half radial, again

JG said:

Am I right that torque puts much greater tension (~100x) on a radial
spoke than a tangentially laced spoke?

In the absence of tangential spokes, yes.

Quoted message said:

Doesn't this mean a R/3X wheel
puts a uniform, but much greater tension on the non-drive spokes under
torque?

No, because the crossed spokes on the right side don't let the wheel
wind up enough to affect the tension on the left by much.

Chalo

Like Chalo said... if you have nearly tangential spokes on the drive
side, then the NDS radial spokes are essentially taken out of the
torque picture.

As Sheldon mentions, a stiff hub body combined with crossed spokes and
low tension (necessary because of dish), will make it more likely that
combined torque, radial, and lateral loads will cause the NDS spokes
to go slack. Radial lacing the NDS removes one of these components...
at the expense of making the drive side carry all the torque.

Ron Ruff said:

Like Chalo said... if you have nearly tangential spokes on the drive
side, then the NDS radial spokes are essentially taken out of the
torque picture.

As Sheldon mentions, a stiff hub body combined with crossed spokes and
low tension (necessary because of dish), will make it more likely that
combined torque, radial, and lateral loads will cause the NDS spokes
to go slack. Radial lacing the NDS removes one of these components...
at the expense of making the drive side carry all the torque.

My Shimano-built rear wheel has radial lacing on the drive side (and
some cross pattern on the NDS). I was concerned that this was
incorrect and would lead to failure...but 200# and 3 years of riding
and the wheel is still true.

According to: http://www.geocities.com/spokeanwheel/lacingsr.htm

"A multi-speed rear wheel that is half radially laced, with the non
drive side being the side with the radial lacing, can only have
increased imbalance in the lateral strengths of the two sides of the
wheel. The radial spokes on the non drive side will also need a lower
build tension when compared to a standard wheel. Since radial lacing
has the highest lateral strength, the spokes do not need to be as
tight to counter the lateral forces from the drive side in order to
achieve a true wheel. Also, with the non drive side spokes even looser
than normal, the likelihood that the wheel will go out of true from
road shock is increased. "

and...

"A properly built half radial rear wheel will be radially laced on the
drive side. It would also be best to lace the drive side so the spoke
heads are on the inside of the flange. "

Sheldon's article talks about the "right" side in his discussion about
half-radial lacing...which I infer to be the drive side. He also
refers to "wrong-way half radial". So these two sites are in complete
contradiction to one another. Any comments?

David Bonnell said:

So these two sites are in complete
contradiction to one another. Any comments?

The site you linked to has many strange ideas that aren't supported by
physics. I have no idea what "imbalance in lateral strength" could
mean in reality... stiffness is the same from either side, although
spokes will go slack sooner with a force towards the looser spokes.
The most important contributor to lateral stiffness is the bacing
angle of the NDS spokes... but the tension of these spokes must also
be considered... since the greater the bracing angle, the looser they
will be... and the looser they are, then the more likely they will go
slack from radial loads (which are generally the highest). So... we
are trying to achieve "high enough" tension on the NDS, and "high
enough" lateral stability... and these work against each other. On the
DS we are basically stuck with a flange spacing of ~19mm from the
centerline, because this is the most we can get with a 130mm dropout
while leaving room for the 8,9,10 spd cassette. If we space the NDS
38mm over then the ratio of NDS/DS tension will be 1/2 (assuming the
pattern is the same on both sides). On some hubs the NDS spacing is
only 30mm (AC comes to mind). This helps to balance the tension but
compromises the lateral stiffness a lot. I think something in the
35-38mm range is the best compromise. For shallow rims, which are
generally lighter and cannot take high tension, an OC rear is a good
idea. This gives a greater bracing angle to the DS, which allows a
higher NDS tension, and provided that the flange spacing is fairly
high, the lateral stiffness will still be good. Deeper and heavier
rims can be safely built with higher tensions, so this is less of an
issue for them.

Heads in radial on the drive side allows for a *slightly* greater
spacing there (maybe 20-21mm), which will balance the tension a little
better. But don't try this with a normal J bend hub... they are not
designed for this type of lacing (combined with high DS tension) and
the flange will fail. That is the main thing that is wrong with the
author's recommendation. The other is that having the loose NDS spokes
take all the torque load means they will likely go slack when radial
and torque loads are high... *if* the hub has a normal small flange on
that side. Also since normal hubs aren't designed to transfer torque
loads through hub body, the hub body can also twist and fail. Some
hubs (like your Shimano) use straight pull spokes with radial lacing
on the DS. There is nothing wrong with this provided that the NDS
flange is large... which minimizes the tension change due to torque
loads... and of course the hub body and flanges must be designed for
the loads that they will see.

Another variation on the theme is triplet lacing (Campy G3 is one
example)... where there are twice as many spokes on the DS compared
to the NDS. This effectively doubles the tension in the NDS spokes.
I've built several wheels with this pattern and it works well, but to
get the maximum benefit from it you need a hub with ~45mm NDS
spacing... or in other words a hub specially made for this pattern.
Normal hubs loose some lateral stiffness, but if the NDS spacing is
high you can have very good lateral stiffness and still have high NDS
tension. The down side is that if you break a NDS spoke, the wheel
goes *way* out of true.