Wednesday is supposed to be my day off, but I've been coming in at irregular hours
to work through the backlog of Nomu Lab wheel builds. Actually, I've been doing this
for about a month straight now. There's a wheel I really want to build for myself,
and yesterday I ended up taking time out for it.
Since it's my day off, I hope you'll forgive me.
This double freewheel hub I built has two weak points, or vulnerabilities.
The main weakness is that the flange width is narrow — about 42.5mm.
There's no way you can get proper lateral stiffness with such narrow flanges.
The vulnerability is a different issue from this unavoidable design constraint —
it's something that should be solvable. I can't write what it is though.
The spokes are Hoshi (Hoshi) No. 15 Aero — not CX-RAY.
I took the hub apart. As for the left and right hub shells:
they're not glued together as you can see.
That means I can change the phase relationship between left and right.
I can build it fine even with a 32H rim that has pair-spoke phasing.
I split it apart. I'm laying paper towels on the freewheel body
as a pillow because I don't want anyone to see the cross-section.
If I can drive 16 rivets in before joining, that's an 80,000 point bonus.
In the middle of a rough build for reassembly.
For a particular reason, I'm doing a rough build without the freewheel body and shaft.
↑Narrow!
With twist-lacing, you need longer spokes than with standard lacing.
The amount of increase varies depending on whether you twist a 4-cross
or a 6-cross pattern. I need to build with this pattern to investigate that,
but also to overcome the vulnerability.
Built it.
I've got the spoke tension at a level that's fine for riding.
The runout is true and the centering is good.
But then I took it apart again.
The spokes have taken a set, so I can't reuse them.
I released the tension before cutting them off.
A wheel that only existed for 10 minutes.
But I did learn something from it. Hmm, interesting.
Next I'm going to do a 10-spoke lacing.
The maximum tangential lacing for 32H is 8-cross, so spokes
coming from holes beyond the hub flange bisector will cross.
Since X-spoke = X/2 cross, with this pattern
a single spoke will cross 5 times with other spokes when viewed from the side.
Of those, the 3rd, 4th, and 5th crossings from the hub side can be woven,
and I'll weave all of them. Since that's 3 weaves, it's a triple-cross.
Rough build is done.
Looking at a spoke, the 3rd crossing is below, the 4th is above,
and the 5th is below. I've woven it 3 times.
Built it.
This lacing pattern has nothing to do with solving the vulnerability.
When you run two or more experiments at once,
you don't get clean data.
Weakness #1 is something you'd notice instantly upon riding, and it was solved.
I probably solved #2 as well, but it requires monitoring over time
so I can't draw conclusions yet.
The spokes I used for this wheel build
were DT Competition, both times.
to work through the backlog of Nomu Lab wheel builds. Actually, I've been doing this
for about a month straight now. There's a wheel I really want to build for myself,
and yesterday I ended up taking time out for it.
Since it's my day off, I hope you'll forgive me.
This double freewheel hub I built has two weak points, or vulnerabilities.
The main weakness is that the flange width is narrow — about 42.5mm.
There's no way you can get proper lateral stiffness with such narrow flanges.
The vulnerability is a different issue from this unavoidable design constraint —
it's something that should be solvable. I can't write what it is though.
The spokes are Hoshi (Hoshi) No. 15 Aero — not CX-RAY.
I took the hub apart. As for the left and right hub shells:
they're not glued together as you can see.
That means I can change the phase relationship between left and right.
I can build it fine even with a 32H rim that has pair-spoke phasing.
I split it apart. I'm laying paper towels on the freewheel body
as a pillow because I don't want anyone to see the cross-section.
If I can drive 16 rivets in before joining, that's an 80,000 point bonus.
In the middle of a rough build for reassembly.
For a particular reason, I'm doing a rough build without the freewheel body and shaft.
↑Narrow!
With twist-lacing, you need longer spokes than with standard lacing.
The amount of increase varies depending on whether you twist a 4-cross
or a 6-cross pattern. I need to build with this pattern to investigate that,
but also to overcome the vulnerability.
Built it.
I've got the spoke tension at a level that's fine for riding.
The runout is true and the centering is good.
But then I took it apart again.
The spokes have taken a set, so I can't reuse them.
I released the tension before cutting them off.
A wheel that only existed for 10 minutes.
But I did learn something from it. Hmm, interesting.
Next I'm going to do a 10-spoke lacing.
The maximum tangential lacing for 32H is 8-cross, so spokes
coming from holes beyond the hub flange bisector will cross.
Since X-spoke = X/2 cross, with this pattern
a single spoke will cross 5 times with other spokes when viewed from the side.
Of those, the 3rd, 4th, and 5th crossings from the hub side can be woven,
and I'll weave all of them. Since that's 3 weaves, it's a triple-cross.
Rough build is done.
Looking at a spoke, the 3rd crossing is below, the 4th is above,
and the 5th is below. I've woven it 3 times.
Built it.
This lacing pattern has nothing to do with solving the vulnerability.
When you run two or more experiments at once,
you don't get clean data.
Weakness #1 is something you'd notice instantly upon riding, and it was solved.
I probably solved #2 as well, but it requires monitoring over time
so I can't draw conclusions yet.
The spokes I used for this wheel build
were DT Competition, both times.