DT's entry into the hub business wasn't all that long ago.
Before the current models, they were sold under the brand name "Hugi" (フギ).
(I typed "u" because entering the umlaut is a hassle)
Now, this Hugi hub doesn't have anything particularly noteworthy about it as a hub.
But there is one remarkably unusual point—
this is an 18-hole rear hub.
18 holes means 9 holes per side.
↑I often think about hub flanges by laying them out like this.
I call it a Mercator projection-style layout,
but you could also call it an Admiral-style or Mizkan-style layout.
Edit: I corrected my cut line placement as I had it wrong initially.
Thanks for pointing that out in the comments.
Now, if you try to use a 2-cross lacing pattern with 9 holes per side,
since it's an odd number you end up with one extra spoke. There's no way around it—this would create poor wheel balance,
so you need to think about a lacing pattern that doesn't completely mess up the balance.
↑And so you're left with only these two options.
Either alternate between 2-cross and radial lacing, or run radial spokes through a 2-cross pattern.
The former results in less runout variation across the entire rim,
so if I ever need to build an 18-hole rear wheel, I lace it X-I style.
Because one flange has both tangential and radial lacing mixed together,
naturally the spoke tension differs between each lacing method.
Wheel building is the process of positioning the rim at the center of the hub's overall width
while reducing both radial and lateral runout to the best possible compromise level.
This radial runout shown as a blue line—
↑you want to make it as circular as possible,
but the tangential and radial spoke tensions will necessarily differ in achieving this.
↑If you tried to equalize spoke tension instead,
roughly speaking, you'd end up with something like this.
Manage runout and tension gets uneven; equalize tension and runout appears. As for which is better for actual wheel use, it's the former.
Spoke tension is the tension in the spoke itself.
This differs from what I'm calling RK in the diagram above.
Spoke tension ≠ RK.
It's RK that relates to nipple-induced rim damage.
When RK is the same, thicker spokes result in lower spoke tension.
Since RK is practically unmeasurable in reality,
rim manufacturers specify spoke tension as the threshold for damage resistance.
But even with the same spoke tension, RK changes depending on spoke thickness.
Since thicker spokes produce larger RK at the same tension,
for example, if you take a rim with a 120kgf upper limit
and tension a 2.0mm plain spoke to 120kgf,
the rim's damage risk at that point (= RK) might be comparable to
a CX-RAY's 140kgf.
However, the rim states "Up to 120kgf!" so
we have no choice but to follow that spoke tension limit.
Spoke tension, unlike RK, can be approximated quite accurately,
so as a manufacturer, that's the only way to set the threshold.
Anyway, the point is: spoke diameter and spoke tension have an inverse relationship—
"thicker = lower, thinner = higher."
So, if you lace an X-I pattern using all the same spoke model,
this is what I call same-diameter, mixed-lace-pattern lacing on one flange.
It's possible to express same-diameter mixed-lace-pattern lacing
in terms of spoke tension using different-diameter, same-lace-pattern lacing.
Imagine if you could infinitely vary the proportions of the two spoke types in a different-diameter, same-lace-pattern build.
(And imagine you could freely set the plain spoke diameter yourself, including the nipple hole size)
Then you could probably adjust the spoke tension ratio
between the tangential and radial spokes in an 18H same-diameter mixed-lace-pattern build
to match that of different-diameter, same-lace-pattern spokes
using fine and thick variants.
In reality, though, tangential and radial lacing behave differently under torsional stress, so
if it came to that, the same-diameter mixed-lace-pattern build with two-thirds tangential lacing remaining
would probably be stronger against twisting.
With the DT Hugi 18H rear hub,
you have no choice but to use mixed-lace-pattern lacing on one flange.
Also, for rear wheels, pure radial lacing on both sides is absolutely not an option,
so at least one side must use tangential lacing.
But what if this were an 18H front wheel instead?
In that case, you'd normally just go with radial lacing across the board.
Different-diameter lacing on one flange would, in spoke tension terms,
be equivalent to doing radial lacing with mismatched spoke thicknesses.
If I could explain the advantages of using mismatched spoke thicknesses for radial lacing,
then I could also explain the advantages of mixing tangential and radial lacing on one flange.
But I can't. And that's true both in theory and from experience.
I once built what I call a "stupid wheel" (※a badly designed wheel)
with a 32-hole rim where one side had 16 holes laced in a 1-cross twisting pattern mixed with 2 radial spokes.
The radial runout wouldn't clean up properly.
The radial spokes shown in red had lower tension,
and lower tension means closer to zero.
When I actually rode it, the radial spoke side felt like it wasn't doing any work transmitting power or maintaining stiffness.
To exaggerate, it felt barely different from a wheel with only twisted spokes.
And those radial spokes on that side kept loosening.
So I modified the lacing pattern, changing the radial spokes to 1-cross.
Now the only difference was whether they were twisted or not,
but within about two months, spoke nipples started breaking, so I disassembled it.
I wasn't planning to sell it anyway, but I got the obvious answer that it definitely wouldn't be sellable.
Different-diameter lacing on one flange is basically a no-go.
DT makes a wheel concept called Tricon,
and I think they avoid the frequent spoke-nipple-failure problem
because they alter the flange height where spokes exit on the hub side
and the orientation of the non-radial spokes is really well thought out.
Plus, they use straight spokes. That's also a big factor.
↑Found it while searching.
This is an image from when I was building wheels like this.
This is "freewheel side plain spoke X-I lacing / non-freewheel side non-plain spoke radial lacing."
Later I changed the non-freewheel side to full plain spoke X-I lacing too, but
I don't have images from that.
I use neologisms like "X-I lacing," but I'm not trying to claim I invented it.
Anyone handed an 18H rear hub would basically have to lace it this way.
The matching front wheel was a Tni carbon hub, 18H, X-I laced,
but back then I didn't have the idea of different-diameter spokes,
and since it was a prototype I was nervous about thin spokes, so I built the whole thing with 2.0mm plain spokes.
If you coin terminology like "plain spoke lacing" or "46-spoke lacing" or "X-I lacing,"
you can fit pretty much any lacing pattern into a category, which is convenient.
↑If you obsess only over spoke tension,
making the 3 radial spokes thinner would supposedly
(though I can't say exactly what the ratio should be)
approximate the tension of the 6 tangential spokes.
The idea would be to mix different-spoke-count different-diameter lacing to get as close to zero difference as possible.
But this only equalizes spoke tension—it doesn't equalize spoke deformation or
nipple-side RK.
However, there is a breakthrough lacing method that equalizes all of these (tension, deformation, RK).
Feel free to copy it!
It's same-spoke-count same-diameter lacing! With 18H tangential lacing you can't use same-spoke-count for all spokes, so just make it radial! If you make all spokes the same model and lace them radially,
on a front wheel you'll have matching spoke tension, deformation, and RK!
18H X-I lacing is a last-resort approach forced upon you only when you "happen to be building a rear wheel with an 18H rear hub," and it has almost no advantages!
The biggest drawback is spoke tension variation.
Well, this isn't something you'd notice without a tension meter.
Before the current models, they were sold under the brand name "Hugi" (フギ).
(I typed "u" because entering the umlaut is a hassle)
Now, this Hugi hub doesn't have anything particularly noteworthy about it as a hub.
But there is one remarkably unusual point—
this is an 18-hole rear hub.
18 holes means 9 holes per side.
↑I often think about hub flanges by laying them out like this.
I call it a Mercator projection-style layout,
but you could also call it an Admiral-style or Mizkan-style layout.
Edit: I corrected my cut line placement as I had it wrong initially.
Thanks for pointing that out in the comments.
Now, if you try to use a 2-cross lacing pattern with 9 holes per side,
since it's an odd number you end up with one extra spoke. There's no way around it—this would create poor wheel balance,
so you need to think about a lacing pattern that doesn't completely mess up the balance.
↑And so you're left with only these two options.
Either alternate between 2-cross and radial lacing, or run radial spokes through a 2-cross pattern.
The former results in less runout variation across the entire rim,
so if I ever need to build an 18-hole rear wheel, I lace it X-I style.
Because one flange has both tangential and radial lacing mixed together,
naturally the spoke tension differs between each lacing method.
Wheel building is the process of positioning the rim at the center of the hub's overall width
while reducing both radial and lateral runout to the best possible compromise level.
This radial runout shown as a blue line—
↑you want to make it as circular as possible,
but the tangential and radial spoke tensions will necessarily differ in achieving this.
↑If you tried to equalize spoke tension instead,
roughly speaking, you'd end up with something like this.
Manage runout and tension gets uneven; equalize tension and runout appears. As for which is better for actual wheel use, it's the former.
Spoke tension is the tension in the spoke itself.
This differs from what I'm calling RK in the diagram above.
Spoke tension ≠ RK.
It's RK that relates to nipple-induced rim damage.
When RK is the same, thicker spokes result in lower spoke tension.
Since RK is practically unmeasurable in reality,
rim manufacturers specify spoke tension as the threshold for damage resistance.
But even with the same spoke tension, RK changes depending on spoke thickness.
Since thicker spokes produce larger RK at the same tension,
for example, if you take a rim with a 120kgf upper limit
and tension a 2.0mm plain spoke to 120kgf,
the rim's damage risk at that point (= RK) might be comparable to
a CX-RAY's 140kgf.
However, the rim states "Up to 120kgf!" so
we have no choice but to follow that spoke tension limit.
Spoke tension, unlike RK, can be approximated quite accurately,
so as a manufacturer, that's the only way to set the threshold.
Anyway, the point is: spoke diameter and spoke tension have an inverse relationship—
"thicker = lower, thinner = higher."
So, if you lace an X-I pattern using all the same spoke model,
this is what I call same-diameter, mixed-lace-pattern lacing on one flange.
It's possible to express same-diameter mixed-lace-pattern lacing
in terms of spoke tension using different-diameter, same-lace-pattern lacing.
Imagine if you could infinitely vary the proportions of the two spoke types in a different-diameter, same-lace-pattern build.
(And imagine you could freely set the plain spoke diameter yourself, including the nipple hole size)
Then you could probably adjust the spoke tension ratio
between the tangential and radial spokes in an 18H same-diameter mixed-lace-pattern build
to match that of different-diameter, same-lace-pattern spokes
using fine and thick variants.
In reality, though, tangential and radial lacing behave differently under torsional stress, so
if it came to that, the same-diameter mixed-lace-pattern build with two-thirds tangential lacing remaining
would probably be stronger against twisting.
With the DT Hugi 18H rear hub,
you have no choice but to use mixed-lace-pattern lacing on one flange.
Also, for rear wheels, pure radial lacing on both sides is absolutely not an option,
so at least one side must use tangential lacing.
But what if this were an 18H front wheel instead?
In that case, you'd normally just go with radial lacing across the board.
Different-diameter lacing on one flange would, in spoke tension terms,
be equivalent to doing radial lacing with mismatched spoke thicknesses.
If I could explain the advantages of using mismatched spoke thicknesses for radial lacing,
then I could also explain the advantages of mixing tangential and radial lacing on one flange.
But I can't. And that's true both in theory and from experience.
I once built what I call a "stupid wheel" (※a badly designed wheel)
with a 32-hole rim where one side had 16 holes laced in a 1-cross twisting pattern mixed with 2 radial spokes.
The radial runout wouldn't clean up properly.
The radial spokes shown in red had lower tension,
and lower tension means closer to zero.
When I actually rode it, the radial spoke side felt like it wasn't doing any work transmitting power or maintaining stiffness.
To exaggerate, it felt barely different from a wheel with only twisted spokes.
And those radial spokes on that side kept loosening.
So I modified the lacing pattern, changing the radial spokes to 1-cross.
Now the only difference was whether they were twisted or not,
but within about two months, spoke nipples started breaking, so I disassembled it.
I wasn't planning to sell it anyway, but I got the obvious answer that it definitely wouldn't be sellable.
Different-diameter lacing on one flange is basically a no-go.
DT makes a wheel concept called Tricon,
and I think they avoid the frequent spoke-nipple-failure problem
because they alter the flange height where spokes exit on the hub side
and the orientation of the non-radial spokes is really well thought out.
Plus, they use straight spokes. That's also a big factor.
↑Found it while searching.
This is an image from when I was building wheels like this.
This is "freewheel side plain spoke X-I lacing / non-freewheel side non-plain spoke radial lacing."
Later I changed the non-freewheel side to full plain spoke X-I lacing too, but
I don't have images from that.
I use neologisms like "X-I lacing," but I'm not trying to claim I invented it.
Anyone handed an 18H rear hub would basically have to lace it this way.
The matching front wheel was a Tni carbon hub, 18H, X-I laced,
but back then I didn't have the idea of different-diameter spokes,
and since it was a prototype I was nervous about thin spokes, so I built the whole thing with 2.0mm plain spokes.
If you coin terminology like "plain spoke lacing" or "46-spoke lacing" or "X-I lacing,"
you can fit pretty much any lacing pattern into a category, which is convenient.
↑If you obsess only over spoke tension,
making the 3 radial spokes thinner would supposedly
(though I can't say exactly what the ratio should be)
approximate the tension of the 6 tangential spokes.
The idea would be to mix different-spoke-count different-diameter lacing to get as close to zero difference as possible.
But this only equalizes spoke tension—it doesn't equalize spoke deformation or
nipple-side RK.
However, there is a breakthrough lacing method that equalizes all of these (tension, deformation, RK).
Feel free to copy it!
It's same-spoke-count same-diameter lacing! With 18H tangential lacing you can't use same-spoke-count for all spokes, so just make it radial! If you make all spokes the same model and lace them radially,
on a front wheel you'll have matching spoke tension, deformation, and RK!
18H X-I lacing is a last-resort approach forced upon you only when you "happen to be building a rear wheel with an 18H rear hub," and it has almost no advantages!
The biggest drawback is spoke tension variation.
Well, this isn't something you'd notice without a tension meter.