wheel tells us is that the load at the thread-end section
is indeed substantial.
↑This is the rim-side portion of spokes that I disassembled on the non-freewheel side
and extracted by cutting without loosening the nipples at all.
Once spokes that have undergone unyo-n are further tightened,
the tightening doesn't convert to spoke tension but rather to spoke elongation.
So even though the spoke length was originally flush with the nipple end face,
several thread pitches have advanced on some spokes.
↑These are two spokes from the final crossing pair
extracted by fully loosening the nipples.
Flipping one of them over looks like this
I aligned the ends at the spoke head.
The appearance that the right spoke's head is lifting is
the difference between pull-through and counter pull-through spokes.
Because the angle from hub flange to thread start is larger
than on ordinary wheels,
the difference in head deformation becomes more pronounced.
When oriented the same way, they fit together quite cleanly.
And there was about this much difference in length.
↑A diagram of a butted spoke, with the section from thread-end to
where it widens to the nipple diameter drawn in red.
The diagram to the right shows a spoke with evenly spaced horizontal lines,
and the section corresponding to the red in the left diagram is also marked in red on the right.
↑The unyo-n in butted spokes
occurs like this, it seems.
Whether it's "stress concentrating due to the short length of the 1.8mm diameter section,
resulting in unyo-n within the actual spoke tension range"
or simply "1.8mm diameter just can't handle it" is unclear.
If it's the former, with 2.0–1.8–2.0mm butted spokes,
there should be a range where unyo-n occurs but with 1.8mm plain spokes it doesn't.
If it's the latter, then doing Molanbon lacing with 1.8mm diameter would
inevitably cause unyo-n
within the scope of ordinary wheel building.
Also, "spoke count" seems to be a fairly significant factor
in unyo-n occurrence too.
It's plausible that 24H exhibits it while 32H doesn't.
This is an extreme diagram, but
when force is applied to a butted spoke in the bending direction
without causing plastic deformation,
the curvature of the arch varies between the thick and thin diameters.
And if the butted section's length were extremely short,
like an apple dimple or chopstick groove,
the deformation shape would be angular rather than arched.
If both spokes had the same dimensions for thick and thin sections
and were pulled in the extension direction rather than bending,
the top spoke would at most exhibit unyo-n whereas
the bottom spoke would snap (fracture)—
if that's the outcome, it's an extreme comparison,
but the condition of Molanbon lacing's thread-end and beyond
has one foot in the state shown in the bottom spoke in the diagram above.
With 15-gauge plain, the bending deformation's arch curvature is constant throughout,
so butted spokes with 15-gauge sections might actually be stronger
specifically regarding unyo-n resistance during Molanbon lacing
than those with 15-gauge butted sections.
Since testing that would be gambling, I decided to rebuild the non-freewheel side spokes
with 14-gauge plain, which is reliably confirmed to resist unyo-n
even through years of use
within the scope of ordinary wheel building.
I replaced the spokes and roughly trued the wheel radially and laterally.
The wheel in the image above is rotating.
The wheel center at that point.
The rim is significantly shifted toward the freewheel side,
but from here I can build the wheel by just tensioning the non-freewheel side
and bringing the center out.
Though it's not quite that simple—I end up needing to fine-tune radial truing again
toward the end.
Still going.
Center is out.
Before disassembly, I attempted some truing on the original wheel too,
but there was a significant radial deflection, and the spoke at that phase
was undergoing unyo-n, so no matter how much I tightened it, the spoke just kept extending
and the runout never disappeared. Once I replaced the spokes though,
it came together cleanly and snapped into place.
The Campagnolo spokes I reused on the freewheel side
all have spoke lengths flush with the nipple end face, confirmed
with no unyo-n.
Oh right, I also did the main task of replacing the freewheel body bearings.
Since I did that after wheel building,
I checked the center with the centering gauge one more time just in case,
but this bearing replacement didn't shift the center at all.
Done. It went from Campagnolo/Compey to all Campagnolo,
but it's still 46-hole Italian lacing.
It's a rear wheel built with Molanbon lacing.
Originally it was built normally,
but the customer requested it be rebuilt.
The current lacing pattern has been like this since August 2017 (→here),
so it's been over 5 years.
This is apparently the wheel the customer uses most for practice and such.
It's a Rolf (Leaf hub) 24H Campagnolo/Compey 46-hole Molanbon lacing.
It's laced with different diameters on left and right.
The customer brought it in for freewheel body bearing replacement, but
separately from that,
the twisted section where the spokes thread is slipping progressively,
and even after truing it goes out of true quickly.
↑This area
The thing is, at the thin diameter portion where it's butted,
plastic deformation (which I call "unyo-n") is occurring.
Normally, at 1.8mm diameter (15-gauge) there's almost no unyo-n,
but with Molanbon lacing, the yielding forces concentrate on the outer circumference
beyond where the rim-side threading ends,
and unyo-n can occur even with Campagnolo spokes.
With Revolutions where the butted section is 1.5mm diameter (17-gauge),
unyo-n starts occurring probabilistically around 100 kgf
even on ordinary wheels,
but when you do Molanbon lacing, by the time you finish building the wheel
nearly all spokes exhibit unyo-n after the thread end,
making wheel building itself difficult—or rather impossible
(I've actually tested this).
With Campagnolo, there's almost no unyo-n at the wheel-building stage,
but I'd never seen it occur this dramatically during years of use.
The customer marked one spot with tape, but
to varying degrees, there are similar traces across the entire thread-end section
on the non-freewheel side.
This is a separate matter, but here's an example of a spoke breaking with Molanbon lacing.
It was built on November 10, 2018 (→here),
and was repaired on November 18 of this year, so exactly 4 years have passed.
It fractured at the thread-end section.
This wheel has Stronger (フリー側) on the freewheel side and Campagnolo on the non-freewheel side,
and the Stronger on the freewheel side is 13-gauge at the head (2.3mm diameter)
and 14-gauge (2.0mm diameter) after that—a single-butted design,
so the diameter doesn't change beyond the thread end.
Also, there were no spokes showing unyo-n on either side.
What this wheel tells us is that the load at the thread-end section
is indeed substantial.
↑This is the rim-side portion of spokes that I disassembled on the non-freewheel side
and extracted by cutting without loosening the nipples at all.
Once spokes that have undergone unyo-n are further tightened,
the tightening doesn't convert to spoke tension but rather to spoke elongation.
So even though the spoke length was originally flush with the nipple end face,
several thread pitches have advanced on some spokes.
↑These are two spokes from the final crossing pair
extracted by fully loosening the nipples.
Flipping one of them over looks like this
I aligned the ends at the spoke head.
The appearance that the right spoke's head is lifting is
the difference between pull-through and counter pull-through spokes.
Because the angle from hub flange to thread start is larger
than on ordinary wheels,
the difference in head deformation becomes more pronounced.
When oriented the same way, they fit together quite cleanly.
And there was about this much difference in length.
↑A diagram of a butted spoke, with the section from thread-end to
where it widens to the nipple diameter drawn in red.
The diagram to the right shows a spoke with evenly spaced horizontal lines,
and the section corresponding to the red in the left diagram is also marked in red on the right.
↑The unyo-n in butted spokes
occurs like this, it seems.
Whether it's "stress concentrating due to the short length of the 1.8mm diameter section,
resulting in unyo-n within the actual spoke tension range"
or simply "1.8mm diameter just can't handle it" is unclear.
If it's the former, with 2.0–1.8–2.0mm butted spokes,
there should be a range where unyo-n occurs but with 1.8mm plain spokes it doesn't.
If it's the latter, then doing Molanbon lacing with 1.8mm diameter would
inevitably cause unyo-n
within the scope of ordinary wheel building.
Also, "spoke count" seems to be a fairly significant factor
in unyo-n occurrence too.
It's plausible that 24H exhibits it while 32H doesn't.
This is an extreme diagram, but
when force is applied to a butted spoke in the bending direction
without causing plastic deformation,
the curvature of the arch varies between the thick and thin diameters.
And if the butted section's length were extremely short,
like an apple dimple or chopstick groove,
the deformation shape would be angular rather than arched.
If both spokes had the same dimensions for thick and thin sections
and were pulled in the extension direction rather than bending,
the top spoke would at most exhibit unyo-n whereas
the bottom spoke would snap (fracture)—
if that's the outcome, it's an extreme comparison,
but the condition of Molanbon lacing's thread-end and beyond
has one foot in the state shown in the bottom spoke in the diagram above.
With 15-gauge plain, the bending deformation's arch curvature is constant throughout,
so butted spokes with 15-gauge sections might actually be stronger
specifically regarding unyo-n resistance during Molanbon lacing
than those with 15-gauge butted sections.
Since testing that would be gambling, I decided to rebuild the non-freewheel side spokes
with 14-gauge plain, which is reliably confirmed to resist unyo-n
even through years of use
within the scope of ordinary wheel building.
I replaced the spokes and roughly trued the wheel radially and laterally.
The wheel in the image above is rotating.
The wheel center at that point.
The rim is significantly shifted toward the freewheel side,
but from here I can build the wheel by just tensioning the non-freewheel side
and bringing the center out.
Though it's not quite that simple—I end up needing to fine-tune radial truing again
toward the end.
Still going.
Center is out.
Before disassembly, I attempted some truing on the original wheel too,
but there was a significant radial deflection, and the spoke at that phase
was undergoing unyo-n, so no matter how much I tightened it, the spoke just kept extending
and the runout never disappeared. Once I replaced the spokes though,
it came together cleanly and snapped into place.
The Campagnolo spokes I reused on the freewheel side
all have spoke lengths flush with the nipple end face, confirmed
with no unyo-n.
Oh right, I also did the main task of replacing the freewheel body bearings.
Since I did that after wheel building,
I checked the center with the centering gauge one more time just in case,
but this bearing replacement didn't shift the center at all.
Done. It went from Campagnolo/Compey to all Campagnolo,
but it's still 46-hole Italian lacing.