radial drive wind-up

steve said:

I would like to question the theory that a radial drive rear wheel
increases spoke tension significantly assuming the non-drive is laced
in a 2x or 3x pattern and your hub body is reasonably stiff. In fact I
would argue that given these conditions the non-drive spokes see a
greater tension change than the drive spokes during tortional loads. I
have done tests with shimano 7800 dura-ace and ksrium ssc sl wheels
both which use radial drive. Using a dt tension meter I had a large
person apply the rim brake to the rear wheel and crank down as hard as
they could on the pedal to induce a large tortional load. When they
did this I measured the change in spoke tension of drive and non-drive
pull spokes. The change in the drive was so small it could have easily
been error due to hand pressure. On the non-drive pull spokes I found
tensions varying 20-30kg. To me this seems to be a large difference and
it makes me think that there must be some other reason besides large
tension changes for premature spoke fatigue on a radial drive.
I also think that since the tensions of the spokes are so high hub
rotation will not cause the bedded spoke to rotate in the hub flange.
There for it seems that the only cause of premature failure would be
the constant bending and unbending of the spoke right after it leaves
the hub flange from hub rotation. This is obviously very amatuer
thinking and would appreciate any words of wisdom that you might have
to set me straight.

Great idea to actually measure the tension; thanks for posting your
observations!

As you theorize, certainly a stiff hub shell helps transfer the torque
to the crossed left side spokes. The existence of wheels like the ones
you checked shows they "work". In wheels like that, the "left side"
spokes *are* the "drive side" spokes -- they do the driving, not the
right side spokes.

Do such wheels even suffer premature spoke failure? I don't know -- I'm
asking because I haven't heard of any premature spoke failures in
factory wheels that have radial right-side spokes.

Quoted message said:

steve said:

I would like to question the theory that a radial drive rear wheel
increases spoke tension significantly assuming the non-drive is laced
in a 2x or 3x pattern and your hub body is reasonably stiff. In fact I
would argue that given these conditions the non-drive spokes see a
greater tension change than the drive spokes during tortional loads. I
have done tests with shimano 7800 dura-ace and ksrium ssc sl wheels
both which use radial drive. Using a dt tension meter I had a large
person apply the rim brake to the rear wheel and crank down as hard as
they could on the pedal to induce a large tortional load. When they
did this I measured the change in spoke tension of drive and non-drive
pull spokes. The change in the drive was so small it could have easily
been error due to hand pressure. On the non-drive pull spokes I found
tensions varying 20-30kg. To me this seems to be a large difference and
it makes me think that there must be some other reason besides large
tension changes for premature spoke fatigue on a radial drive.
I also think that since the tensions of the spokes are so high hub
rotation will not cause the bedded spoke to rotate in the hub flange.
There for it seems that the only cause of premature failure would be
the constant bending and unbending of the spoke right after it leaves
the hub flange from hub rotation. This is obviously very amatuer
thinking and would appreciate any words of wisdom that you might have
to set me straight.

Great idea to actually measure the tension; thanks for posting your
observations!

As you theorize, certainly a stiff hub shell helps transfer the torque
to the crossed left side spokes. The existence of wheels like the ones
you checked shows they "work". In wheels like that, the "left side"
spokes *are* the "drive side" spokes -- they do the driving, not the
right side spokes.

Do such wheels even suffer premature spoke failure? I don't know -- I'm
asking because I haven't heard of any premature spoke failures in
factory wheels that have radial right-side spokes.

Steve and Di -

Interesting topic, somewhat near and dear to my heart.

I have some experience with the Shimano wheels - the RD5XX variety.

These I bought as an experiment because they were uniquely enough
constructed to pique my curiosity. I fully expected them to fail on me,
as I'm relatively large and have a reputation for exerting considerable
forces on stuff. I nick-named them the "bumble-bee" wheels, because
there was no way they could actually work - like bumble-bees can't fly.

These wheels are radially laced on the right side and crossed on the
left (due to my experience with these wheels I have abandoned the
terminology of "drive side" and "non-drive side"😉.

They also have the spoke heads at the rim, and are reduced spoke count
wheels. There were times when they felt a bit soft, which I attribute to
the reduced spoke count.

The placement of the spoke head at the rim seems like a Good Idea. Most
of the movement that a spoke sees is at the hub/head junction. Putting a
large nipple there in a good seat certainly seems like a stronger
engineering solution, as the weakest portion of the spoke is in a place
where it doesn't see as much flex or movement. The spokes are
conventional (but extremely long and bladed), but require a special
washer at the rim. Obviously the rims are non-standard, having the spoke
hole in the side of the semi-aero rim as are the hubs. These are not
wheels one can really rebuild, with the exception of spoke replacement.

I rode these wheels for about 6000 miles with absolutely no issues with
spoke breakage. In fact, the wheels stayed true the entire time. They
were among the best wheels I have ever had over the past 35 years. The
took a lickin' and kept on tickin'.

I wish I could continue the experiment with the wheels, but the bike was
stolen last weekend.

Based upon my experiences with these wheels I decided to build a new set
of wheels for my commuter bike. These are 36 spoke wheels, but the right
side is laced radially and the left side is three cross.

My commute is fairly rough, as I have occasion to ride down stairs, down
the middle of railroad tracks (it's easier at speed!) and through
unpaved areas. I usually have the panniers loaded pretty well, as well
as my 200+ pound lard ass on board. I run a Continental 700X37 tire. I
have had some problems with spoke breakage in the past on this bike in
this usage.

So far they have a thousand miles or so on them and have worked well.
They have remained true and there has been no spoke damage (which I
would not expect in any wheel with only 1000 mile son it).

Part of the thought process that lead me to do this is that, with
today's drive trains, rear wheels have a severe dish. If the right side
is the drive side, then those spokes need to provide most of the motive
torque as well as holding the dish. If they are radially laced, then
they cannot really transmit any torque and they are left with the job of
holding the dish, while the left side spokes provide the driving torque.

Admittedly, I'm running on intuition and not FEA....

For now I will say that in my usage, on my bike, with me riding it, they
work fine. Let's get together in a couple of years and I'll tell you how
they have fared over 10,000 miles ;^) .

Cheers,

Tom

Tom Schmitz said:

I have some experience with the Shimano wheels - the RD5XX variety.
There were times when they felt a bit soft, which I attribute to
the reduced spoke count.

They are about the flexiest wheels ever made. The fact that you had no
trouble shows how little flex matters...

Quoted message said:

Part of the thought process that lead me to do this is that, with
today's drive trains, rear wheels have a severe dish. If the right side
is the drive side, then those spokes need to provide most of the motive
torque as well as holding the dish. If they are radially laced, then
they cannot really transmit any torque and they are left with the job of
holding the dish, while the left side spokes provide the driving torque.

This is a solution to a "problem" that doesn't exist though, for a
wheel that is well built. The increased tension from driving torque is
small and will still not exceed the spoke's capabilities.

There are some issues with using radial spokes on the right. The right
hub flange needs to be quite strong to take radial spokes with such
high tension... especially with 36 spokes. Also your hub needs to
relatively beefy in the center to transfer the torque without yielding.
And the crossed spokes will be more likely to go slack during torque +
bumps because they are already rather loose (on a normal hub anyway).
This will result in fatigue and failure of the spokes eventually. If
the left flange is moved in to balance the tension better, this isn't
an issue... but of course that increases lateral flex, too.

One thing that radial on the drive side (of the bike, not the hub) can
give you is a better bracing angle to the rim... if you lace heads in.
You may not have enough derailleur clearance though. Because of this
one advantage, it is an ok idea for a manufacturer to design and build
a wheel this way... but be careful when you do it yourself and use
parts that *weren't* designed for it.

Ron Ruff said:

Tom Schmitz said:

I have some experience with the Shimano wheels - the RD5XX variety.
There were times when they felt a bit soft, which I attribute to
the reduced spoke count.

They are about the flexiest wheels ever made. The fact that you had no
trouble shows how little flex matters...

My sentiments as well. I noticed it, but was not bothered by it.

Quoted message said:

Quoted message said:

Part of the thought process that lead me to do this is that, with
today's drive trains, rear wheels have a severe dish. If the right side
is the drive side, then those spokes need to provide most of the motive
torque as well as holding the dish. If they are radially laced, then
they cannot really transmit any torque and they are left with the job of
holding the dish, while the left side spokes provide the driving torque.

This is a solution to a "problem" that doesn't exist though, for a
wheel that is well built. The increased tension from driving torque is
small and will still not exceed the spoke's capabilities.

Hmmm.. Not sure I agree 100% with that. "A wheel that is well built" is
somewhat nebulous, and I might argue that a well built wheel is a bit at
the ragged edge of performance with severely dished designs. Relieving
the right side of some duty creates an additional margin of strength. Of
course, lacing the right side radially places the strain vector through
a smaller section on the hub. It all could be [censored]-for-tat.

Alas, this is a bit of an argument similar to how many angels can dance
on the head of a pin. There are so many variables and intended usages to
cope with. You are correct in that currently designed conventional
wheels are strong enough as-is.

Quoted message said:

There are some issues with using radial spokes on the right. The right
hub flange needs to be quite strong to take radial spokes with such
high tension... especially with 36 spokes. Also your hub needs to
relatively beefy in the center to transfer the torque without yielding.

Yes to all of that. The hub I'm using is similar in design to a Shimano
Parallax, which has a rather beefy center. I'm not aware of any
conventional rear hub, however, that is designed for radial build and
this one is no different in that respect. Whether or not it will
withstand such treatment only time will tell. This is, after all, an
experiment (I'm easily amused).

Quoted message said:

And the crossed spokes will be more likely to go slack during torque +
bumps because they are already rather loose (on a normal hub anyway).
This will result in fatigue and failure of the spokes eventually. If
the left flange is moved in to balance the tension better, this isn't
an issue... but of course that increases lateral flex, too.

I have not noticed any strain-induced slackness; time will tell on the
fatigue aspect.

Quoted message said:

One thing that radial on the drive side (of the bike, not the hub) can
give you is a better bracing angle to the rim... if you lace heads in.
You may not have enough derailleur clearance though. Because of this
one advantage, it is an ok idea for a manufacturer to design and build
a wheel this way... but be careful when you do it yourself and use
parts that *weren't* designed for it.

I did build "heads in" for that very reason. I didn't do the math, but
it must be at least .005% better than real life.... :^)

Cheers,

Tom

Ron Ruff said:

Quoted message said:

This is a solution to a "problem" that doesn't exist though, for a

wheel that is well built. The increased tension from driving torque is
small and will still not exceed the spoke's capabilities.

There are some issues with using radial spokes on the right. The right
hub flange needs to be quite strong to take radial spokes with such
high tension... especially with 36 spokes. Also your hub needs to
relatively beefy in the center to transfer the torque without yielding.
And the crossed spokes will be more likely to go slack during torque +
bumps because they are already rather loose (on a normal hub anyway).
This will result in fatigue and failure of the spokes eventually. If
the left flange is moved in to balance the tension better, this isn't
an issue... but of course that increases lateral flex, too.

One thing that radial on the drive side (of the bike, not the hub) can
give you is a better bracing angle to the rim... if you lace heads in.
You may not have enough derailleur clearance though. Because of this
one advantage, it is an ok idea for a manufacturer to design and build
a wheel this way... but be careful when you do it yourself and use
parts that *weren't* designed for it.

So from what I understand you are saying the largest issues with radial
lacing on the drive side is hub body and flange strength as well as
higher fatigue rates for "push" spokes on the non-drive side. Lets say
we lace a wheel up to a philwood hub which does allow radial lacing and
has a large diameter hub body. If the drive side was radially laced
and the non-drive was 3x would you be more concerned with the non-drive
spokes failing prematurely or the drive side? As you said earlier, we
must consider that tensions are more even between drive and non-drive
since heads in radial lacing gives a greater bracing angle.
Steve

steve said:

Lets say
we lace a wheel up to a philwood hub which does allow radial lacing and
has a large diameter hub body. If the drive side was radially laced
and the non-drive was 3x would you be more concerned with the non-drive
spokes failing prematurely or the drive side? As you said earlier, we
must consider that tensions are more even between drive and non-drive
since heads in radial lacing gives a greater bracing angle.

Sheldon Brown postulates that in a normal wheel (3x both sides) it is
not uncommon for "pushing" spokes on the NDS to go slack and fatigue.
He recommends radial lacing on the NDS for this reason. By making the
DS radial so the NDS has to transfer all the torque, it will be even
more likely that the "pushing" spokes go slack. So... I'd be more
concerned about the NDS... the radial spokes on the DS should be
fine... so long as the hub is ok.

If you are curious you might look at the bracing angles and what the
resulting ratios in spoke tensions would be for the two cases. An
offset rim (like an Aerohead OC) would also help.

BTW steve, when you found tensions varying 20-30kg, that was between
the lowest and highest... yes? So a single spoke would see a 10-15kg
tension change from it's unloaded state?

On 10 Sep 2006 21:27:06 -0700, "Ron Ruff" <[email hidden]>

Quoted message said:

steve said:

Lets say
we lace a wheel up to a philwood hub which does allow radial lacing and
has a large diameter hub body. If the drive side was radially laced
and the non-drive was 3x would you be more concerned with the non-drive
spokes failing prematurely or the drive side? As you said earlier, we
must consider that tensions are more even between drive and non-drive
since heads in radial lacing gives a greater bracing angle.

Sheldon Brown postulates that in a normal wheel (3x both sides) it is
not uncommon for "pushing" spokes on the NDS to go slack and fatigue.
He recommends radial lacing on the NDS for this reason. By making the
DS radial so the NDS has to transfer all the torque, it will be even
more likely that the "pushing" spokes go slack. So... I'd be more
concerned about the NDS... the radial spokes on the DS should be
fine... so long as the hub is ok.

If you are curious you might look at the bracing angles and what the
resulting ratios in spoke tensions would be for the two cases. An
offset rim (like an Aerohead OC) would also help.

Dear Steve and Ron,

Browse down to "Half-radial Spoking" and then "Wrong-way Half Radial":

http://sheldonbrown.com/wheelbuild.html#radial

Cheers,

Carl Fogel

Ron Ruff said:

BTW steve, when you found tensions varying 20-30kg, that was between
the lowest and highest... yes? So a single spoke would see a 10-15kg
tension change from it's unloaded state?

Yes, a single pull spoke would increase 10-15kg after a torque was
applied. I understand that this might not be a huge amount but it is
still a huge difference compared to the drive side tension changes
according to this test. So my question still stands as to why most
people think that radial drive will cause drive side spokes to
prematurely fail.

Quoted message said:

On 10 Sep 2006 21:27:06 -0700, "Ron Ruff" <[email hidden]>
wrote:

Dear Steve and Ron,

Browse down to "Half-radial Spoking" and then "Wrong-way Half Radial":

http://sheldonbrown.com/wheelbuild.html#radial

Cheers,

Carl Fogel

Carl,
I did browse "Half-radial Spoking" and the "Wrong-way Half Radial" but
it doesn't seem to explain anything new. I think that some companies
might have laced the drive side radially for the reasons Sheldon brown
states but, I think it is quite clear that heads in radial lacing on
the drive side does provide a better bracing angle. This being the
case I find most people disaprove of radial drive lacing not because of
the fatigue of non-drive spokes but because of premature fatigue on the
drive side. In fact I believe even Joe Brandt stated in his book that
a radial dive lacing would cause substantial increase in tension on the
drive side from tortional loads causing excellerated fatigue, though
please correct me if I am wrong since I do not have the text in front
of me and could be mis-interpreting what he wrote. This is what I am
arguing, that the drive side of a radial laced spoke experiences very
little tension change through tortional loads. So, if we were able to
tension the non-drive crossed spokes to the point were the "push"
spokes would not completly detension what would be the problem in
biulding a wheel like this assuming the hub is capable of handling this
lacing pattern.

Steve

steve said:

Quoted message said:

On 10 Sep 2006 21:27:06 -0700, "Ron Ruff" <[email hidden]>
wrote:

Dear Steve and Ron,

Browse down to "Half-radial Spoking" and then "Wrong-way Half Radial":

http://sheldonbrown.com/wheelbuild.html#radial

Cheers,

Carl Fogel

Carl,
I did browse "Half-radial Spoking" and the "Wrong-way Half Radial" but
it doesn't seem to explain anything new. I think that some companies
might have laced the drive side radially for the reasons Sheldon brown
states but, I think it is quite clear that heads in radial lacing on
the drive side does provide a better bracing angle. This being the
case I find most people disaprove of radial drive lacing not because of
the fatigue of non-drive spokes but because of premature fatigue on the
drive side. In fact I believe even Joe Brandt stated in his book that
a radial dive lacing would cause substantial increase in tension on the
drive side from tortional loads causing excellerated fatigue, though
please correct me if I am wrong since I do not have the text in front
of me and could be mis-interpreting what he wrote. This is what I am
arguing, that the drive side of a radial laced spoke experiences very
little tension change through tortional loads. So, if we were able to
tension the non-drive crossed spokes to the point were the "push"
spokes would not completly detension what would be the problem in
biulding a wheel like this assuming the hub is capable of handling this
lacing pattern.

Steve

Dear Steve,

Here are Jobst's comments on radial spoking:

Radial spokes carry loads just as well as crossed spokes, but they
cannot transmit torque. They transmit torque only after the hub
rotates ahead of the rim, making the spokes no longer truly radial.
This rotation produces a small tangential offset, or lever, on which
spoke tension can act to produce torque. This lever is the distance
between the as of the hub and the extended as of the no-longer-radial
spokes. The driving torque is the product of this small offset and the
tension of all the spokes.

In a radial rear wheel the windup that occurs while riding is small
(less than two degrees). However, this motion increases spoke fatigue,
and spoke rotation in the flange causes wear. As radial spokes wind up
under torque, they become appreciably tighter causing high rim stress
and, in some instances, flange or rim failure. Looser spoking would
reduce windup induced tension, but it would also reduce wheel
strength.

Even though they transmit no torque, front wheels should not be spoked
radially because high radial stress can cause fatigue failure of their
flanges. The spoke holes of aluminum alloy hubs can break out causing
wheel collapse. Flange fatigue takes time, so these failures do not
occur immediately. Some lightweight hubs carry specific warnings
against radial spoking.

--"The Bicycle Wheel," 3rd edition, p. 67

I'm not sure that I follow which spokes you plan to tension how much,
but I suspect that any scheme that relies on radically changing the
tension on one side without affecting the other will fail.

Cheers,

Carl Fogel

Ron Ruff said:

Quoted message said:

Lets say we lace a wheel up to a Phil Wood hub which does allow
radial lacing and has a large diameter hub body. If the drive side
was radially laced and the non-drive was 3x would you be more
concerned with the non-drive spokes failing prematurely or the
drive side? As you said earlier, we must consider that tensions
are more even between drive and non-drive since heads in radial
lacing gives a greater bracing angle.

Quoted message said:

Sheldon Brown postulates that in a normal wheel (3x both sides) it
is not uncommon for "pushing" spokes on the NDS to go slack and
fatigue. He recommends radial lacing on the NDS for this reason. By
making the DS radial so the NDS has to transfer all the torque, it
will be even more likely that the "pushing" spokes go
slack. So... I'd be more concerned about the NDS... the radial
spokes on the DS should be fine... so long as the hub is OK.

Quoted message said:

If you are curious you might look at the bracing angles and what the
resulting ratios in spoke tensions would be for the two cases. An
offset rim (like an Aerohead OC) would also help.

If you look at the graphs in "the Bicycle Wheel" you can assure
yourself that this will not occur, the torque load shown there is the
same as the radial load and the braking load, just to make the
relative effects more comparable. In effect, the wheel shown is
climbing a vertical wall.

Jobst Brandt

Quoted message said:

Here are Jobst's comments on radial spoking:
Even though they transmit no torque, front wheels should not be spoked
radially because high radial stress can cause fatigue failure of their
flanges. The spoke holes of aluminum alloy hubs can break out causing
wheel collapse. Flange fatigue takes time, so these failures do not
occur immediately. Some lightweight hubs carry specific warnings
against radial spoking.

--"The Bicycle Wheel," 3rd edition, p. 67

Don't the spokes on a front wheel transmit torque under braking?

Cam

Cam said:

Quoted message said:

Here are Jobst's comments on radial spoking:
Even though they transmit no torque, front wheels should not be spoked
radially because high radial stress can cause fatigue failure of their
flanges. The spoke holes of aluminum alloy hubs can break out causing
wheel collapse. Flange fatigue takes time, so these failures do not
occur immediately. Some lightweight hubs carry specific warnings
against radial spoking.

--"The Bicycle Wheel," 3rd edition, p. 67

Don't the spokes on a front wheel transmit torque under braking?

This depends on where the braking surface is located relative to the
hub.

Cam said:

Don't the spokes on a front wheel transmit torque under braking?

No, since the brake acts directly on the rim rather than the hub.

Quoted message said:

If you look at the graphs in "the Bicycle Wheel"

Gotta get that book someday...

steve said:

Joe Brandt stated in his book that
a radial dive lacing would cause substantial increase in tension on the
drive side from tortional loads causing excellerated fatigue

He is refering to having an entire rear wheel radial laced. In that
case the hub/rim would wind up and increase fatigue. In a half radial
wheel the crossed spokes carry the torque with very little
displacement, so the radial spokes are fine.

Quoted message said:

Quoted message said:

If you look at the graphs in "the Bicycle Wheel" you can assure
yourself that this will not occur, the torque load shown there is the
same as the radial load and the braking load, just to make the
relative effects more comparable. In effect, the wheel shown is
climbing a vertical wall.

Jobst Brandt

Jobst,
If you don't mind, I would like to comment on your statement that a
rear wheel, which is radially laced, increases spoke tensions
significantly when a tortional load is applied. Using my amateur math,
it seems that spoke tensions would not change a significant amount. As
I understand it, when a tortional load is applied to a wheel the
tension changes in the spoke as a result of the hub flange rotating a
certain amount there for increasing the distance between the flange
hole and the rim which stretches the spoke and causes it to increase in
tension. In your book I believe you stated that the wind-up of a rear
wheel that is radially laced would be less than two degrees. If this
is the case it seems to me that the distance between the flange hole
and the rim would change very little which should result in a
relatively small tension increase.

Thank you for your time and comments

Steve

Steve Sauter said:

Quoted message said:

If you look at the graphs in "the Bicycle Wheel" you can assure
yourself that this will not occur, the torque load shown there is
the same as the radial load and the braking load, just to make the
relative effects more comparable. In effect, the wheel shown is
climbing a vertical wall.

Quoted message said:

If you don't mind, I would like to comment on your statement that a
rear wheel, which is radially laced, increases spoke tensions
significantly when a tortional load is applied. Using my amateur
math, it seems that spoke tensions would not change a significant
amount. As I understand it, when a tortional load is applied to a
wheel the tension changes in the spoke as a result of the hub flange
rotating a certain amount there for increasing the distance between
the flange hole and the rim which stretches the spoke and causes it
to increase in tension. In your book I believe you stated that the
wind-up of a rear wheel that is radially laced would be less than
two degrees. If this is the case it seems to me that the distance
between the flange hole and the rim would change very little which
should result in a relatively small tension increase.

Quoted message said:

Thank you for your time and comments

Torsional windup of two degrees has a cosine effect of 0.1, and that
times a flange radius of 22mm gives 2.2mm elongation over the initial
stress of spoke tension. That is more than double the tension and
cyclically would rip the spoke out of the flange radially in a short
time. Even at x3 spokes tear out of rear hubs. Beyond this, the
spokes and nipples would also be under a greater cyclic stress than
cross laced spoking causes.

As you see, broken spokes are discussed here off and on, showing that
they occur under apparently far better conditions than on radially
spoked drive wheels. You can spoke your wheels any way you like but
don't encourage others do choose your method.

Jobst Brandt