A customer brought in the front and rear wheels of a fixed-gear bike built with Araya 16B gold rims.
They wanted both wheels to have their tubular tires replaced using rim cement,
but since the customer is capable of doing tubular tire installation with rim cement themselves,
the fact that they're asking our shop specifically means
that in addition to the straightforward tire replacement work,
they're vaguely sensing some pitfalls unique to this particular job.
For what it's worth, I'm sensing them too.
Dura-Ace small-flange hubs, 36H,
laced in Italian (crossed) style.
This wasn't a request from a track cyclist, but I used to think
that anything not laced in JIS pattern couldn't be used for keirin racing.
Apparently, as long as all the components are NJS-certified and the wheel is at least 8-spoke laced,
it passes inspection.
I've actually seen keirin wheels laced Italian style with keirin valves installed
that passed inspection.
When I aligned the printing on the hub shell of the Dura-Ace in the image above,
the Italian lacing's left and right hand characteristics also lined up.
And looking at the wheels from the right side in that orientation,
the rear wheel's label was facing backwards, but
we shouldn't worry about that.
The tires are Vittoria Rally with red sidewalls and Lion Tire GT30,
and on both wheels the tread pattern was installed in the reverse direction.
But this is somewhat unavoidable, you see,
because both tires have single-sided labels, and if you follow the convention that
"when a label is single-sided, the label should be on the right,"
you'd end up mounting them backwards.
The tires themselves don't have any instructions specifying reverse mounting.
TUFO has tires like this too, but
I prioritize the tread pattern over the label position, even if it means the label ends up on the left.
MOST's tubular tire levers are super easy to use!(not an ad)
The rear wheel in particular was glued down so firmly
that I struggled with the initial insertion of the tire lever.
The rim cement was the slow-drying variety.
As I've written before, Panaracer rim cement comes in cans or tubes marked as "fast-drying,"
but considering the history of rim cement,
that's arguably ultra-fast-drying stuff.
(It's better to avoid it if possible.)
You can ride a few hours after installation, and
leaving it for a day guarantees it's definitely ready.
Truly slow-drying rim cement needs to cure for about a week.
You might think, "Why not just use fast-drying cement for everything?"
but slow-drying rim cement has the advantage that
between the tire and rim,
the cement remains tacky and effective rather than drying completely hard,
so even if you forcefully pry the wheel, the tire
doesn't easily come off the rim.
...Or so the theory goes,
but since rim cement itself has been improving over the years,
fast-drying rim cement doesn't lose adhesion
so quickly that you'd need to re-glue tires
within the lifespan of regular use.
Even with cyclocross tires that luckily never punctured
and weren't used at all during the off-season,
they'll hold for 2 years (2 seasons).
Of course, if you're running something for 3, 5, or more years,
you should re-glue.
When fast-drying rim cement fully cures,
it dries into a hard, flaky powder that
comes off the rim easily
(riding while it's in this near-final state—at least with road bikes—
I'm not aware of any cases of tires coming off),
but when slow-drying rim cement fully cures,
it forms quite a strong bed,
making the tire hard to peel off, and
the fundoshi (base tape) of the tubular tire
often remains stuck to the rim.
The customer wasn't deliberately using
slow-drying rim cement—they apparently used
a bulk large-can rim cement sitting in a corner of the workshop
to install the tires.
While that can may say "fast-drying,"
if we're comparing it to Panaracer's fast-drying cement,
it's considerably slow-drying in character.
This wheel had rim cement stuck to the hub shell and rim sides too,
but small cans with brushes or
tube-type rim cement (like toothpaste tubes)
are less prone to such mistakes.
↑When the valve hole was pointing straight up,
it apparently dripped toward the hub shell.
This hub is only NJS-certified for 36H builds,
and since it's actually 36H, it carries the NJS stamp.
Apart from the current spoke orientation,
there are traces of old spoke holes, so this hub has been
laced into a wheel at least twice before.
The rear hub has threaded ends on both sides.
With Italian lacing, theoretically
you shouldn't flip the hub and use it the other way,
but there were marks from a small cog.
However, theoretically the right side
was definitely used more,
↑the left flange compared to
the right flange is dirtier,
and the old spoke holes shown in the image had deformation.
There was runout related to this phase orientation,
which I corrected, but these spokes are
15-gauge based double-butted spokes,
and since our shop doesn't have
steady stock of that same gauge,
I didn't replace the spokes.
The spokes on both wheels, front and rear, both sides,
are Hoshi (Star) "Stainless."
Here "Stainless" isn't a generic term but
the brand name of that particular spoke model.
The head is stamped with a ☆ mark as shown in the image,
whereas Starlight spokes have a capital H stamp.
I've written this many times before,
but Stainless spokes barely react to magnets.
By contrast, Starlight spokes attract almost like iron—
they snap right onto a magnet.
There was a period when Hoshi's quality control on these two models was sloppy,
and when I inquired with them directly, they gave me a dishonest response saying
"That's not the case; we do things properly."
So I hold a deep distrust of Hoshi spokes.
A few years ago, I did get confirmation that what I was saying wasn't wrong,
but that doesn't mean I'll suddenly start using Hoshi spokes,
so when building keirin wheels, I use
NJS-certified spokes, and except for the old Aero Starlight
(which Hoshi apparently lost the manufacturing know-how for),
I don't use Hoshi spokes.
The currently available Aero Starlight model with NJS non-certified status
appears to be 100% compositionally Starlight spokes.
With this wheel,
the front wheel's ☆-stamped spokes barely reacted to magnets,
but the rear wheel's ☆-stamped spokes attracted magnets
as strongly as Starlight spokes.
What I call "fake Starlight" or "shitty Starlight"—
spokes with H stamps—react slightly to magnets,
more than Stainless but clearly not with Starlight's magnetism,
a middle-ground property spoke.
If we rate magnetic response as Stainless = 1
and Starlight = 10, these feel like 2-3.
"Not Stainless but practically Stainless"—a neither-nor spoke.
In contrast, this fake Stainless had
about the same magnetism as Starlight.
Using the same scale, I'd say 9 or 10.
There are cases of getting counterfeits when actually seeking Starlight,
but since we don't stock Stainless originally,
I rarely encounter fake Stainless except
in situations like this.
With Dura-Ace fixed-gear hubs and Araya 16B gold rims,
you can build 305mm even with 8-spoke lacing regardless of whether the hub is small or large flange,
so if you manage spokes by putting them in one drawer and restocking as needed,
one wheel could end up with mixed Stainless and
fake Stainless, but
on this wheel's front and rear,
whether the spoke batches were different or they were built at different times is unclear,
but the front wheel was entirely Stainless spokes,
and the rear wheel had fake Stainless for all but one spoke.
Theoretically the rear wheel would have spokes less prone to breaking,
but the rear wheel has a history of one spoke breaking,
and an Asahi spoke of the same gauge was used as a repair
(it's not impossible that a spoke was mixed in because one was short
during new wheel building, but probably not).
The stamp is a capital A.
This spoke has the same composition as Hoshi's Stainless equivalent,
so on the rear wheel this one spoke alone
barely reacted to magnets.
Hoshi's double-butted spokes have
a pronounced step where the diameter changes,
whereas Asahi's transition is less distinct.
DT's taper where diameter changes is longer,
so the transition is even less clear,
and Sapim has cutting marks
like a cut-glass pattern.
The front Lion Tire had
carcass damage at a spot where the open side
had apparently been crushed flat from sitting on the ground for years,
and twisting it slightly reveals the inner tube.
This is a budget-tier tire
that was once called a "1000-yen tire."
Those were the days.
Incidentally, there was a model called Formula Uno (F1)
with the same tread pattern as Rally, but
going further back, Formula Uno had
a center tread in slick pattern and
an orange-only tread surface.
Rally and Formula Uno from that era were also 1000-yen tires, and
about 25 years ago I regularly used the orange Formula Uno.
Back then, even if tubular tires were inferior in
longitudinal runout and tread center precision,
the 1000-yen tubular was somewhat better in performance than
premium WO tires, but
for the past 20 years or so, WO tire evolution has been remarkable,
and WO tires are no longer something you "tolerate using"
in terms of performance and ride quality.
With Continental, for example, the GP 5000 WO tire
rolls more lightly and has higher cornering limits
than the Giro tubular, so
while putting competition tubulars on your best wheels is one thing,
there's no need to use tubulars for everyday wheels anymore.
That said, between the absolute best tubulars and
the best WO (including tubeless) tires,
I still think tubulars have overall better performance.
Back to the main topic now.
The fundoshi of the tubular tire had firmly dried in the slow-drying cement,
leaving residue on the rim in quite a few places all the way around.
Wild guess, but the customer was anticipating this.
Oops—
I had to partially disassemble an STI lever at year-end last year
before I'd fully mapped it out,
and I modified a chipping tool tip
to press and remove a small snap ring.
Well, whatever.
↑This is mid-work, but
scraping off the fundoshi was incredibly tedious.
The new tires fitted are
Panaracer Practice 270,
a model name that screams "practice tire."
270 refers to the claimed weight,
indicating the weight of the narrower 22.5C size in the lineup.
There's also a 25C size available,
with a claimed weight of 320g.
By the way, this tire has printing on its fundoshi and
a "MADE IN THAILAND" marking in one spot on the tread,
so I can definitively say
it's an OEM product from Lion Tire.
The original rear tire, the Rally, and
Continental's Giro mentioned earlier
are also made by Lion Tire (→here).
Before installing the tires, I inspected them
because I planned to correct any runout or center misalignment,
but both wheels had only slight lateral runout,
and no center misalignment.
They wanted both wheels to have their tubular tires replaced using rim cement,
but since the customer is capable of doing tubular tire installation with rim cement themselves,
the fact that they're asking our shop specifically means
that in addition to the straightforward tire replacement work,
they're vaguely sensing some pitfalls unique to this particular job.
For what it's worth, I'm sensing them too.
Dura-Ace small-flange hubs, 36H,
laced in Italian (crossed) style.
This wasn't a request from a track cyclist, but I used to think
that anything not laced in JIS pattern couldn't be used for keirin racing.
Apparently, as long as all the components are NJS-certified and the wheel is at least 8-spoke laced,
it passes inspection.
I've actually seen keirin wheels laced Italian style with keirin valves installed
that passed inspection.
When I aligned the printing on the hub shell of the Dura-Ace in the image above,
the Italian lacing's left and right hand characteristics also lined up.
And looking at the wheels from the right side in that orientation,
the rear wheel's label was facing backwards
we shouldn't worry about that
The tires are Vittoria Rally with red sidewalls and Lion Tire GT30,
and on both wheels the tread pattern was installed in the reverse direction.
But this is somewhat unavoidable, you see,
because both tires have single-sided labels, and if you follow the convention that
"when a label is single-sided, the label should be on the right,"
you'd end up mounting them backwards.
The tires themselves don't have any instructions specifying reverse mounting.
TUFO has tires like this too, but
I prioritize the tread pattern over the label position, even if it means the label ends up on the left.
MOST's tubular tire levers are super easy to use!
The rear wheel in particular was glued down so firmly
that I struggled with the initial insertion of the tire lever.
The rim cement was the slow-drying variety.
As I've written before, Panaracer rim cement comes in cans or tubes marked as "fast-drying,"
but considering the history of rim cement,
that's arguably ultra-fast-drying stuff.
(It's better to avoid it if possible.)
You can ride a few hours after installation, and
leaving it for a day guarantees it's definitely ready.
Truly slow-drying rim cement needs to cure for about a week.
You might think, "Why not just use fast-drying cement for everything?"
but slow-drying rim cement has the advantage that
between the tire and rim,
the cement remains tacky and effective rather than drying completely hard,
so even if you forcefully pry the wheel, the tire
doesn't easily come off the rim.
...Or so the theory goes,
but since rim cement itself has been improving over the years,
fast-drying rim cement doesn't lose adhesion
so quickly that you'd need to re-glue tires
within the lifespan of regular use.
Even with cyclocross tires that luckily never punctured
and weren't used at all during the off-season,
they'll hold for 2 years (2 seasons).
Of course, if you're running something for 3, 5, or more years,
you should re-glue.
When fast-drying rim cement fully cures,
it dries into a hard, flaky powder that
comes off the rim easily
(riding while it's in this near-final state—at least with road bikes—
I'm not aware of any cases of tires coming off),
but when slow-drying rim cement fully cures,
it forms quite a strong bed,
making the tire hard to peel off, and
the fundoshi (base tape) of the tubular tire
often remains stuck to the rim.
The customer wasn't deliberately using
slow-drying rim cement—they apparently used
a bulk large-can rim cement sitting in a corner of the workshop
to install the tires.
While that can may say "fast-drying,"
if we're comparing it to Panaracer's fast-drying cement,
it's considerably slow-drying in character.
This wheel had rim cement stuck to the hub shell and rim sides too,
but small cans with brushes or
tube-type rim cement (like toothpaste tubes)
are less prone to such mistakes.
↑When the valve hole was pointing straight up,
it apparently dripped toward the hub shell.
This hub is only NJS-certified for 36H builds,
and since it's actually 36H, it carries the NJS stamp.
Apart from the current spoke orientation,
there are traces of old spoke holes, so this hub has been
laced into a wheel at least twice before.
The rear hub has threaded ends on both sides.
With Italian lacing, theoretically
you shouldn't flip the hub and use it the other way,
but there were marks from a small cog.
However, theoretically the right side
was definitely used more,
↑the left flange compared to
the right flange is dirtier,
and the old spoke holes shown in the image had deformation.
There was runout related to this phase orientation,
which I corrected, but these spokes are
15-gauge based double-butted spokes,
and since our shop doesn't have
steady stock of that same gauge,
I didn't replace the spokes.
The spokes on both wheels, front and rear, both sides,
are Hoshi (Star) "Stainless."
Here "Stainless" isn't a generic term but
the brand name of that particular spoke model.
The head is stamped with a ☆ mark as shown in the image,
whereas Starlight spokes have a capital H stamp.
I've written this many times before,
but Stainless spokes barely react to magnets.
By contrast, Starlight spokes attract almost like iron—
they snap right onto a magnet.
There was a period when Hoshi's quality control on these two models was sloppy,
and when I inquired with them directly, they gave me a dishonest response saying
"That's not the case; we do things properly."
So I hold a deep distrust of Hoshi spokes.
A few years ago, I did get confirmation that what I was saying wasn't wrong,
but that doesn't mean I'll suddenly start using Hoshi spokes,
so when building keirin wheels, I use
NJS-certified spokes, and except for the old Aero Starlight
(which Hoshi apparently lost the manufacturing know-how for),
I don't use Hoshi spokes.
The currently available Aero Starlight model with NJS non-certified status
appears to be 100% compositionally Starlight spokes.
With this wheel,
the front wheel's ☆-stamped spokes barely reacted to magnets,
but the rear wheel's ☆-stamped spokes attracted magnets
as strongly as Starlight spokes.
What I call "fake Starlight" or "shitty Starlight"—
spokes with H stamps—react slightly to magnets,
more than Stainless but clearly not with Starlight's magnetism,
a middle-ground property spoke.
If we rate magnetic response as Stainless = 1
and Starlight = 10, these feel like 2-3.
"Not Stainless but practically Stainless"—a neither-nor spoke.
In contrast, this fake Stainless had
about the same magnetism as Starlight.
Using the same scale, I'd say 9 or 10.
There are cases of getting counterfeits when actually seeking Starlight,
but since we don't stock Stainless originally,
I rarely encounter fake Stainless except
in situations like this.
With Dura-Ace fixed-gear hubs and Araya 16B gold rims,
you can build 305mm even with 8-spoke lacing regardless of whether the hub is small or large flange,
so if you manage spokes by putting them in one drawer and restocking as needed,
one wheel could end up with mixed Stainless and
fake Stainless, but
on this wheel's front and rear,
whether the spoke batches were different or they were built at different times is unclear,
but the front wheel was entirely Stainless spokes,
and the rear wheel had fake Stainless for all but one spoke.
Theoretically the rear wheel would have spokes less prone to breaking,
but the rear wheel has a history of one spoke breaking,
and an Asahi spoke of the same gauge was used as a repair
(it's not impossible that a spoke was mixed in because one was short
during new wheel building, but probably not).
The stamp is a capital A.
This spoke has the same composition as Hoshi's Stainless equivalent,
so on the rear wheel this one spoke alone
barely reacted to magnets.
Hoshi's double-butted spokes have
a pronounced step where the diameter changes,
whereas Asahi's transition is less distinct.
DT's taper where diameter changes is longer,
so the transition is even less clear,
and Sapim has cutting marks
like a cut-glass pattern.
The front Lion Tire had
carcass damage at a spot where the open side
had apparently been crushed flat from sitting on the ground for years,
and twisting it slightly reveals the inner tube.
This is a budget-tier tire
that was once called a "1000-yen tire."
Incidentally, there was a model called Formula Uno (F1)
with the same tread pattern as Rally, but
going further back, Formula Uno had
a center tread in slick pattern and
an orange-only tread surface.
Rally and Formula Uno from that era were also 1000-yen tires, and
about 25 years ago I regularly used the orange Formula Uno.
Back then, even if tubular tires were inferior in
longitudinal runout and tread center precision,
the 1000-yen tubular was somewhat better in performance than
premium WO tires, but
for the past 20 years or so, WO tire evolution has been remarkable,
and WO tires are no longer something you "tolerate using"
in terms of performance and ride quality.
With Continental, for example, the GP 5000 WO tire
rolls more lightly and has higher cornering limits
than the Giro tubular, so
while putting competition tubulars on your best wheels is one thing,
there's no need to use tubulars for everyday wheels anymore.
That said, between the absolute best tubulars and
the best WO (including tubeless) tires,
I still think tubulars have overall better performance.
Back to the main topic now.
The fundoshi of the tubular tire had firmly dried in the slow-drying cement,
leaving residue on the rim in quite a few places all the way around.
Wild guess, but the customer was anticipating this.
Oops—
I had to partially disassemble an STI lever at year-end last year
before I'd fully mapped it out,
and I modified a chipping tool tip
to press and remove a small snap ring.
Well, whatever.
↑This is mid-work, but
scraping off the fundoshi was incredibly tedious.
The new tires fitted are
Panaracer Practice 270,
a model name that screams "practice tire."
270 refers to the claimed weight,
indicating the weight of the narrower 22.5C size in the lineup.
There's also a 25C size available,
with a claimed weight of 320g.
By the way, this tire has printing on its fundoshi and
a "MADE IN THAILAND" marking in one spot on the tread,
so I can definitively say
it's an OEM product from Lion Tire.
The original rear tire, the Rally, and
Continental's Giro mentioned earlier
are also made by Lion Tire (→here).
Before installing the tires, I inspected them
because I planned to correct any runout or center misalignment,
but both wheels had only slight lateral runout,
and no center misalignment.