105 flange failure pics

cashrefundman said:

This summer I snapped a newish 105 hub flange/shell across 4 spokes.
Take a look

http://freeengineer.org/flangefailure.html

And you guys said pulling spokes on the outside was OK 🙂

CRM

that's an unusually large chunk of flange you had fail there. what was
your spoke tension?

In article <[email hidden]>,

cashrefundman said:

This summer I snapped a newish 105 hub flange/shell across 4 spokes.
Take a look

http://freeengineer.org/flangefailure.html

And you guys said pulling spokes on the outside was OK 🙂

CRM

That's an interesting failure.

Don't take this the wrong way, but how did you take five pictures of
that failure, and not get one of them where the metal faces were in
focus? Seeing detail on those areas can help the smart guys here
diagnose the failure mode.

--
Ryan Cousineau, [email hidden] http://www.wiredcola.com
Verus de parvis; verus de magnis.

cashrefundman <[email hidden]> wrote in message news:<[email hidden]>...

Quoted message said:

This summer I snapped a newish 105 hub flange/shell across 4 spokes.
Take a look

http://freeengineer.org/flangefailure.html

And you guys said pulling spokes on the outside was OK 🙂

CRM

That's an impressive failure. FWIW, I'm a bit lighter (probably 250
max total, including bike, me, baggage, etc.) and I have a set of
'new' 105 32h holes with about 8,000 miles on 'em, and I've never had
a problem with them. I assume Shimano warranteed the hub?

jim beam said:

that's an unusually large chunk of flange you had fail there. what was
your spoke tension?

Don't know the tension but I built and maintained it to be "pretty
tight" on the drive side.

CRM

Ryan Cousineau said:

Don't take this the wrong way, but how did you take five pictures of
that failure, and not get one of them where the metal faces were in
focus?

I tried but my camera is autofocus and won't on that spot.

Seeing detail on those areas can help the smart guys here

Quoted message said:

diagnose the failure mode.

Well it's just grainy looking cast aluminum. I don't doubt that it is a
fatigue failure.

To me it looks like it started in at the corner of recess where the
freehub body nests into the hub shell (The thinnest part of the hub
shell) and propagated from there to the spoke holes

CRM

cashrefundman said:

jim beam said:

that's an unusually large chunk of flange you had fail there. what
was your spoke tension?

Don't know the tension but I built and maintained it to be "pretty
tight" on the drive side.

CRM

well, i know there are years of archive "advice" on this forum telling
people to build wheels with spokes as tight as possible, but i'm telling
you this is flawed. excess tension does /not/ improve wheel rigidity,
only spoke & rim modulus do that. excess tension does not improve spoke
fatigue resistance, only the manufacturer of the spokes does that.

what excess spoke tension /does/ achieve is rim cracking and hub flange
failure. if i rebuilt this wheel, i would;

1. use the rim manufacturer's specified spoke tension.
2. use the hub manufacturer's specified spoke lacing pattern.
3. enjoy long & trouble-free service.

imo, this excess spoke tension stuff is probably the single biggest
factor in the rapid rise of the pre-built wheel market. if i were
mavic, and kept being presented with warranty returns for cracked rims
that my lab testing showed to have been built with excess tension, there
would come a point where i would want to control the lunacy. the only
solution is to get into the wheel building market - that way, product
leaving the factory is /known/ to be built to spec. and warranty becomes
managable.

jim beam said:

cashrefundman said:

Don't know the tension but I built and maintained it to be "pretty
tight" on the drive side.

CRM

well, i know there are years of archive "advice" on this forum telling
people to build wheels with spokes as tight as possible, but i'm telling
you this is flawed. excess tension does /not/ improve wheel rigidity,
only spoke & rim modulus do that. excess tension does not improve spoke
fatigue resistance, only the manufacturer of the spokes does that.

People who have read and understood the years of archived advice from this
forum know that high tension does not improve wheel rigidity and they know
that high tension does not improve spoke fatigue resistance. They also
know that neither of these benefits has ever been claimed. The only
claim is that high tension increases the ultimate strength of the wheel.
This claim has been repeatedly shown to be true. Please explain how this
proven relationship between spoke tension and wheel strength is flawed.

--
Todd Bryan
Santa Barbara, CA
bryan at cs dot utk dot edu

cashrefundman said:

This summer I snapped a newish 105 hub flange/shell across 4 spokes.
Take a look

http://freeengineer.org/flangefailure.html

And you guys said pulling spokes on the outside was OK 🙂

CRM

I am glad that you were able to ride home.
36 spokes built in semi-tangential way allowed you the redundancy to do this.
Phil Wood flanges are very thick and tough, but I think you could use another Shimano hub without a similar failure.

1. Double Butted spokes help spread the load amongst themselves and over larger portions of the rim and *HUB*.
2. The hub could have been damaged during the build if the builder used a spoke head setting tool and hit the hub flange hard enough. Other damage to the hub flange could have happened, but it doesn't help to speculate further without more knowledge of this particular hub.
3. Spoke tension balancing is important to have spokes share the load and the rim/hub interfaces have even load distributio as well.
4. Spoke support angle makes a larger difference as the load increases. If you can use a wider OLD and/or a rim with OFF CENTER spoke bead, you can accomplish spreading the load more effectively.
5. Stiffer rims help spread the load over more spokes, and becuase of this over a larger number of spoke/hub hole interfaces.

Chalo is a very big rider with considerable wheel building and riding experience.
Now that he is healing from his recent crash, he may add his thoughts.
His suggestions could help even more.

In article <[email hidden]>,

cashrefundman said:

Ryan Cousineau said:

Don't take this the wrong way, but how did you take five pictures of
that failure, and not get one of them where the metal faces were in
focus?

I tried but my camera is autofocus and won't on that spot.

Dang.

Quoted message said:

Seeing detail on those areas can help the smart guys here

Quoted message said:

diagnose the failure mode.

Well it's just grainy looking cast aluminum. I don't doubt that it is a
fatigue failure.

One issue is that you may see two different looks on the broken bits:
one where the tearing initiated slowly, until enough of the part failed
that it breaks the flange section off catastrophically.

Quoted message said:

To me it looks like it started in at the corner of recess where the
freehub body nests into the hub shell (The thinnest part of the hub
shell) and propagated from there to the spoke holes

Any signs of scratches or other features that might have initiated a
crack in that area?

--
Ryan Cousineau, [email hidden] http://www.wiredcola.com
Verus de parvis; verus de magnis.

cashrefundman said:

This summer I snapped a newish 105 hub flange/shell across 4 spokes.
Take a look

http://freeengineer.org/flangefailure.html

And you guys said pulling spokes on the outside was OK 🙂

Ouch. I have one of those on the back of my racing bike (and, worse, a
radially-laced Ultegra on the front).

Ryan Cousineau said:

Quoted message said:

Well it's just grainy looking cast aluminum. I don't doubt that it is a
fatigue failure.

Quoted message said:

Any signs of scratches or other features that might have initiated a
crack in that area?

I have now disassembled the wheel and cleaned the hub and must withdrawl
my previous theory.

It was a spoke hole failure.

How do I know? There are 3 other drive side pulling spoke holes with
partial or full cracks of exactly the same shape as the failure.

So there ya go.

Guess it was just the wrong hub for my application, weight, abuse
coefficient and sense of spoke tension.

I think I'll stick with my phil 7 speed for a now.

CRM

In article <[email hidden]>,
[email hidden] says...

Quoted message said:

I tried but my camera is autofocus and won't on that spot.

Many autofocus cameras can be fooled into focusing on what you want.

Say you want to focus with the hub 12 inches from the lens. Simply
point the camera at the floor from 12 inches away and partly depress the
shutter. This will lock in the focus and aperture. Now, point your
camera at the portion of the hub you want to snap from a distance of 12
inches, and it will be in focus. Depress the shutter and you've got it.

Rick

Todd Bryan said:

jim beam said:

cashrefundman said:

Don't know the tension but I built and maintained it to be "pretty
tight" on the drive side.

CRM

well, i know there are years of archive "advice" on this forum telling
people to build wheels with spokes as tight as possible, but i'm telling
you this is flawed. excess tension does /not/ improve wheel rigidity,
only spoke & rim modulus do that. excess tension does not improve spoke
fatigue resistance, only the manufacturer of the spokes does that.

People who have read and understood the years of archived advice from this
forum know that high tension does not improve wheel rigidity and they know
that high tension does not improve spoke fatigue resistance. They also
know that neither of these benefits has ever been claimed. The only
claim is that high tension increases the ultimate strength of the wheel.
This claim has been repeatedly shown to be true. Please explain how this
proven relationship between spoke tension and wheel strength is flawed.

look at this graph.

http://www.damonrinard.com/wheel/tension.gif

origin:
http://www.damonrinard.com/wheel/index.htm

see how the front end of the graph is a flat line vs. tension? that's
because the stiffness of the wheel is a function of the modulus of the
components used, not the tension. if the components are constant and
remain below yield, there's no way stiffness can change.

by the same rationale, if the materials are constant, the strength
[yield] of the wheel is not increasing with tension either - there's
simply no mechanism for it. does a spring get stronger the more load
you place on it? the force exerted by the spring is more, but the
strength, as defined by displacement per unit load, remains the same.

regarding fatigue strength, that decreases as a function of increasing load.

jim beam said:

by the same rationale, if the materials are constant, the strength
[yield] of the wheel is not increasing with tension either - there's
simply no mechanism for it. does a spring get stronger the more load
you place on it? the force exerted by the spring is more, but the
strength, as defined by displacement per unit load, remains the same.

A wire-spoked wheel's load bearing capacity before loss of spoke
tension is directly proportional to spoke tension. Cyclical loss of
all spoke preload tension causes structural instability that leads to
rapid failure of the wheel.

If that were not the case, then we could all just lace our wheels
finger-tight with an effective threadlocking compound, and there would
be no sacrifice of longevity or durability.

Optimum spoke tension is as high as the rim will support. If holes
crack or bulge, or if eyelets pop out, that the rim won't support that
tension. The reason Jobst specifies the buckling strength of the rim
instead is that he only uses Mavic MA2s. The MA2's sturdy stainless
steel socketed eyelets and conventional hole counts mean that the
rim's Euler limit dictates its maximum spoke tension.

Jobst would likely say that a rim whose spoke holes imposed its spoke
tension limit was poorly designed, and I'm inclined to agree.

Chalo Colina

jim beam said:

Todd Bryan said:

People who have read and understood the years of archived advice from
this
forum know that high tension does not improve wheel rigidity and they
know
that high tension does not improve spoke fatigue resistance. They also
know that neither of these benefits has ever been claimed. The only
claim is that high tension increases the ultimate strength of the wheel.
This claim has been repeatedly shown to be true. Please explain how this
proven relationship between spoke tension and wheel strength is flawed.

look at this graph.

http://www.damonrinard.com/wheel/tension.gif

origin:
http://www.damonrinard.com/wheel/index.htm

see how the front end of the graph is a flat line vs. tension? that's
because the stiffness of the wheel is a function of the modulus of the
components used, not the tension. if the components are constant and
remain below yield, there's no way stiffness can change.

by the same rationale, if the materials are constant, the strength
[yield] of the wheel is not increasing with tension either - there's
simply no mechanism for it. does a spring get stronger the more load
you place on it? the force exerted by the spring is more, but the
strength, as defined by displacement per unit load, remains the same.

That graph is good to show the role of spoke tension, but
you aren't getting the entire story because you are ignoring
an important part of the graph. As you point out, the left
side of the graph is flat, showing that the stiffness of the
wheel under a fixed load is constant - but only within that
range of spoke tensions. In the right side of the graph, at
lower tensions, the graph shoots upward, showing that under
a certain amount of spoke tension, the deflection of the
wheel increases. Why should this be? Clearly, the
elasticity of the individual wheel components is the same.
But, when the spoke tension is too low, some of the spokes
go completely slack when a load is applied, and when the rim
is no longer constrained by the stiffness of these spokes,
the rim is free to deflect.

This is the important consequence when the spoke tension is
low - the rim can deflect (strain) more under a given load.
If the load increases and the rim strains too far, it will
permanently deform (yield) - in other words, the wheel will
fail. Higher tension in the spokes increases the amount of
load required to slacken the spokes, keeping most of the
deflection in the high stiffness region, thereby increasing
the load required to reach the rim's yield strain. Because
the stiffness of the spokes support the rim both
horizontally and vertically, the magnitude of the tension in
the spokes plays a role in the strength of the wheel both
vertically and horizontally.

I have performed an experiment which demonstrates the
importance of spoke tension on wheel strength. I had two
front wheels built with equivalent components, one with high
spoke tension and one with low spoke tension. I laid the
high spoke tension wheel on the ground, resting on one the
end of the axle. I then stepped onto the rim with my feet
on opposite sides (180 degrees apart) and with my entire
wheel on the wheel, I balanced the wheel on the end of the
axle (I pressed my hands against a wall in order to stay
balanced on the axle, but my entire weight was borne at two
points on opposite points of the rim). This represented an
extreme case of Rinard's lateral load test. The wheel
easily bore my weight, and showed no ill aftereffects from
this test. I then repeated the test with the low tension
wheel. This time, the rim collapsed, and yielded into a
saddle shape (aka, the wheel tacoed, to use the common
vernacular). The wheel with the higher spoke tension was
clearly stronger.

Strength to vertical forces should be similarly affected by
spoke tension. Unfortunately, I don't have a way to
generate repeatable forces high enough to test wheel
vertical strength.

When you consider that external forces on the wheel
primarily cause spoke tension decreases, and that when the
spokes are completely slackened they no longer support the
rim, and that an unsupported rim can be yielded more easily
than a well supported rim, it is clear that the magnitude of
the spoke tension plays a vital role in the ultimate
strength of a wheel.

Mark McMaster
[email hidden]

Mark McMaster said:

jim beam said:

Todd Bryan said:

People who have read and understood the years of archived advice from
this
forum know that high tension does not improve wheel rigidity and they
know
that high tension does not improve spoke fatigue resistance. They also
know that neither of these benefits has ever been claimed. The only
claim is that high tension increases the ultimate strength of the wheel.
This claim has been repeatedly shown to be true. Please explain how
this
proven relationship between spoke tension and wheel strength is flawed.

look at this graph.

http://www.damonrinard.com/wheel/tension.gif

origin:
http://www.damonrinard.com/wheel/index.htm

see how the front end of the graph is a flat line vs. tension? that's
because the stiffness of the wheel is a function of the modulus of the
components used, not the tension. if the components are constant and
remain below yield, there's no way stiffness can change.

by the same rationale, if the materials are constant, the strength
[yield] of the wheel is not increasing with tension either - there's
simply no mechanism for it. does a spring get stronger the more load
you place on it? the force exerted by the spring is more, but the
strength, as defined by displacement per unit load, remains the same.

That graph is good to show the role of spoke tension, but you aren't
getting the entire story because you are ignoring an important part of
the graph. As you point out, the left side of the graph is flat, showing
that the stiffness of the wheel under a fixed load is constant - but
only within that range of spoke tensions. In the right side of the
graph, at lower tensions, the graph shoots upward, showing that under a
certain amount of spoke tension, the deflection of the wheel increases.
Why should this be? Clearly, the elasticity of the individual wheel
components is the same. But, when the spoke tension is too low, some of
the spokes go completely slack when a load is applied, and when the rim
is no longer constrained by the stiffness of these spokes, the rim is
free to deflect.

This is the important consequence when the spoke tension is low - the
rim can deflect (strain) more under a given load. If the load increases
and the rim strains too far, it will permanently deform (yield) - in
other words, the wheel will fail. Higher tension in the spokes
increases the amount of load required to slacken the spokes, keeping
most of the deflection in the high stiffness region, thereby increasing
the load required to reach the rim's yield strain. Because the
stiffness of the spokes support the rim both horizontally and
vertically, the magnitude of the tension in the spokes plays a role in
the strength of the wheel both vertically and horizontally.

I have performed an experiment which demonstrates the importance of
spoke tension on wheel strength. I had two front wheels built with
equivalent components, one with high spoke tension and one with low
spoke tension. I laid the high spoke tension wheel on the ground,
resting on one the end of the axle. I then stepped onto the rim with my
feet on opposite sides (180 degrees apart) and with my entire wheel on
the wheel, I balanced the wheel on the end of the axle (I pressed my
hands against a wall in order to stay balanced on the axle, but my
entire weight was borne at two points on opposite points of the rim).
This represented an extreme case of Rinard's lateral load test. The
wheel easily bore my weight, and showed no ill aftereffects from this
test. I then repeated the test with the low tension wheel. This time,
the rim collapsed, and yielded into a saddle shape (aka, the wheel
tacoed, to use the common vernacular). The wheel with the higher spoke
tension was clearly stronger.

Strength to vertical forces should be similarly affected by spoke
tension. Unfortunately, I don't have a way to generate repeatable
forces high enough to test wheel vertical strength.

When you consider that external forces on the wheel primarily cause
spoke tension decreases, and that when the spokes are completely
slackened they no longer support the rim, and that an unsupported rim
can be yielded more easily than a well supported rim, it is clear that
the magnitude of the spoke tension plays a vital role in the ultimate
strength of a wheel.

Mark McMaster
[email hidden]

mark [& chalo]

think big picture a sec. how does just increasing pre-load change a
material's ultimate load capacity? the only way that can happen is if
the material is a variable, which is clearly not the case. that's why
the graph is flat line. increasing the deflection force just shifts the
position of the line, not its slope.

simple example, & one on which tom sherman may care to elaborate:
pre-stressed steel reinforced concrete has reinforcements which are
tensioned to increase the load the concrete can bear before the onset of
cracking. fairly obvious, right? however, that pre-load reduces the
overall load the concrete component can resist before failure - maybe
not so obvious, but logical if you think about it. the premise that
"increasing tension increases strength" is more an article of faith than
logic.

regarding the taco wheel experiment, i too have tried that. my
experience was different to yours in that the rim itself makes a huge
difference. spoke tension, within moderate levels of rim max spec, does
not. excess tension makes it /easy/ to taco the wheel. when you say
"eqivalent components" were used in your two wheels, can you confirm
that the rims were identical? and what were the spoke tensions?

jim beam said:

Mark McMaster said:

jim beam said:

Todd Bryan wrote:

>
> People who have read and understood the years of archived advice
> from this
> forum know that high tension does not improve wheel rigidity and
> they know
> that high tension does not improve spoke fatigue resistance. They also
> know that neither of these benefits has ever been claimed. The only
> claim is that high tension increases the ultimate strength of the
> wheel.
> This claim has been repeatedly shown to be true. Please explain how
> this
> proven relationship between spoke tension and wheel strength is flawed.
>
>

look at this graph.

http://www.damonrinard.com/wheel/tension.gif

origin:
http://www.damonrinard.com/wheel/index.htm

see how the front end of the graph is a flat line vs. tension?
that's because the stiffness of the wheel is a function of the
modulus of the components used, not the tension. if the components
are constant and remain below yield, there's no way stiffness can
change.

by the same rationale, if the materials are constant, the strength
[yield] of the wheel is not increasing with tension either - there's
simply no mechanism for it. does a spring get stronger the more load
you place on it? the force exerted by the spring is more, but the
strength, as defined by displacement per unit load, remains the same.

That graph is good to show the role of spoke tension, but you aren't
getting the entire story because you are ignoring an important part of
the graph. As you point out, the left side of the graph is flat,
showing that the stiffness of the wheel under a fixed load is constant
- but only within that range of spoke tensions. In the right side of
the graph, at lower tensions, the graph shoots upward, showing that
under a certain amount of spoke tension, the deflection of the wheel
increases. Why should this be? Clearly, the elasticity of the
individual wheel components is the same. But, when the spoke tension
is too low, some of the spokes go completely slack when a load is
applied, and when the rim is no longer constrained by the stiffness of
these spokes, the rim is free to deflect.

This is the important consequence when the spoke tension is low - the
rim can deflect (strain) more under a given load. If the load
increases and the rim strains too far, it will permanently deform
(yield) - in other words, the wheel will fail. Higher tension in the
spokes increases the amount of load required to slacken the spokes,
keeping most of the deflection in the high stiffness region, thereby
increasing the load required to reach the rim's yield strain. Because
the stiffness of the spokes support the rim both horizontally and
vertically, the magnitude of the tension in the spokes plays a role in
the strength of the wheel both vertically and horizontally.

I have performed an experiment which demonstrates the importance of
spoke tension on wheel strength. I had two front wheels built with
equivalent components, one with high spoke tension and one with low
spoke tension. I laid the high spoke tension wheel on the ground,
resting on one the end of the axle. I then stepped onto the rim with
my feet on opposite sides (180 degrees apart) and with my entire wheel
on the wheel, I balanced the wheel on the end of the axle (I pressed
my hands against a wall in order to stay balanced on the axle, but my
entire weight was borne at two points on opposite points of the rim).
This represented an extreme case of Rinard's lateral load test. The
wheel easily bore my weight, and showed no ill aftereffects from this
test. I then repeated the test with the low tension wheel. This
time, the rim collapsed, and yielded into a saddle shape (aka, the
wheel tacoed, to use the common vernacular). The wheel with the
higher spoke tension was clearly stronger.

Strength to vertical forces should be similarly affected by spoke
tension. Unfortunately, I don't have a way to generate repeatable
forces high enough to test wheel vertical strength.

When you consider that external forces on the wheel primarily cause
spoke tension decreases, and that when the spokes are completely
slackened they no longer support the rim, and that an unsupported rim
can be yielded more easily than a well supported rim, it is clear that
the magnitude of the spoke tension plays a vital role in the ultimate
strength of a wheel.

Mark McMaster
[email hidden]

mark [& chalo]

think big picture a sec. how does just increasing pre-load change a
material's ultimate load capacity? the only way that can happen is if
the material is a variable, which is clearly not the case. that's why
the graph is flat line. increasing the deflection force just shifts the
position of the line, not its slope.

Actually, I think you are concentrating on the small
picture, which perhaps keeps you from seeing the big
picture. Yes, it is true that applying a pre-load to an
individual component can not change it's ultimate strength.
But we are talking about the strength of a structure, not
its individual components. Yes, increasing spoke tension
does in fact decrease the amount of additional load the
spokes can take in tension, and decreases the amount of
additional load the rim can take in circumferential
compression - but these are typically not the failure modes
of a wheel under use.

Riding loads applied to a wheel in use cause the rim to be
loaded in bending (and the spokes to be loaded in
tension/compression). Lightweight rims (<400 grams) only
have a radial (bending) strength of 200 lb at best. Yet
when the rim is built into a wheel, the wheel can have a
strength of 800 lb. or more. How can this be? It is
because the spokes support the rim, and the pre-load and
high stiffness of the spokes limit the amount the rim bends
inward under an applied load, allowing the rim bear far more
load before failing.

As long as the spokes remain under tension, they will
support the rim, and prevent it from failing under a bending
load. A radial load on a wheel causes the spokes nearest
the load to reduce tension. If and when a spoke losses all
its static tension, it will no longer support the rim, and
the rim will be free to bend under additional load. When
the load exceeds both the static tension of the adjacent
spokes and the bending strength of the rim, the rim will
yield and the wheel will fail. A higher static tension in
the spokes will increase the point at which the spokes
completely de-tension, and correspondingly increase the load
at which the wheel will fail.

Back to the graph - you haven't explained what you think is
going on on the right side of the graph. With the same
applied load, the deflection increases as the spoke tension
decreases. And as we know, if the rim deflects too far, it
will yield. Would you not agree the right most data point
shows a wheel closer to failure than the data points on the
left?

Quoted message said:

regarding the taco wheel experiment, i too have tried that. my
experience was different to yours in that the rim itself makes a huge
difference. spoke tension, within moderate levels of rim max spec, does
not. excess tension makes it /easy/ to taco the wheel. when you say
"eqivalent components" were used in your two wheels, can you confirm
that the rims were identical? and what were the spoke tensions?

No, the rims weren't completely identical, but were very
similar. The rim on the low tension wheel was a Mavic Open
4, and the rim on the high tension wheel was a Sun M14A.
Both rims were approx. 425 grams, both were 19mm wide, and
18mm deep. If anything the rim on the failed wheel (Mavic
Open 4) has a better reputation for strength than the
non-failed wheel (Sun M14A).

Mark McMaster
[email hidden]

Quoted message said:

This summer I snapped a newish 105 hub flange/shell across 4 spokes.

I've seen a failure like that only once before, but it was on a
radially-laced wheel.

--
"Bicycling is a healthy and manly pursuit with much
to recommend it, and, unlike other foolish crazes,
it has not died out." -- The Daily Telegraph (1877)

Mark McMaster said:

jim beam said:

Mark McMaster said:

jim beam wrote:

> Todd Bryan wrote:
>
>>
>> People who have read and understood the years of archived advice
>> from this
>> forum know that high tension does not improve wheel rigidity and
>> they know
>> that high tension does not improve spoke fatigue resistance. They
>> also
>> know that neither of these benefits has ever been claimed. The only
>> claim is that high tension increases the ultimate strength of the
>> wheel.
>> This claim has been repeatedly shown to be true. Please explain
>> how this
>> proven relationship between spoke tension and wheel strength is
>> flawed.
>>
>>
>
> look at this graph.
>
> http://www.damonrinard.com/wheel/tension.gif
>
> origin:
> http://www.damonrinard.com/wheel/index.htm
>
> see how the front end of the graph is a flat line vs. tension?
> that's because the stiffness of the wheel is a function of the
> modulus of the components used, not the tension. if the components
> are constant and remain below yield, there's no way stiffness can
> change.
>
> by the same rationale, if the materials are constant, the strength
> [yield] of the wheel is not increasing with tension either - there's
> simply no mechanism for it. does a spring get stronger the more
> load you place on it? the force exerted by the spring is more, but
> the strength, as defined by displacement per unit load, remains the
> same.

That graph is good to show the role of spoke tension, but you aren't
getting the entire story because you are ignoring an important part
of the graph. As you point out, the left side of the graph is flat,
showing that the stiffness of the wheel under a fixed load is
constant - but only within that range of spoke tensions. In the
right side of the graph, at lower tensions, the graph shoots upward,
showing that under a certain amount of spoke tension, the deflection
of the wheel increases. Why should this be? Clearly, the elasticity
of the individual wheel components is the same. But, when the spoke
tension is too low, some of the spokes go completely slack when a
load is applied, and when the rim is no longer constrained by the
stiffness of these spokes, the rim is free to deflect.

This is the important consequence when the spoke tension is low - the
rim can deflect (strain) more under a given load. If the load
increases and the rim strains too far, it will permanently deform
(yield) - in other words, the wheel will fail. Higher tension in the
spokes increases the amount of load required to slacken the spokes,
keeping most of the deflection in the high stiffness region, thereby
increasing the load required to reach the rim's yield strain.
Because the stiffness of the spokes support the rim both horizontally
and vertically, the magnitude of the tension in the spokes plays a
role in the strength of the wheel both vertically and horizontally.

I have performed an experiment which demonstrates the importance of
spoke tension on wheel strength. I had two front wheels built with
equivalent components, one with high spoke tension and one with low
spoke tension. I laid the high spoke tension wheel on the ground,
resting on one the end of the axle. I then stepped onto the rim with
my feet on opposite sides (180 degrees apart) and with my entire
wheel on the wheel, I balanced the wheel on the end of the axle (I
pressed my hands against a wall in order to stay balanced on the
axle, but my entire weight was borne at two points on opposite points
of the rim). This represented an extreme case of Rinard's lateral
load test. The wheel easily bore my weight, and showed no ill
aftereffects from this test. I then repeated the test with the low
tension wheel. This time, the rim collapsed, and yielded into a
saddle shape (aka, the wheel tacoed, to use the common vernacular).
The wheel with the higher spoke tension was clearly stronger.

Strength to vertical forces should be similarly affected by spoke
tension. Unfortunately, I don't have a way to generate repeatable
forces high enough to test wheel vertical strength.

When you consider that external forces on the wheel primarily cause
spoke tension decreases, and that when the spokes are completely
slackened they no longer support the rim, and that an unsupported rim
can be yielded more easily than a well supported rim, it is clear
that the magnitude of the spoke tension plays a vital role in the
ultimate strength of a wheel.

Mark McMaster
[email hidden]

mark [& chalo]

think big picture a sec. how does just increasing pre-load change a
material's ultimate load capacity? the only way that can happen is if
the material is a variable, which is clearly not the case. that's why
the graph is flat line. increasing the deflection force just shifts
the position of the line, not its slope.

Actually, I think you are concentrating on the small picture, which
perhaps keeps you from seeing the big picture. Yes, it is true that
applying a pre-load to an individual component can not change it's
ultimate strength. But we are talking about the strength of a
structure, not its individual components. Yes, increasing spoke tension
does in fact decrease the amount of additional load the spokes can take
in tension, and decreases the amount of additional load the rim can take
in circumferential compression - but these are typically not the failure
modes of a wheel under use.

Riding loads applied to a wheel in use cause the rim to be loaded in
bending (and the spokes to be loaded in tension/compression).
Lightweight rims (<400 grams) only have a radial (bending) strength of
200 lb at best. Yet when the rim is built into a wheel, the wheel can
have a strength of 800 lb. or more. How can this be? It is because the
spokes support the rim, and the pre-load and high stiffness of the
spokes limit the amount the rim bends inward under an applied load,
allowing the rim bear far more load before failing.

As long as the spokes remain under tension, they will support the rim,
and prevent it from failing under a bending load.

that's not what i've seen. i don't advocate this, but i have a friend
that rode mountain for /months/ on a wheel with spokes so loose, the
thing literally made a grinding noise as the hub center shifted while
rolling. that wheel remained as straight as any i've seen, and believe
me, it was /not/ babied.

Quoted message said:

A radial load on a
wheel causes the spokes nearest the load to reduce tension. If and when
a spoke losses all its static tension, it will no longer support the
rim, and the rim will be free to bend under additional load. When the
load exceeds both the static tension of the adjacent spokes and the
bending strength of the rim, the rim will yield and the wheel will
fail. A higher static tension in the spokes will increase the point at
which the spokes completely de-tension,

we agree.

Quoted message said:

and correspondingly increase the
load at which the wheel will fail.

we don't agree. that's why i used the pre-stressed concrete example.
high tension in the reinforcing bars pushes the onset of cracking up the
graph, but it subtracts from the ultimate load the piece can bear. the
piece /does/ need to be pre-loaded, but no more than necessary for the
reason above, particularly as there is no overall strength benefit.

Quoted message said:

Back to the graph - you haven't explained what you think is going on on
the right side of the graph. With the same applied load, the deflection
increases as the spoke tension decreases.

right, but that graph's "knee" is where the spokes are slack. it's no
different to having play in a wheel bearing with insufficient preload.

Quoted message said:

And as we know, if the rim
deflects too far, it will yield. Would you not agree the right most
data point shows a wheel closer to failure than the data points on the
left?

being as the rim is unsupported, yes. it's the same as any cantilever
vs simple beam. and if i'm not being clear about that, my mistake. i'm
_not_ saying that spokes need to be slack. i'm saying that, as observed
by the flat line portion of the graph, deflection is independant of
pre-load - again, a cantilever where the tensile component is obeying
Hookes Law. excess pre-load reduces overall load capacity, both in
static & dynamic [fatigue] terms.

Quoted message said:

Quoted message said:

regarding the taco wheel experiment, i too have tried that. my
experience was different to yours in that the rim itself makes a huge
difference. spoke tension, within moderate levels of rim max spec,
does not. excess tension makes it /easy/ to taco the wheel. when you
say "eqivalent components" were used in your two wheels, can you
confirm that the rims were identical? and what were the spoke tensions?

No, the rims weren't completely identical, but were very similar. The
rim on the low tension wheel was a Mavic Open 4, and the rim on the high
tension wheel was a Sun M14A. Both rims were approx. 425 grams, both
were 19mm wide, and 18mm deep. If anything the rim on the failed wheel
(Mavic Open 4) has a better reputation for strength than the non-failed
wheel (Sun M14A).

they are indeed similar - thanks for sharing.

Quoted message said:

Mark McMaster
[email hidden]

mark, let me say how much i appreciate a reasoned debate with you. thanks!