How much does tension rise on squeezed spokes?

[Joe Riel]

[email protected] writes:

> Both spokes are on the non-drive side, a parallel pair, not
> one on the left and one on the right.

I thought the plane of the wheel was parallel to the floor. Didn't
you say that one spokes was above the other? I'm confused. However,
if the most of the tension is going into one spoke, then its paired
spoke should see a decrease, since the rim will be radially deflected
by the tensioned spoke. My previous conjecture about the rim
deflecting laterally may not apply.

--
Joe Riel

 

[[email protected]]

[email protected] wrote:

> Some posts have led me to wonder whether my powers of
> description are ever feebler than I thought, so here's a
> crude diagram of the setup:
>
> http://home.comcast.net/~carlfogel/download/biketest.jpg

> The question is whether the weight hanging down is going to
> give a strikingly different result than a single hand
> squeeze.
>
> I can use another pair of spokes, or just a single spoke for
> a comparison, if someone has an idea. Joe Riel has suggested
> supporting the rim (which has a 110 psi tire on it), but I'm
> not sure where the support would go.

Carl,

Tie the rope to some independent anchor point, not the
second spoke. Then you know you're only loading one
spoke. It isn't what is done in stress relieving but it
eliminates a variable (the unmeasured effect of friction
in the rope, which decreases the side load on the
lower spoke).

I'll guess that you'll see that the upper spoke increase
in tension and the lower spoke decrease, because the
upper spoke brings the rim inward.

 

[Marvin]

[email protected] wrote:
> On Sat, 06 May 2006 15:49:51 -0600, [email protected]
> wrote:
>
> >Elsewhere we've wandered off into how to calculate and test
> >the rise in tension when a spoke is squeezed. That thread
> >(as usual) is getting a bit large and awkward for people to
> >follow, so here's a new thread.
> >
> >Joe Riel has a page with calculations indicating that rim
> >stiffness plays a large role in the rise in tension when a
> >tensioned spoke is displaced sideways:
> >
> >www.k-online.com/~joer/cycling/spoke-tension.pdf
> >
> >Joe recently commented:
> >
> >"I think it's pretty clear from the final graph that there
> >is a large variation in the final tension, dependent upon
> >the rim stiffness---that was the point of the article."
> >
> >Testing the rise in tension is a little trickier than you'd
> >expect. The gauge is so large that it gets in the way, you
> >could use another pair of hands, it's hard to do two spokes
> >on one side with your left hand and two on the other with
> >your right for balance, and everything will vary with the
> >next tester--initial tension, strength of squeeze, kind of
> >spoke, stiffness of rim.
> >
> >But I took a stab at it with an old 36-spoke 700c rim,
> >straight 2mm spokes, and probably less initial tension than
> >it ought to have.
> >
> >Unsqueezed, the spoke registered 15 on my Park gauge, which
> >is probably around a pathetic 41 kgf (the Park conversion
> >chart doesn't go below 17 for such spokes.)
> >
> >Squeezed as hard as I could manage with a rope, the spoke
> >tension increased to 20 on the Park gauge, 68 kgf.
> >
> >(In both measurements, I squeezed and released the gauge
> >three times--without that, it would have stayed on 15
> >instead of rising to 20.)
> >
> >I think that this roughly 30 kgf (65-lb) tension increase is
> >far less than most people expect from spoke squeezing, but I
> >hasten to add that a greater increase might be obtained with
> >better techniques--which is what this post is meant to
> >provoke.
> >
> >A stiffer rim, hauling harder on the rope, more initial
> >tension, and perhaps a more sensitive gauge might improve
> >the results.
> >
> >The rim is plain and not boxed, so it may not be very stiff.
> >
> >The rope squeeze may not be as good as one hand, even though
> >it looks pretty good. The rope is right in the midspan of
> >the spoke, but the area squeezed is much smaller than a
> >hand, which might affect leverage. (With the gauge on the
> >spoke, I can't get my whole hand in to squeeze.)
> >
> >The initial tension was embarrassingly low, which could
> >confuse the rim stiffness and other factors.
> >
> >The Park gauge is said to use a lot of force itself (you can
> >see the bending within the gauge in the picture) and might
> >confuse things).
> >
> >Here's a Park tension gauge showing 15 on the target spoke:
> >http://www.filelodge.com/files/room19/497501/spoke1.jpg
> >
> >Here it shows 20 with the spoke squeezed with a rope:
> >http://www.filelodge.com/files/room19/497501/spoke2.jpg
> >
> >A blurry picture of gauge and rope setup:
> >http://www.filelodge.com/files/room19/497501/spoke3.jpg
> >
> >Better focus of gauge and rope setup:
> >http://www.filelodge.com/files/room19/497501/spoke4.jpg
> >
> >Closer view of gauge and rope, showing bending of spoke:
> >http://www.filelodge.com/files/room19/497501/spoke5.jpg
> >
> >Here's Park's 0-50 scale conversion chart for 2mm straight
> >round steel spokes:
> >
> >scale_marking
> > Park_kgf
> > estimate_kgf
> >13 --
> >14 --
> >15 -- (41?) <--initial tension
> >16 -- (46?)
> >17 51
> >18 56
> >19 62
> >20 68 <--squeezed tension
> >21 76
> >22 85
> >23 95 <--recommended tension for most wheels
> >24 107
> >25 121
> >26 137
> >27 156
> >28 178
> >29 ---
> >30 ---
> >
> >I hope that this provokes pictures from people with better
> >cameras, wheels, grips, and testing techniques. If nothing
> >else, I learned that the rope will stay in place, unknotted,
> >under heavy tension because the spokes are so thin that the
> >loop jams solid.
> >
> >If anyone lacking a gauge can suggest a different way to do
> >it, I'll give it a try, with the obvious warning that my
> >experimental skills are already pushing their limits.
> >
> >Cheers,
> >
> >Carl Fogel
>
> Some posts have led me to wonder whether my powers of
> description are ever feebler than I thought, so here's a
> crude diagram of the setup:
>
> http://home.comcast.net/~carlfogel/download/biketest.jpg

Hey, I understood what was going on.

The issue with the tension in the lower spoke is because the majority
of the weight is actually being taken by the upper spoke. It's
interesting that the rim deflection is sufficient to actually detension
the lower spoke, but that's not the real problem.

Two problems spring to mind:

1: the knotting lower loop of rope. If that jams up you've effectively
tied the two spokes together and any further weight then pulls down
solely on the upper spoke. This is where friction is causing you the
most trouble.

2: you're setting up what looks like a classic pulley system, so the
force on the upper spoke will be double that applied to the top spoke.
It's tricky to calculate the forces on the lower spoke with that
knotting arrangement, but I reckon that would explain what you're
seeing.

If the aim of the game here is to test Jobst's spoke squeezing
technique you really need to squeeze spokes on both sides of the rim at
once. Squeezing one side will introduce rim deflections that wouldn't
be there (or at least wouldn't be anywhere near as bad) if you squeezed
spokes on both sides.

Now what I can't do is suggest a good solution to the problem. If you
had a spring that could clip onto both spokes and apply a reasonable
tension that would be ideal, but heaven help your fingers trying to
attach a spring with 30lb of tension.

Is it possible that you have a couple of toe straps in your garage?
Hanging a weight off a toestrap might not be easy but once the weight
was removed you'd be sure the tensions on the two spokes were even, and
you could squeeze spokes on the other side as well.

I'd also suggest sticking a bit more tension onto that wheel if you
think it'll possibly take it. Being off the bottom of the Park scale
cannot be helping the accuracy of your measurements

 

[Marvin]

Marvin wrote:
> [email protected] wrote:
> > On Sat, 06 May 2006 15:49:51 -0600, [email protected]
> > wrote:
> >

<snipped>

> >
> > Some posts have led me to wonder whether my powers of
> > description are ever feebler than I thought, so here's a
> > crude diagram of the setup:
> >
> > http://home.comcast.net/~carlfogel/download/biketest.jpg
>
> Hey, I understood what was going on.

....actually it seems that I didn't, or rather, got the two experiments
confused. I understand both of them now though...

> The issue with the tension in the lower spoke is because the majority
> of the weight is actually being taken by the upper spoke. It's
> interesting that the rim deflection is sufficient to actually detension
> the lower spoke, but that's not the real problem.
>
> Two problems spring to mind:
>
> 1: the knotting lower loop of rope. If that jams up you've effectively
> tied the two spokes together and any further weight then pulls down
> solely on the upper spoke. This is where friction is causing you the
> most trouble.
>
> 2: you're setting up what looks like a classic pulley system, so the
> force on the upper spoke will be double that applied to the top spoke.
> It's tricky to calculate the forces on the lower spoke with that
> knotting arrangement, but I reckon that would explain what you're
> seeing.

Sorry, I'm a muppet, I didn't see you'd readjusted the experiment.

In which case, you've definitely got a pulley setup going so the force
applied to the upper spoke is higher than the lower. Between rope
friction and rim displacement it's not surprising that you're seeing an
actual drop in spoke tension.

Your original idea was better for avoiding these issues, but not as
good for getting an accurate measurement of applied force.

What might be better on both counts is to tie a rope to the ceiling,
loop it around your pair of spokes and then dangle a weight from the
bottom. You'll be applying double the rope tension to each spoke, give
or take frictional losses, but at least those losses should be the same
on both spokes.

You'll also be able to do away with the brakes which ISTR were
concerning you.

> If the aim of the game here is to test Jobst's spoke squeezing
> technique you really need to squeeze spokes on both sides of the rim at
> once. Squeezing one side will introduce rim deflections that wouldn't
> be there (or at least wouldn't be anywhere near as bad) if you squeezed
> spokes on both sides.
>
> Now what I can't do is suggest a good solution to the problem. If you
> had a spring that could clip onto both spokes and apply a reasonable
> tension that would be ideal, but heaven help your fingers trying to
> attach a spring with 30lb of tension.
>
> Is it possible that you have a couple of toe straps in your garage?
> Hanging a weight off a toestrap might not be easy but once the weight
> was removed you'd be sure the tensions on the two spokes were even, and
> you could squeeze spokes on the other side as well.
>
> I'd also suggest sticking a bit more tension onto that wheel if you
> think it'll possibly take it. Being off the bottom of the Park scale
> cannot be helping the accuracy of your measurements

 

[Peter Cole]

[email protected] wrote:
>
> Some posts have led me to wonder whether my powers of
> description are ever feebler than I thought, so here's a
> crude diagram of the setup:
>
> http://home.comcast.net/~carlfogel/download/biketest.jpg

If you squeeze a single pair of spokes on one side of a wheel, you'll
notice that the rim deflects laterally quite readily, so it's not clear
exactly what you're measuring.

 

[jim beam]

[email protected] wrote:
> On Sat, 06 May 2006 15:49:51 -0600, [email protected]
> wrote:
>
>
>>Elsewhere we've wandered off into how to calculate and test
>>the rise in tension when a spoke is squeezed. That thread
>>(as usual) is getting a bit large and awkward for people to
>>follow, so here's a new thread.
>>
>>Joe Riel has a page with calculations indicating that rim
>>stiffness plays a large role in the rise in tension when a
>>tensioned spoke is displaced sideways:
>>
>>www.k-online.com/~joer/cycling/spoke-tension.pdf
>>
>>Joe recently commented:
>>
>>"I think it's pretty clear from the final graph that there
>>is a large variation in the final tension, dependent upon
>>the rim stiffness---that was the point of the article."
>>
>>Testing the rise in tension is a little trickier than you'd
>>expect. The gauge is so large that it gets in the way, you
>>could use another pair of hands, it's hard to do two spokes
>>on one side with your left hand and two on the other with
>>your right for balance, and everything will vary with the
>>next tester--initial tension, strength of squeeze, kind of
>>spoke, stiffness of rim.
>>
>>But I took a stab at it with an old 36-spoke 700c rim,
>>straight 2mm spokes, and probably less initial tension than
>>it ought to have.
>>
>>Unsqueezed, the spoke registered 15 on my Park gauge, which
>>is probably around a pathetic 41 kgf (the Park conversion
>>chart doesn't go below 17 for such spokes.)
>>
>>Squeezed as hard as I could manage with a rope, the spoke
>>tension increased to 20 on the Park gauge, 68 kgf.
>>
>>(In both measurements, I squeezed and released the gauge
>>three times--without that, it would have stayed on 15
>>instead of rising to 20.)
>>
>>I think that this roughly 30 kgf (65-lb) tension increase is
>>far less than most people expect from spoke squeezing, but I
>>hasten to add that a greater increase might be obtained with
>>better techniques--which is what this post is meant to
>>provoke.
>>
>>A stiffer rim, hauling harder on the rope, more initial
>>tension, and perhaps a more sensitive gauge might improve
>>the results.
>>
>>The rim is plain and not boxed, so it may not be very stiff.
>>
>>The rope squeeze may not be as good as one hand, even though
>>it looks pretty good. The rope is right in the midspan of
>>the spoke, but the area squeezed is much smaller than a
>>hand, which might affect leverage. (With the gauge on the
>>spoke, I can't get my whole hand in to squeeze.)
>>
>>The initial tension was embarrassingly low, which could
>>confuse the rim stiffness and other factors.
>>
>>The Park gauge is said to use a lot of force itself (you can
>>see the bending within the gauge in the picture) and might
>>confuse things).
>>
>>Here's a Park tension gauge showing 15 on the target spoke:
>>http://www.filelodge.com/files/room19/497501/spoke1.jpg
>>
>>Here it shows 20 with the spoke squeezed with a rope:
>>http://www.filelodge.com/files/room19/497501/spoke2.jpg
>>
>>A blurry picture of gauge and rope setup:
>>http://www.filelodge.com/files/room19/497501/spoke3.jpg
>>
>>Better focus of gauge and rope setup:
>>http://www.filelodge.com/files/room19/497501/spoke4.jpg
>>
>>Closer view of gauge and rope, showing bending of spoke:
>>http://www.filelodge.com/files/room19/497501/spoke5.jpg
>>
>>Here's Park's 0-50 scale conversion chart for 2mm straight
>>round steel spokes:
>>
>>scale_marking
>> Park_kgf
>> estimate_kgf
>>13 --
>>14 --
>>15 -- (41?) <--initial tension
>>16 -- (46?)
>>17 51
>>18 56
>>19 62
>>20 68 <--squeezed tension
>>21 76
>>22 85
>>23 95 <--recommended tension for most wheels
>>24 107
>>25 121
>>26 137
>>27 156
>>28 178
>>29 ---
>>30 ---
>>
>>I hope that this provokes pictures from people with better
>>cameras, wheels, grips, and testing techniques. If nothing
>>else, I learned that the rope will stay in place, unknotted,
>>under heavy tension because the spokes are so thin that the
>>loop jams solid.
>>
>>If anyone lacking a gauge can suggest a different way to do
>>it, I'll give it a try, with the obvious warning that my
>>experimental skills are already pushing their limits.
>>
>>Cheers,
>>
>>Carl Fogel
>
>
> Some posts have led me to wonder whether my powers of
> description are ever feebler than I thought, so here's a
> crude diagram of the setup:
>
> http://home.comcast.net/~carlfogel/download/biketest.jpg
>
> Imagine the small circle representing weights to be as many
> as ten weights dangling half-way to the floor like a bunch
> of bananas.
>
> The rotary vise was completely beyond my pictorial skills,
> so just imagine a monster bench vise that can be rotated to
> clamp a seat post.
>
> The question is whether the weight hanging down is going to
> give a strikingly different result than a single hand
> squeeze.
>
> The weight trying to turn the wheel is resisted by the brake
> pads, so maybe the rim deforms much more (or in a different
> direction) than it would if we could use a calibrated
> squeeze-grip of some kind.
>
> I can use another pair of spokes, or just a single spoke for
> a comparison, if someone has an idea. Joe Riel has suggested
> supporting the rim (which has a 110 psi tire on it), but I'm
> not sure where the support would go.
>
> Cheers,
>
> Carl Fogel

carl, while your experiment is terrific and a much more positive
contribution than the usual re-hashing of wildly misconceived
"engineering theory", to some extent, you're on a wild goose chase. by
definition, when you load something, it deflects. a wheel is a complex
3-d structure, so a load on one component has an effect on the whole.
to observe all this, you really need to set up the wheel with
micrometers oriented for all 3 axes at each spoke position on the rim -
that will give you the full extent of what's going on. or you can model
it, but that's a more involved exercise and you'd need data on the rim
which you'd probably have to obtain by experiment.

regarding spoke tension differential between the two you're "squeezing",
there are two issues:

1. you likely have a friction problem.
2. you need to take into account the effect of the brake reaction.

you can get around this by horizontally mounting the wheel and using
pulleys and two sets of weights. the whole structure has to be in
equilibrium, so the system you have right now is effectively 3-point
loading, not 2.

 

[Joe Riel]

jim beam <[email protected]> writes:

>
> 1. you likely have a friction problem.
> 2. you need to take into account the effect of the brake reaction.

On reflection I wonder whether either of these are significant
contributors to the what Carl is seeing, i.e. the tension decrease in
the lower spoke. That can be explained simply: even in the absence of
friction, the upper spoke is going to see more (twice) the lateral
force than the lower spoke. The resulting radial deflection of the
rim can/will decrease the tension in the lower spoke. The lateral
deflection of the rim, caused because only one pair of spokes is being
squeezed, also contributes. Of course, with friction the situation is
exacerbated.

--
Joe Riel

 

[Peter Cole]

[email protected] wrote:
> [email protected] wrote:
>
>> On Mon, 08 May 2006 09:34:51 -0500, Tim McNamara
>> <[email protected]> wrote:
>>
>>> OK: 300 mm spoke displaced 10 mm results in stretching the spokes .66
>>> mm. That's assuming that the rim does not deflect towards the hub.
>>>
>>> How much does a stretch of .66 mm raise the spoke tension?
>>>
>>> If we know the yield point of the spokes, how far to we have to deflect
>>> them to raise tension enough to achieve stress reliving?
>> Dear Tim,
>>
>> Unfortunately, the whole point of Joe Riel's article is that
>> the rim is not likely to be infinitely stiff and that
>> individual aluminum rims may deflect significantly toward
>> the hub.
>>
>> Private emails from other experimenters have shown a
>> noticeably greater tension rise than I got for the same
>> squeezing force--but they were using shorter spokes, which
>> suggests bigger and stiffer rims.
>>
>> The rim that I used is just a plain rim, not deep, no box
>> section, with 294mm spokes. The other testers mentioned
>> 270mm, 275mm, and 285mm spokes in their emails. I don't know
>> if any of them are going to post their results, which all
>> used squeeze forces around 30-50 pounds.
>>
>> Cheers,
>>
>> Carl Fogel
>
> The exact deflection of a given aluminum (i.e., isotropic) rim under a
> spoke-squeezy loading condition would be easy to predict with FEA. All
> we would need is the moment of inertia of the rim cross-section in the
> plane of the wheel. If someone could provide me with that, I'd be happy
> to run a model using beam elements.
>
> Jason
>

Have you seen:
<http://www.duke.edu/~hpgavin/papers/HPGavin-Wheel-Paper.pdf>

 

[Peter Cole]

[email protected] wrote:

> It's far simpler than that. Rim deflection is lateral spoke
> deflection divided by the tangent function from spoke deflection and
> length, something easy to measure. All deflection comes from rim
> deflection, the spokes not changing length. To hell with isotropic or
> Hook's Law and the like.

It seems that when stress relieving by squeezing 4 spokes, given the
rough stiffness of spokes around 1KN/mm and rim (built) of 3KN/mm that
the change of length of the spokes would be roughly equal to the rim
deflection.

 

[[email protected]]

On Tue, 09 May 2006 13:33:19 GMT, Joe Riel
<[email protected]> wrote:

>jim beam <[email protected]> writes:
>
>>
>> 1. you likely have a friction problem.
>> 2. you need to take into account the effect of the brake reaction.
>
>On reflection I wonder whether either of these are significant
>contributors to the what Carl is seeing, i.e. the tension decrease in
>the lower spoke. That can be explained simply: even in the absence of
>friction, the upper spoke is going to see more (twice) the lateral
>force than the lower spoke. The resulting radial deflection of the
>rim can/will decrease the tension in the lower spoke. The lateral
>deflection of the rim, caused because only one pair of spokes is being
>squeezed, also contributes. Of course, with friction the situation is
>exacerbated.

Dear Joe,

I think that at least part of the solution is going to be to
hang the weight downward from a rope on the upper spoke
alone, while running a second, independent rope upward from
the lower spoke to the ceiling:

----|----- ceiling
|
|
------k------------------- upper spoke-----hub
| | upside-down
| | bike
----k--------------------- lower spoke-----hub
|
| k=knot
|
|
xx-pound weight

Then no brake is needed, and instead of friction-confused
pulley, xx pounds up should be balanced by the rope to the
ceiling's resistance.

For sideways deflection because only a non-drive-side pair
is being being squeezed, things get trickier.

A crude balance with just a rope wound around the opposite
drive-side-pair might be enough.

A duplicate ceiling rope and weight might be possible, but
you run into the problem of the weights being much too wide
and bulky to hang only a rim-width apart conveniently
without pushing each other apart, twirling, and so forth.

Duplicate ceiling ropes and a single weight with two ropes,
one from the upper spoke on each side of the rim would give
half the xx weight, but is getting tricky again.

I need to work on a stout, convenient ceiling anchors in
just the right place and see how the practical side works
out. Back to bed.

Cheers,

Carl Fogel

 

[Joe Riel]

[email protected] writes:

> I think that at least part of the solution is going to be to
> hang the weight downward from a rope on the upper spoke
> alone, while running a second, independent rope upward from
> the lower spoke to the ceiling:
>
> ----|----- ceiling
> |
> |
> ------k------------------- upper spoke-----hub
> | | upside-down
> | | bike
> ----k--------------------- lower spoke-----hub
> |
> | k=knot
> |
> |
> xx-pound weight
>
> Then no brake is needed, and instead of friction-confused
> pulley, xx pounds up should be balanced by the rope to the
> ceiling's resistance.

I don't see an obvious need for having the bike attached to the wheel;
it complicates the situation in that it provides an additional anchor
so the tension in the two ropes may not be equal. It does provide one
benefit, it prevents the wheel from rotating laterally. However, it
is unnecessary, instead of squeezing the spokes together, pull them
apart---the situation is identical but the setup is then stable.

--
Joe Riel

 

[Leo Lichtman]

Jobst Brandt wrote: Hook's Law
>Jim Beam wrote: hooke
^^^^^^^^^^^^^^^^^^
Jim, if we;re going to concentrate on trivia, lets make that "Hooke's law."

However, I believe the important point is that stress relieving a wheel by
squeezing the spokes does two things: 1.) It may cause the spoke ends to
seat better in their holes. 2.) It can cause the bends at the hub holes to
change their angle (by metal yield.) Neither of these effects follows
Hooke's law.

 

[Peter Cole]

Leo Lichtman wrote:
> Jobst Brandt wrote: Hook's Law
>> Jim Beam wrote: hooke
> ^^^^^^^^^^^^^^^^^^
> Jim, if we;re going to concentrate on trivia, lets make that "Hooke's law."
>
> However, I believe the important point is that stress relieving a wheel by
> squeezing the spokes does two things: 1.) It may cause the spoke ends to
> seat better in their holes. 2.) It can cause the bends at the hub holes to
> change their angle (by metal yield.) Neither of these effects follows
> Hooke's law.
>
>

Nor do either come from stress relieving.

 

[[email protected]]

Hi Peter,

No, I haven't seen the Gavin paper, and wow! That's beyond what I was
hoping to get. He approximates Izz, Irr and GJ[1] very well, and
provides the constants I would need to build a model. That's very
exciting. Now developing my own FE model is just a question of how much
free time I have. Aside from that, the paper is fascinating and
provides hard data about spoke fatigue cycles and a few other things.
This is great.

Thank you!

Jason



[1] Izz, Irr and GJ are moments of inertia for the rim, which are a way
of describing stiffness about various axes. They depend on the geometry
and material of the rim.

 

[[email protected]]

On Sat, 06 May 2006 15:49:51 -0600, [email protected]
wrote:

Here are the new test results. Here's a diagram of the new
setup, which uses no brake--the upper spoke is pulled down
by a weight on a rope, while the lower spoke is held up by
two ends of a rope from the ceiling:

http://home.comcast.net/~carlfogel/download/newspoke.jpg

As I expected, without the friction and pulley effect, the
spoke tension increase was even smaller. Adding a 100-lb
weight raised the upper/lower spoke tension from 185/144 lbs
to 263/273 lbs, only about 80 to 130 pounds.

Again, this unexpectedly low tension rise could be due to
the rim not being anywhere near infinite stiffness, the
initial tension, or some goof in the setup.

At 100 lbs of weight, a clamp was used to squeeze the
opposite drive-side pair of parallel spokes, but almost no
change in tension was measured.

Here's a picture of the spoke pair with 100 pounds of
weights pulling the upper spoke down (middle rope) and two
ends of a separate rope going up to the ceiling and holding
the lower spoke up. The sheet of paper is just to make it
easier to see how the spokes are bending toward each other,
but the shadows confuse the view of ropes (and even the
spokes):

http://home.comcast.net/~carlfogel/download/spoke100lb.jpg

Sorry about the lack of overall pictures, but anything big
enough to show the whole setup fails to show the details,
and I had to remove the fluorescent light over the workbench
just to get the bike to fit.

Cheers,

Carl Fogel

***

700c rear wheel retired last week after 30,000 miles
shallow, unimpressive rim, practically no brake wear
(my daily ride just doesn't involve much braking)
non-drive-side spokes
36 2.0mm straight stainless steel

this is the new setup
rope from ceiling instead of brake pad
noticeably smaller tension increases

parallel pair @ 90 degrees to valve hole
different pair than last time

clamp bike upside-down in rotary vise
back of wheel sticks out over edge of workbench
wheel is upright, vertical, bike is upside down

tire inflated to 110 psi

lower spoke hangs from two ends of rope to ceiling
this just holds lower spoke up, nothing fancy
wheel is free, no brake pad clamping

tie rope to midspan of upper spoke in parallel pair

hang 5-pound weights from 0 to 100 pounds
weight of rope trivial

just enough room toward rim to use Park gauge
ropes not touching anything else
not tire, rim, each other, or other spokes

5-lb weight added to rope
all weight lifted, then lowered and left to hang

Park spoke tension gauge squeezed 6 times
left edge near spoke nipple
(6 times instead of 3, smaller changes when weights added)

at midspan, gap narrowed from ~75mm to ~35mm at 100 lbs

at 100 lbs weight, opposite parallel spoke pair clamped
welder's vise-grip, narrowed drive-side pair gap to 44mm
scarcely any tension change measured
263/273 pounds rose to 270/273 pounds

decimal points only generallly reliable
best estimate, influenced by memory of previous estimate
ropes stayed in midspan better with this setup
no rise noted or slight drop in gauge reading at low weight

extrapolate Park chart to decimal values

raw data convert park mark to spoke tension

upper lower weight upper lower upper lower total
spoke spoke in spoke spoke spoke spoke spokes
gauge gauge pounds kgf kgf lbs lbs lbs

21.9 19.6 0 84.1 65.6 185 144 329.3
21.7 19.6 5 82.3 65.6 181 144 325.4
21.2 19.6 10 77.8 65.6 171 144 315.5
21.5 20.2 15 80.5 69.6 177 153 330.2
21.7 20.7 20 82.3 73.6 181 162 343.0
21.3 21.7 25 78.7 82.3 173 181 354.2
21.9 21.7 30 84.1 82.3 185 181 366.1
22.0 21.8 35 85.0 83.2 187 183 370.0
22.2 22.2 40 87.0 87.0 191 191 382.8
22.4 22.5 45 89.0 90.0 196 198 393.8
22.5 23.0 50 90.0 95.0 198 209 407.0
23.0 23.2 55 95.0 97.4 209 214 423.3
22.5 23.5 60 90.0 101.0 198 222 420.2
23.2 23.8 65 97.4 104.6 214 230 444.4
23.4 23.9 70 99.8 105.8 220 233 452.3
23.8 24.1 75 104.6 108.4 230 238 468.6
24.0 24.6 80 107.0 115.4 235 254 489.3
24.0 25.0 85 107.0 121.0 235 266 501.6
24.9 24.8 90 119.6 118.2 263 260 523.2
24.6 25.0 95 115.4 121.0 254 266 520.1
24.9 25.2 100 119.6 124.2 263 273 536.4

w/clamp
25.1 25.0 100 122.6 124.2 270 273 543.0
on opposite
spoke pair

gauge chart gauge chart
mark kgf mark kgf
estimate extrapolated
17 51
18 56 18.0 56 0.6 steps
19 62 18.1 56.6
20 68 18.2 57.2
21 76 18.3 57.8
22 85 18.4 58.4
23 95 18.5 59.0
24 107 18.6 59.6
25 121 18.7 60.2
26 137 18.8 60.8
27 156 18.9 61.4
28 179 19.0 62 0.6 steps
19.1 62.6
19.2 63.2
19.3 63.8
19.4 64.4
19.5 65.0
19.6 65.6
19.7 66.2
19.8 66.8
19.9 67.4
20.0 68 0.8 steps
20.1 68.8
20.2 69.6
20.3 70.4
20.4 71.2
20.5 72.0
20.6 72.8
20.7 73.6
20.8 74.4
20.9 75.2
21.0 76 0.9 steps
21.1 76.9
21.2 77.8
21.3 78.7
21.4 79.6
21.5 80.5
21.6 81.4
21.7 82.3
21.8 83.2
21.9 84.1
22.0 85 1.0 steps
22.1 86.0
22.2 87.0
22.3 88.0
22.4 89.0
22.5 90.0
22.6 91.0
22.7 92.0
22.8 93.0
22.9 94.0
23.0 95 1.2 steps
23.1 96.2
23.2 97.4
23.3 98.6
23.4 99.8
23.5 101.0
23.6 102.2
23.7 103.4
23.8 104.6
23.9 105.8
24.0 107 1.4 steps
24.1 108.4
24.2 109.8
24.3 111.2
24.4 112.6
24.5 114.0
24.6 115.4
24.7 116.8
24.8 118.2
24.9 119.6
25.0 121 1.6 steps
25.1 122.6
25.2 124.2
25.3 125.8
25.4 127.4
25.5 129.0
25.6 130.6
25.7 132.2
25.8 133.8
25.9 135.4
26.0 137 1.9 steps
26.1 138.9
26.2 140.8
26.3 142.7
26.4 144.6
26.5 146.5
26.6 148.4
26.7 150.3
26.8 152.2
26.9 154.1
27.0 156
end of table


 

[grant]

> Some posts have led me to wonder whether my powers of
> description are ever feebler than I thought, so here's a
> crude diagram of the setup:
>
> http://home.comcast.net/~carlfogel/download/biketest.jpg

Apologies if I have missed the point of this exercise, but your diagram
indicates that you are squeezing two spokes ON THE SAME SIDE of the
wheel, two non-drive side spokes in this case.

I always assumed squeezing spokes (for stress relieving) refers to
spokes on opposite sides of the wheel.

Surely squeezing on the same side you'll be deflecting the rim to that
side (just as truing would) and thus not seeing as much increase in
tension as if you had squeezed spokes on opposite side.

If this thread is deliberately about spokes-on-the-same-side, ignore the
above

gsf.

 

[[email protected]]

Carl Fogel writes:

>>>> (Look carefully at the raw data, and you'll see a spot or two
>>>> where it didn't seem to budge with added weight and I irritably
>>>> recorded the data as such. And I was certainly baffled when I
>>>> added weight after weight and the lower spoke lost tension
>>>> instead of gaining it as I expected.)

>>> Presumably that was caused by the rim deflecting downward,
>>> effectively shortening the path for the bottom spoke. Supporting
>>> the rim so it cannot deflect laterally (downward) should stop
>>> that.

>> Oooh! I'm gradually getting the picture. The bicycle is lying on
>> its side and there is a a rope around a pair of spokes at their
>> midpoint pulling down.

> [snip]

> Sorry, the bike is not on its side. As I keep saying, it's upside
> down.

> Here's a sketch that I just posted:

> http://home.comcast.net/~carlfogel/download/biketest.jpg

> But after you see how it's working, maybe you'll see whatever is
> leading to unexpected results.

Now I see what's happening. Ideally if there were no friction where
the rope goes over the upper spoke the forces could be assessed... but
there is friction and the rope is not a smooth surface without
features. Beyond that, there is a reaction force trying to turn the
wheel and this puts other forces into the spoke complement.

> Joe Riel has wondered if supporting the wheel might help, but I have
> no idea if the forces are downward or circular or even affecting the
> result.

That might well do it but supporting the wheel would have to be done
by a bridge under the spoke nipple to prevent wheel rotation at the
source. There are too many unresolved reaction forces just the same.

> The brake pads stop the wheel from turning when the weight hangs
> from the horizontal spokes.

Yes, but they are at the other side of the wheel and act through the
entire wheel to do that task. They are "pulling" down on the opposite
side of the wheel. This is like braking forces (shown in "the Bicycle
Wheel") and cause spoke tension changes in the wheel (see diagram in
book).

> Joe Riel has wondered if supporting the wheel might help, but I have
> no idea if the forces are downward or circular or both or even
> affecting the result.

This is not a good experimental arrangement. I don't see a way to
clean it up.

Jobst Brandt

 

[[email protected]]

Jason Krantz writes:

> No, I haven't seen the Gavin paper, and wow! That's beyond what I
> was hoping to get. He approximates Izz, Irr and GJ[1] very well, and
> provides the constants I would need to build a model. That's very
> exciting. Now developing my own FE model is just a question of how
> much free time I have. Aside from that, the paper is fascinating and
> provides hard data about spoke fatigue cycles and a few other
> things. This is great.

> [1] Izz, Irr and GJ are moments of inertia for the rim, which are a
> way of describing stiffness about various axes. They depend on the
> geometry and material of the rim.

Meanwhile, don't forget that rims have spoke holes in them and
therefore, a discontinuous cross sectional area. Of course this has
little to do with the matter at hand because this does not affect the
input force while stress relieving.

Jobst Brandt

 

[[email protected]]

Carl Fogel writes:

> Here are the new test results. Here's a diagram of the new setup,
> which uses no brake--the upper spoke is pulled down by a weight on a
> rope, while the lower spoke is held up by two ends of a rope from
> the ceiling:

http://home.comcast.net/~carlfogel/download/newspoke.jpg

> As I expected, without the friction and pulley effect, the spoke
> tension increase was even smaller. Adding a 100-lb weight raised the
> upper/lower spoke tension from 185/144 lbs to 263/273 lbs, only
> about 80 to 130 pounds.

> Again, this unexpectedly low tension rise could be due to the rim
> not being anywhere near infinite stiffness, the initial tension, or
> some goof in the setup.

Could you pick a pair of spokes with nearly equal tension to start
with, or adjust the looser spoke to be as tight as the other so this
disparity in deflection and tension change is less obscure? The
arrangement you have should cause the same change in both spokes
unless there is some unseen influence being overlooked.

> At 100 lbs of weight, a clamp was used to squeeze the opposite
> drive-side pair of parallel spokes, but almost no change in tension
> was measured.

Something seems wrong about the results. If I hang a 100lb weight on
a clothesline there has got to be more than 100lbs tension in the line
if it is anywhere near horizontal, with a sag of less than 10 degrees,
which we seem to have in your wheel.

> Here's a picture of the spoke pair with 100 pounds of weights
> pulling the upper spoke down (middle rope) and two ends of a
> separate rope going up to the ceiling and holding the lower spoke
> up. The sheet of paper is just to make it easier to see how the
> spokes are bending toward each other, but the shadows confuse the
> view of ropes (and even the spokes):

http://home.comcast.net/~carlfogel/download/spoke100lb.jpg

> Sorry about the lack of overall pictures, but anything big enough to
> show the whole setup fails to show the details, and I had to remove
> the fluorescent light over the workbench just to get the bike to
> fit.


> Park spoke tension gauge squeezed 6 times
> left edge near spoke nipple
> (6 times instead of 3, smaller changes when weights added)

> at midspan, gap narrowed from ~75mm to ~35mm at 100 lbs

> at 100 lbs weight, opposite parallel spoke pair clamped welder's
> vise-grip, narrowed drive-side pair gap to 44mm scarcely any tension
> change measured 263/273 pounds rose to 270/273 pounds

> decimal points only generally reliable best estimate, influenced by
> memory of previous estimate ropes stayed in midspan better with this
> setup no rise noted or slight drop in gauge reading at low weight

> extrapolate Park chart to decimal values

> raw data convert park mark to spoke tension
> ... ... ...

This would be a bit easier to decipher if it were shown graphically.
Stacks of numbers look like a maze without exit.

Jobst Brandt

 

[[email protected]]

Grant who? writes:

>> Some posts have led me to wonder whether my powers of description
>> are ever feebler than I thought, so here's a crude diagram of the
>> setup:

http://home.comcast.net/~carlfogel/download/biketest.jpg

> Apologies if I have missed the point of this exercise, but your
> diagram indicates that you are squeezing two spokes ON THE SAME SIDE
> of the wheel, two non-drive side spokes in this case.

> I always assumed squeezing spokes (for stress relieving) refers to
> spokes on opposite sides of the wheel.

The method is shown pictorially in "the Bicycle Wheel" and involves
grasping pairs of nearly parallel spokes on opposite sides of the
wheel squeezing them together forcefully. It is this force, or
better, the increase in spoke tension that it causes that is being
investigated.

> Surely squeezing on the same side you'll be deflecting the rim to
> that side (just as truing would) and thus not seeing as much
> increase in tension as if you had squeezed spokes on opposite side.

That is why it is done manually on opposite sides of the wheel so that
there is little lateral deflection. This is done mainly to prevent
loss of wheel truing, not to change the resulting tension. Therefore,
the current experimental method is a good approximation of that
process.

> If this thread is deliberately about spokes-on-the-same-side, ignore
> the above

Jobst Brandt