SAPIM Spoke FAtigue Estimate

[[email protected]]

Is it not odd that people don't record the "actual" fatique test
results of these Sapim and other spokes. I find it strange that bike
enthusiasts don't club round and make a site that displays actual real
life tests on various spoke/build/rim combinations. You could have a
simple data base of rim, spoke and hand/machine build information
which has come from individual riders having actually riden these
combos.
I have just had a Sapim "Strong" 13 single butted spokes
and washered wheel, using a stainless steel double box section
westrick and Sturmey AW hub built up by Fairlight cycles. I'm going to
keep a log and see for definite how many miles i actually get. I have
just got 800 miles from a 13g terrible handbuild before spoke
breakage(with extremely light road riding), so that's my bench mark.
Wheel two is on 300 miles using a better builder and 14 DT straights.
Hopefully we'll see 1k this time round.

If someone wanted to end this apparently perpetual discussion over
spoke life, it would be great if a novice could go to a dedicated
website and look-up a section that tells him that if he uses x rim and
y spokes in conjuction with handbuild/machine build and a normal load,
then he should, on average expect z miles before spoke breakage. I
realise there are numerous extraneous variables, but a rough average
would be possible for each combination. Maybe if that happened, some
of the spoke manufacturers would get miffed because real life would
show conclusively that their spokes are inferior.
I reckon it would save a lot of time for people so that they could
avoid making the same mistakes every time, though the bike industry
would probably suffer financially.
Has anyone actually tried this before and found a repeatable average
for any one particular spoke/rim combo?
cheers, Nick.

 

[jim beam]

[email protected] wrote:
> Is it not odd that people don't record the "actual" fatique test
> results of these Sapim and other spokes. I find it strange that bike
> enthusiasts don't club round and make a site that displays actual real
> life tests on various spoke/build/rim combinations. You could have a
> simple data base of rim, spoke and hand/machine build information
> which has come from individual riders having actually riden these
> combos.

it's odd, but not unexpected - after all, you can argue opinion, it's
much harder to argue fact. if you don't collect data, you can inject
[under-informed] opinion into the void.

> I have just had a Sapim "Strong" 13 single butted spokes
> and washered wheel, using a stainless steel double box section
> westrick and Sturmey AW hub built up by Fairlight cycles. I'm going to
> keep a log and see for definite how many miles i actually get. I have
> just got 800 miles from a 13g terrible handbuild before spoke
> breakage(with extremely light road riding), so that's my bench mark.
> Wheel two is on 300 miles using a better builder and 14 DT straights.
> Hopefully we'll see 1k this time round.
>
> If someone wanted to end this apparently perpetual discussion over
> spoke life, it would be great if a novice could go to a dedicated
> website and look-up a section that tells him that if he uses x rim and
> y spokes in conjuction with handbuild/machine build and a normal load,
> then he should, on average expect z miles before spoke breakage. I
> realise there are numerous extraneous variables, but a rough average
> would be possible for each combination. Maybe if that happened, some
> of the spoke manufacturers would get miffed because real life would
> show conclusively that their spokes are inferior.
> I reckon it would save a lot of time for people so that they could
> avoid making the same mistakes every time, though the bike industry
> would probably suffer financially.
> Has anyone actually tried this before and found a repeatable average
> for any one particular spoke/rim combo?
> cheers, Nick.
>

 

[Ben C]

On 2007-02-09, Gary Young <[email protected]> wrote:
[...]
> Imagine you have spokes with a perfect line -- they sit snugly in the
> holes, hug closely to the flange as they exit the hub, and enter the
> nipples at the same angle at which the nipples are seated in the rim. In
> that case, where would the bending stress come from?

In this configuration, assuming everything's perfect, I think you still
have bending, but the perpendicular distance from the force direction to
the fulcrum is half the diameter of the spoke, at the centre of the
spoke. Pulling on the spoke will result in compression on the inside of
the elbow and tension on the outside-- it is bending.

In reality though the moment is likely to be somewhat greater than this,
either due to less than perfect flushness, or due to a more complicated
shape (like a question-mark hook) near where the spoke exits the flange.

> In what way would it matter if the spoke is straight gauge or butted?
> The spoke would change in length as it is loaded and unloaded, but
> would it bend?

In the perfect or near-perfect situation, I would have thought no
bending at the midsection during loading and unloading provided the
spoke doesn't go slack. But you still have bending stresses at the
elbow.

If the spoke is unloaded so far that it goes slack, then it will bend--
pushing two ends of an unsupported wire together is "unstable" in the
sense that it will bend one way or another, not compress perfectly along
its length.

In that situation, being butted rather than PG might reduce elbow flex.

> (I confess that my theory is all wet if such axial changes are
> considered bending.) Do straight-pull, elbow-less spokes experience
> bending stress?

If they don't go slack there should be no bending.

 

[Michael Press]

In article
<[email protected]>
,
[email protected] wrote:

> Is it not odd that people don't record the "actual" fatique test
> results of these Sapim and other spokes. I find it strange that bike
> enthusiasts don't club round and make a site that displays actual real
> life tests on various spoke/build/rim combinations. You could have a
> simple data base of rim, spoke and hand/machine build information
> which has come from individual riders having actually riden these
> combos.
> I have just had a Sapim "Strong" 13 single butted spokes
> and washered wheel, using a stainless steel double box section
> westrick and Sturmey AW hub built up by Fairlight cycles. I'm going to
> keep a log and see for definite how many miles i actually get. I have
> just got 800 miles from a 13g terrible handbuild before spoke
> breakage(with extremely light road riding), so that's my bench mark.
> Wheel two is on 300 miles using a better builder and 14 DT straights.
> Hopefully we'll see 1k this time round.
>
> If someone wanted to end this apparently perpetual discussion over
> spoke life, it would be great if a novice could go to a dedicated
> website and look-up a section that tells him that if he uses x rim and
> y spokes in conjuction with handbuild/machine build and a normal load,
> then he should, on average expect z miles before spoke breakage. I
> realise there are numerous extraneous variables, but a rough average
> would be possible for each combination. Maybe if that happened, some
> of the spoke manufacturers would get miffed because real life would
> show conclusively that their spokes are inferior.
> I reckon it would save a lot of time for people so that they could
> avoid making the same mistakes every time, though the bike industry
> would probably suffer financially.
> Has anyone actually tried this before and found a repeatable average
> for any one particular spoke/rim combo?

A wheel can be made well or poorly. Well enough to
almost entirely compensate for deficient components.
The one deficiency that a thoughtful and well executed
build cannot compensate for is weakness in the rim at
the spoke hole.

How long have you been reading here? Many many people
have testified to the overwhelming benefits derived
from building well and thoughtfully. Many hub-rim-spoke
combinations have been reported upon by professional
and amateur.

Do an archive search, list the reports with citation
and write back with the URL.

--
Michael Press

 

[jim beam]

Gary Young wrote:
<snip for brevity>

> Correcting the spoke line makes the spoke straighter along its unsupported
> length. It brings the elbow-end of the spoke into closer conformity with
> the hub flange, thus reducing the range of bending that's possible.

no. offset is offset is offset. you have to grasp this fundamental
concept to be able to understand anything further.

> It
> seems to me that it has everything to do with the question at hand. If it
> weren't important, why in the past have you emphasized so much the pains
> that spoke makers take to ensure that spoke line is optimized?

this is a separate issue - this is static bending for conformity with
build geometry, not dynamic bending as a function of load.

>
> I think you've unwittingly committed yourself to a position in which you
> essentially agree with Jobst that misaligned spoke lines are a problem
> even after tensioning and bedding-in.

no gary - it's got nothing to do with "correcting the spoke line" so
don't put those words in my mouth. you have to understand the
principles. a straight line is the shortest distance between two
points. a j-bend spoke does not take the straight line path between the
two loading points, it's offset. not my much, only about 3mm, but
that's enough to create a bending moment.

i'm snipping the rest because it's pointless if the above is not understood.

 

[jim beam]

Michael Press wrote:
> In article
> <[email protected]>
> ,
> [email protected] wrote:
>
>> Is it not odd that people don't record the "actual" fatique test
>> results of these Sapim and other spokes. I find it strange that bike
>> enthusiasts don't club round and make a site that displays actual real
>> life tests on various spoke/build/rim combinations. You could have a
>> simple data base of rim, spoke and hand/machine build information
>> which has come from individual riders having actually riden these
>> combos.
>> I have just had a Sapim "Strong" 13 single butted spokes
>> and washered wheel, using a stainless steel double box section
>> westrick and Sturmey AW hub built up by Fairlight cycles. I'm going to
>> keep a log and see for definite how many miles i actually get. I have
>> just got 800 miles from a 13g terrible handbuild before spoke
>> breakage(with extremely light road riding), so that's my bench mark.
>> Wheel two is on 300 miles using a better builder and 14 DT straights.
>> Hopefully we'll see 1k this time round.
>>
>> If someone wanted to end this apparently perpetual discussion over
>> spoke life, it would be great if a novice could go to a dedicated
>> website and look-up a section that tells him that if he uses x rim and
>> y spokes in conjuction with handbuild/machine build and a normal load,
>> then he should, on average expect z miles before spoke breakage. I
>> realise there are numerous extraneous variables, but a rough average
>> would be possible for each combination. Maybe if that happened, some
>> of the spoke manufacturers would get miffed because real life would
>> show conclusively that their spokes are inferior.
>> I reckon it would save a lot of time for people so that they could
>> avoid making the same mistakes every time, though the bike industry
>> would probably suffer financially.
>> Has anyone actually tried this before and found a repeatable average
>> for any one particular spoke/rim combo?
>
> A wheel can be made well or poorly. Well enough to
> almost entirely compensate for deficient components.
> The one deficiency that a thoughtful and well executed
> build cannot compensate for is weakness in the rim at
> the spoke hole.
>
> How long have you been reading here? Many many people
> have testified to the overwhelming benefits derived
> from building well and thoughtfully. Many hub-rim-spoke
> combinations have been reported upon by professional
> and amateur.

hold on a second. if you go to buy a spoke from your lbs, what do you
receive? chances are, it's a spoke made by d.t., wheelsmith or sapim.
all are well made, utilize very high quality materials, and their use
has a significant impact on fatigue results.

now, compare that to r.b.t reports of spoke failures. almost all are
from "pre-built" wheels with no-name spokes. rebuilding with branded
spokes seems always followed by deafening silence, i.e. no further
breakages.

coincidence?

also bear in mind that these quality branded spokes have been ubiquitous
for about 30 years and have pretty much a lock on the domestic market -
there's not much data on rebuilds with off-brand spokes.

>
> Do an archive search, list the reports with citation
> and write back with the URL.
>

backatcha.

 

[[email protected]]

On Fri, 09 Feb 2007 01:55:58 -0600, Gary Young <[email protected]>
wrote:

[snip]

>> For one thing, the different-thickness straight spoke blanks probably
>> start out with different initial strengths due to the work hardening
>> effect that got them to their initial size. That is, a 2.3 mm straight
>> spoke probably starts out with a different strength than a 2.0 mm
>> straight spoke.
>>
>Are you suggesting that the Strong spokes are work-hardened up from a
>smaller diameter? Do you really think that's plausible? Why not just start
>with a 2.3 mm wire and then draw down the thin section?
>
>> Cheers,
>>
>> Carl Fogel

Dear Gary,

I'm baffled by your question.

Where do you see any suggestion that Strong spokes are work-hardened
"up from a smaller diameter"?

As I said, a 2.3 mm straight spoke blank probably starts out with a
different strength than a 2.0 mm straight spoke blank.

Both spokes have been drawn down from thicker original stock. The 2.0
mm spoke section must obviously be drawn down thinner than the 2.3 mm
spoke section.

In wire-drawing operations, the wire stock is repeatedly drawn down.
Since it work-hardens in the process, it eventually has to be annealed
before further drawing.

We don't know the initial state of the 2.3 mm straight spoke blank
before it was drawn down to a 2.0 midsection and rim end.

We do know what Sapim's strength tests measured.

Cheers,

Carl Fogel

 

[[email protected]]

On 9 Feb 2007 07:45:57 -0800, [email protected] wrote:

[snip]

> Has anyone actually tried this before and found a repeatable average
>for any one particular spoke/rim combo?
> cheers, Nick.

Dear Nick,

Briefly, no.

No one has found a repeatable average for any particular spoke/rim
combo.

Even in accelerated testing with exaggerated loads running night and
day at the equivalent of 45 mph, it took about two months to break a
sample set of 78 spokes in the early 1980's when Stanford tested
spokes for WheelSmith.

More modern spokes are thought to be much more durable, so testing
would take even longer.

If you ask how long spokes will last if unsqueezed, you'll just get a
lot of well-it-depends and hard-to-say and doesn't-matter and other
responses that highlight how rarely spokes break and how anecdotal the
data is. You just end up with a vague "they'll break sooner but mine
never break" warning.

Of course, anyone who disagrees with my view of the situation is free
to predict whether unsqueezed spokes will last 5,000, 10,000, 25,000,
50,000, 100,000 or 300,000 miles and explain the basis for their
prediction.

In any case, the effort required for collecting a vast database of
poorly qualified anecdotal information is probably far more than the
effort to replace an occasional broken spoke.

Cheers,

Carl Fogel

 

[Johan Bornman]

On Feb 9, 9:25 pm, jim beam <[email protected]> wrote:
>
> no [Gary]. offset is offset is offset. you have to grasp this fundamental
> concept to be able to understand anything further.
>
Jim, I don't understand your "offset" concept. The fact that the spoke
comes out of the flange 3mm (or whatever) off centre from the flange,
doesn't mean it cannot still follow a straight line to the rim (slight
detour at cross-over negated). We're talking about a straight line
here, essentially from the elbow's exit point to the nipple. Where is
the bending in that?

> > It
> > seems to me that it has everything to do with the question at hand. If it
> > weren't important, why in the past have you emphasized so much the pains
> > that spoke makers take to ensure that spoke line is optimized?
>
> this is a separate issue - this is static bending for conformity with
> build geometry, not dynamic bending as a function of load.
>
>
>
> > I think you've unwittingly committed yourself to a position in which you
> > essentially agree with Jobst that misaligned spoke lines are a problem
> > even after tensioning and bedding-in.
>
> no gary - it's got nothing to do with "correcting the spoke line" so
> don't put those words in my mouth. you have to understand the
> principles. a straight line is the shortest distance between two
> points. a j-bend spoke does not take the straight line path between the
> two loading points, it's offset. not my much, only about 3mm, but
> that's enough to create a bending moment.


Jim, explain where that bending moment is. As Gary has already
explained, we're assuming a good flush exit from the hub, where's the
offset? The spoke travels in a straight line.

 

[Johan Bornman]

On Feb 9, 9:43 pm, jim beam <[email protected]> wrote:
>
> hold on a second. if you go to buy a spoke from your lbs, what do you
> receive? chances are, it's a spoke made by d.t., wheelsmith or sapim.
> all are well made, utilize very high quality materials, and their use
> has a significant impact on fatigue results.

How does their use have a significant impact on fatigue results. A
cycle is a cycle. Are you suggesting that abuse of wheels has an
impact on fatigue failures? Please explain, I don't understand "use"
in this context.


> now, compare that to r.b.t reports of spoke failures. almost all are
> from "pre-built" wheels with no-name spokes. rebuilding with branded
> spokes seems always followed by deafening silence, i.e. no further
> breakages.
>
> coincidence?

I think you're drawing conclusions here from suspect data.

Johan Bornman

 

[jim beam]

Johan Bornman wrote:
> On Feb 9, 9:25 pm, jim beam <[email protected]> wrote:
>> no [Gary]. offset is offset is offset. you have to grasp this fundamental
>> concept to be able to understand anything further.
>>
> Jim, I don't understand your "offset" concept. The fact that the spoke
> comes out of the flange 3mm (or whatever) off centre from the flange,
> doesn't mean it cannot still follow a straight line to the rim (slight
> detour at cross-over negated). We're talking about a straight line
> here, essentially from the elbow's exit point to the nipple. Where is
> the bending in that?
>
>>> It
>>> seems to me that it has everything to do with the question at hand. If it
>>> weren't important, why in the past have you emphasized so much the pains
>>> that spoke makers take to ensure that spoke line is optimized?
>> this is a separate issue - this is static bending for conformity with
>> build geometry, not dynamic bending as a function of load.
>>
>>
>>
>>> I think you've unwittingly committed yourself to a position in which you
>>> essentially agree with Jobst that misaligned spoke lines are a problem
>>> even after tensioning and bedding-in.
>> no gary - it's got nothing to do with "correcting the spoke line" so
>> don't put those words in my mouth. you have to understand the
>> principles. a straight line is the shortest distance between two
>> points. a j-bend spoke does not take the straight line path between the
>> two loading points, it's offset. not my much, only about 3mm, but
>> that's enough to create a bending moment.
>
>
> Jim, explain where that bending moment is. As Gary has already
> explained, we're assuming a good flush exit from the hub, where's the
> offset? The spoke travels in a straight line.
>

it's all about the offset johan. one end is not loaded axially to the
spoke.

 

[jim beam]

Johan Bornman wrote:
> On Feb 9, 9:43 pm, jim beam <[email protected]> wrote:
>> hold on a second. if you go to buy a spoke from your lbs, what do you
>> receive? chances are, it's a spoke made by d.t., wheelsmith or sapim.
>> all are well made, utilize very high quality materials, and their use
>> has a significant impact on fatigue results.
>
> How does their use have a significant impact on fatigue results. A
> cycle is a cycle. Are you suggesting that abuse of wheels has an
> impact on fatigue failures? Please explain, I don't understand "use"
> in this context.

a number of factors affect fatigue - these include design, material
quality and surface finish. with a quality branded spoke, all 3 bases
are covered.

>
>
>> now, compare that to r.b.t reports of spoke failures. almost all are
>> from "pre-built" wheels with no-name spokes. rebuilding with branded
>> spokes seems always followed by deafening silence, i.e. no further
>> breakages.
>>
>> coincidence?
>
> I think you're drawing conclusions here from suspect data.

no, i'm drawing conclusions from materials science.

 

[Gary Young]

On Fri, 09 Feb 2007 11:12:40 -0600, Ben C wrote:

> On 2007-02-09, Gary Young <[email protected]> wrote:
> [...]
>> Imagine you have spokes with a perfect line -- they sit snugly in the
>> holes, hug closely to the flange as they exit the hub, and enter the
>> nipples at the same angle at which the nipples are seated in the rim. In
>> that case, where would the bending stress come from?
>
> In this configuration, assuming everything's perfect, I think you still
> have bending, but the perpendicular distance from the force direction to
> the fulcrum is half the diameter of the spoke, at the centre of the
> spoke. Pulling on the spoke will result in compression on the inside of
> the elbow and tension on the outside-- it is bending.
>
> In reality though the moment is likely to be somewhat greater than this,
> either due to less than perfect flushness, or due to a more complicated
> shape (like a question-mark hook) near where the spoke exits the flange.

I was hoping someone else would comment, because I don't know enough
to dispute what you say here. However, I still have to wonder how great an
impact stresses at this small a scale have. After all, if you take the
scale down enough, even straight-pull spokes have bending (I think). The
head of the spoke would bend like a diaphragm as the spoke's tension
increases and decreases. I also suspect that the threads would bend as
they are pulled against the mating surfaces inside the nipple.

One other consideration -- based on the technical documents on their
website, I think Sapim believe that the problem is not at the level of
internal changes. In other words, they point to bending that is visible to
the eye. But I'm going to start another thread on that question.

 

[Gary Young]

On Fri, 09 Feb 2007 13:05:57 -0700, carlfogel wrote:

> On Fri, 09 Feb 2007 01:55:58 -0600, Gary Young <[email protected]>
> wrote:
>
> [snip]
>
>>> For one thing, the different-thickness straight spoke blanks probably
>>> start out with different initial strengths due to the work hardening
>>> effect that got them to their initial size. That is, a 2.3 mm straight
>>> spoke probably starts out with a different strength than a 2.0 mm
>>> straight spoke.
>>>
>>Are you suggesting that the Strong spokes are work-hardened up from a
>>smaller diameter? Do you really think that's plausible? Why not just start
>>with a 2.3 mm wire and then draw down the thin section?
>>
>>> Cheers,
>>>
>>> Carl Fogel
>
> Dear Gary,
>
> I'm baffled by your question.
>
> Where do you see any suggestion that Strong spokes are work-hardened
> "up from a smaller diameter"?
>
> As I said, a 2.3 mm straight spoke blank probably starts out with a
> different strength than a 2.0 mm straight spoke blank.
>
> Both spokes have been drawn down from thicker original stock. The 2.0
> mm spoke section must obviously be drawn down thinner than the 2.3 mm
> spoke section.
>
> In wire-drawing operations, the wire stock is repeatedly drawn down.
> Since it work-hardens in the process, it eventually has to be annealed
> before further drawing.
>
> We don't know the initial state of the 2.3 mm straight spoke blank
> before it was drawn down to a 2.0 midsection and rim end.
>
> We do know what Sapim's strength tests measured.
>
> Cheers,
>
> Carl Fogel

Carl,
You're right, it was a dumb question and not even pertinent to fatigue,
which is the main thing I want to understand.

I do remain skeptical of your explanation of the Strong's stellar fatigue
life, not least because it seems at odd with some of Sapim's technical
documents. But I plan to start a separate thread on those documents.

 

[Gary Young]

On Mon, 12 Feb 2007 07:09:08 -0800, jim beam wrote:

> Johan Bornman wrote:
>> On Feb 9, 9:25 pm, jim beam <[email protected]> wrote:
>>> no [Gary]. offset is offset is offset. you have to grasp this fundamental
>>> concept to be able to understand anything further.
>>>
>> Jim, I don't understand your "offset" concept. The fact that the spoke
>> comes out of the flange 3mm (or whatever) off centre from the flange,
>> doesn't mean it cannot still follow a straight line to the rim (slight
>> detour at cross-over negated). We're talking about a straight line
>> here, essentially from the elbow's exit point to the nipple. Where is
>> the bending in that?
>>
>>>> It
>>>> seems to me that it has everything to do with the question at hand. If it
>>>> weren't important, why in the past have you emphasized so much the pains
>>>> that spoke makers take to ensure that spoke line is optimized?
>>> this is a separate issue - this is static bending for conformity with
>>> build geometry, not dynamic bending as a function of load.
>>>
>>>
>>>
>>>> I think you've unwittingly committed yourself to a position in which you
>>>> essentially agree with Jobst that misaligned spoke lines are a problem
>>>> even after tensioning and bedding-in.
>>> no gary - it's got nothing to do with "correcting the spoke line" so
>>> don't put those words in my mouth. you have to understand the
>>> principles. a straight line is the shortest distance between two
>>> points. a j-bend spoke does not take the straight line path between the
>>> two loading points, it's offset. not my much, only about 3mm, but
>>> that's enough to create a bending moment.
>>
>>
>> Jim, explain where that bending moment is. As Gary has already
>> explained, we're assuming a good flush exit from the hub, where's the
>> offset? The spoke travels in a straight line.
>>
>
> it's all about the offset johan. one end is not loaded axially to the
> spoke.

Suppose I have a test apparatus that allows me to tension a spoke. The
spoke is firmly clamped at the threads at one end and by the head at the
other. Because only the head is clamped, the elbow itself is not
supported. In that case, I can see your point about offset.

Now, suppose we put a triangular-shaped piece under the spoke somewhere
midspan, so that the spoke bends slightly as it goes over the point of the
triangular. Are you saying that the bending moment would be unchanged in
that situation?

 

[Ben C]

On 2007-02-12, Gary Young <[email protected]> wrote:
> On Fri, 09 Feb 2007 11:12:40 -0600, Ben C wrote:
>
>> On 2007-02-09, Gary Young <[email protected]> wrote:
>> [...]
>>> Imagine you have spokes with a perfect line -- they sit snugly in the
>>> holes, hug closely to the flange as they exit the hub, and enter the
>>> nipples at the same angle at which the nipples are seated in the rim. In
>>> that case, where would the bending stress come from?
>>
>> In this configuration, assuming everything's perfect, I think you still
>> have bending, but the perpendicular distance from the force direction to
>> the fulcrum is half the diameter of the spoke, at the centre of the
>> spoke. Pulling on the spoke will result in compression on the inside of
>> the elbow and tension on the outside-- it is bending.
>>
>> In reality though the moment is likely to be somewhat greater than this,
>> either due to less than perfect flushness, or due to a more complicated
>> shape (like a question-mark hook) near where the spoke exits the flange.
>
> I was hoping someone else would comment, because I don't know enough
> to dispute what you say here. However, I still have to wonder how great an
> impact stresses at this small a scale have.

Good point. The smaller the moment the further the tensile stress on the
outside of the elbow is from yield. The further it is from yield, the
longer the fatigue life. That's why long-shank spokes are bad.

You don't need a big offset for spoke tension to bring the elbow to
yield, but the distances are small and the more complex aspects of the
geometry become significant. I went through all this when trying to
understand how the elbow might bend plastically during the build, and
the role of stress relief. My conclusion was that whether the final
configuration is one in which spoke tension is enough to hold the elbow
at near yield stress is not obvious and may vary from wheel to wheel.
It's a factor in whether you need to stress-relieve or not (but you
might as well stress-relieve), but even if you do stress-relieve, the
larger that moment the worse the fatigue's going to be in use anyway.

> After all, if you take the
> scale down enough, even straight-pull spokes have bending (I think). The
> head of the spoke would bend like a diaphragm as the spoke's tension
> increases and decreases.

I think I see what you mean. Not sure if that counts as "bending", but
whatever it is, I would think the mean stress of the cycle would be much
lower than yield.

> I also suspect that the threads would bend as
> they are pulled against the mating surfaces inside the nipple.

Probably. I'm less clear of the mechanisms of fatigue at the threads
(although it is another place where spokes fail quite a lot).

 

[Ron Ruff]

On Feb 9, 12:25 pm, jim beam <[email protected]> wrote:
> no. offset is offset is offset. you have to grasp this fundamental
> concept to be able to understand anything further.
>
> this is a separate issue - this is static bending for conformity with
> build geometry, not dynamic bending as a function of load.
>
> a straight line is the shortest distance between two
> points. a j-bend spoke does not take the straight line path between the
> two loading points, it's offset. not my much, only about 3mm, but
> that's enough to create a bending moment.

I guess I don't understand it either, jim. If the spoke is lined up as
well as possible (laying on the flange) it will have that offset, but
the offset will be constant. The line that the spoke takes to the rim
will be constant as well... so the only dynamic variation in load will
be the same one that the rest of the spoke sees. It *will* be smaller
if the spoke is butted, because the center of the spoke will stretch
more... but I can't see why the *bending moment* of the middle of the
spoke should matter, because the load is 99.99% axial.

If Sapim's test had the spoke sticking out from the flange, with a
load cycling from 0-80kg then I'd expect the results they got... but
my wheels aren't like that, and they don't see that kind of loading.
In normal riding NDS rear, ~70-50kg, DS rear ~115-105kg. If anything
their test should illustrate that 0 tension isn't a good thing to see
on a spoke.

 

[jim beam]

Ron Ruff wrote:
> On Feb 9, 12:25 pm, jim beam <[email protected]> wrote:
>> no. offset is offset is offset. you have to grasp this fundamental
>> concept to be able to understand anything further.
>>
>> this is a separate issue - this is static bending for conformity with
>> build geometry, not dynamic bending as a function of load.
>>
>> a straight line is the shortest distance between two
>> points. a j-bend spoke does not take the straight line path between the
>> two loading points, it's offset. not my much, only about 3mm, but
>> that's enough to create a bending moment.
>
> I guess I don't understand it either, jim. If the spoke is lined up as
> well as possible (laying on the flange) it will have that offset, but
> the offset will be constant.

the offset is constant, but not the bending force imposed as load varies.

> The line that the spoke takes to the rim
> will be constant as well... so the only dynamic variation in load will
> be the same one that the rest of the spoke sees.

no.

imagine you have a builder's "t" square, and you drill holes on the
farthest parts of the arms so you can hang it at one end and load
weights on the other. you can see that the middle experiences a bending
moment with tension on the inside of the elbow, compression on the
outside, right?

now cut one inch off the end of one arm, re-drill, and repeat. still
bending, right? less, but it's still there.

repeat until you have one arm only 3mm long. the bending moment reduces
each time, but it still exists and will continue to exist until the
two loading points are exactly axial.

> It *will* be smaller
> if the spoke is butted, because the center of the spoke will stretch
> more...

stop and think about that a moment. if i had a black box with a hook
poking out of the end and a spring scale attached to the hook, and i
exerted force so that the scale showed 10lbs force, what's in the box?

> but I can't see why the *bending moment* of the middle of the
> spoke should matter, because the load is 99.99% axial.

but it's not 100%.

>
> If Sapim's test had the spoke sticking out from the flange, with a
> load cycling from 0-80kg then I'd expect the results they got...

what is the expectation? how are sapim spokes different from say
wheelsmith or d.t.?

> but
> my wheels aren't like that, and they don't see that kind of loading.
> In normal riding NDS rear, ~70-50kg, DS rear ~115-105kg. If anything
> their test should illustrate that 0 tension isn't a good thing to see
> on a spoke.
>
industry standard is to accelerate fatigue testing. all industries.

bottom line is that it's the small stuff that matters. that's why
metallurgists spend so much time looking don microscopes. surface
finish differences may not be obvious to the naked eye for instance, but
they sure do make a difference to fatigue results.

 

[Ron Ruff]

On Feb 12, 11:50 am, jim beam <[email protected]> wrote:
> repeat until you have one arm only 3mm long. the bending moment reduces
> each time, but it still exists and will continue to exist until the
> two loading points are exactly axial.

The 99.99% axial number, was underestimated... it is actually ~99.995%
axial, if we concede that there is a 3mm offset. But I think you are
oversimplifying the way that a spoke sits in a flange. If the inner
radius of the spoke rests on (or is embedded in) the flange and the
elbow is formed so that it points in the direction of the spoke's
axial load, then is there a 3mm offset? It seems to me that the offset
would be ~half the spoke diameter, or the distance from the spoke
center to the edge of the flange.

> stop and think about that a moment. if i had a black box with a hook
> poking out of the end and a spring scale attached to the hook, and i
> exerted force so that the scale showed 10lbs force, what's in the box?

That's true. If you test by applying a fixed load cycle it won't
matter. In a real wheel this isn't the case though. For a butted spoke
the typical load cycle is lower than for a non-butted one, all else
being equal. A more flexible spoke spreads the load over a greater
portion of the rim.

> what is the expectation? how are sapim spokes different from say
> wheelsmith or d.t.?

They aren't. I was refering to their results which show bladed spokes
lasting longer than butted ones.

> industry standard is to accelerate fatigue testing. all industries.

Makes sense... but the tricky part is applying a load cycle that
mimics real life while also accelerating it. I'm not convinced that
the way their spokes were fixtured at the J was appropriate, and
cycling down to zero load is an abnormal occurrence too. Cycling from
50-150kg would make more sense to me, but maybe they tried that and
the spokes wouldn't break...

 

[jim beam]

Ron Ruff wrote:
> On Feb 12, 11:50 am, jim beam <[email protected]> wrote:
>> repeat until you have one arm only 3mm long. the bending moment reduces
>> each time, but it still exists and will continue to exist until the
>> two loading points are exactly axial.
>
> The 99.99% axial number, was underestimated... it is actually ~99.995%
> axial, if we concede that there is a 3mm offset.

i think you'll find the stress increase in the skin of the elbow is a
good deal greater than 1.005 times load.

> But I think you are
> oversimplifying the way that a spoke sits in a flange. If the inner
> radius of the spoke rests on (or is embedded in) the flange and the
> elbow is formed so that it points in the direction of the spoke's
> axial load, then is there a 3mm offset?

that's a deceptively simple question with a complex answer. it's an
interaction of the stress distribution between the flange, the elbow
shank, the head and of course the spoke proper. i haven't built a
finite element model of a spoke and how it interacts with its anchor,
but as you know, we have a number of facts that lead one to think this
is not a straightforward situation:

1. spokes fatigue at the elbow, with initiation on the inside.
2. spokes fatigue at the elbow, with initiation on the outside.
3. spoke even fatigue under the spoke head.
4. /nothing/ fatigues without reason, even if it's not immediately obvious.

> It seems to me that the offset
> would be ~half the spoke diameter, or the distance from the spoke
> center to the edge of the flange.

that's only if the spoke is hard up against the flange, the flange
doesn't elastically distort on loading, etc.

>
>> stop and think about that a moment. if i had a black box with a hook
>> poking out of the end and a spring scale attached to the hook, and i
>> exerted force so that the scale showed 10lbs force, what's in the box?
>
> That's true. If you test by applying a fixed load cycle it won't
> matter. In a real wheel this isn't the case though. For a butted spoke
> the typical load cycle is lower than for a non-butted one, all else
> being equal.

that's greater elasticity affecting rim interaction.

> A more flexible spoke spreads the load over a greater
> portion of the rim.

well, elasticity allows... see above.

>
>> what is the expectation? how are sapim spokes different from say
>> wheelsmith or d.t.?
>
> They aren't. I was refering to their results which show bladed spokes
> lasting longer than butted ones.

well, my opinion on that subject is as stated /way/ up thread.

>
>> industry standard is to accelerate fatigue testing. all industries.
>
> Makes sense... but the tricky part is applying a load cycle that
> mimics real life while also accelerating it.

that's not realistic. we don't crash planes to test them. we don't
collapse bridges to test them. otoh, accelerated fatigue testing is
very reliable. i know of no situation where fatigue life [cycle count]
suddenly drops off simply as a function of time, hence it's safe to
accelerate. there's over 100 years of very high level research on this
subject - it's not like there's a giant intellectual black hole specific
to bicycles in general and spokes in particular.

> I'm not convinced that
> the way their spokes were fixtured at the J was appropriate,

why? do we know what fixture they use? i know it's a popular position
here on r.b.t, but it's not reasonable to just assume they're goof's
that don't know what they're doing.

> and
> cycling down to zero load is an abnormal occurrence too. Cycling from
> 50-150kg would make more sense to me, but maybe they tried that and
> the spokes wouldn't break...

see above for accelerated fatigue testing. to do anything else is
simply not practical or even reliable since time dictates products would
be a complete generation behind - planes, trains, automobiles...