Lighter gauge spokes on non-drive side?
Erik Voldengen
I've been told that using lighter guage spokes on the non drive
side of the rear wheel will increase the tension on that side,
reducing the difference in tension between the two sides.
BS? I've read "the book," and nothing, not even a mention
of folklore, was written.
--
-Erik
Pete Ruckelshaus
ATB) for years and it's the only way I will build wheels, especially for
myself. Yes, tension seems to be more even (sorry, I don't have a
tension gage, I do it all by "feel" and experience). I'm a heavier
rider and rear wheel strength is pretty ket for me, and with this setup
I have never tacoed a wheel.
Pete
--
Pete's Bikindex - http://www.bikindex.com/
* * A Lycos Top 5% Site - Best of Cycling * *
PPWrench
spoke gauge....different spoke gauges will feel different but as measured by a
tensionometer-the actual tension should be the same-on a properly built
wheel-mixing spoke gauges does nothing to improve a wheel-IMO-I rec. 14-15 dbl
butts(I like DT) in almost all applications....
Peter
ProPeloton
Boulder
Mark Atanowicz
bbc...@user1.teleport.com (Erik Voldengen) writes:
> I've been told that using lighter guage spokes on the non drive
> side of the rear wheel will increase the tension on that side,
> reducing the difference in tension between the two sides.
>
> BS? I've read "the book," and nothing, not even a mention
> of folklore, was written.
The use of thinner spokes on the NDS doesn not increase the tension on
those spokes, but rather the elongation, which is a function of tension
*and* X-sectional area. This increases the rim deformation at which
the spokes go slack, making the wheel more durable. While Jobst has
never acknowledged my repeated claims of this, his non-flaming signals
tacit approval. ;-)
Mark Atanowicz
Chuck Fry
In article <6ns8h3$t5o$1...@user1.teleport.com>,
Erik Voldengen <bbc...@user1.teleport.com> wrote:
>I've been told that using lighter guage spokes on the non drive
>side of the rear wheel will increase the tension on that side,
>reducing the difference in tension between the two sides.
This is BS. The tension in the non-drive side spokes will be the same,
no matter the gauge of the spokes.
However, what will be different is the tension per square millimeter.
The smaller cross-section translates to an increase in unit stress.
According to some books on fasteners for non-engineers that I have read,
what holds threaded fasteners (spokes and nipples) tight is tension in
the threads. The tension causes the threads to deform minutely; the
local deformation is what locks the nut (nipple) in place on the bolt
(spoke).
So for this scheme to work, you would need to use a smaller thread size
on the non-drive side. Just using a smaller cross section is not
enough.
I'd like to try building a rear wheel with 14/15/14 gauge drive side
and 15/16/15 gauge non-drive side spokes. My guess is it will hold its
trueness better than a wheel built with all the same size spokes, or one
built with a mix of butted and non-butted spokes of the same thread
size.
OTOH, maybe I'll just build one with a trick asymmetrical rim instead...
Keith Bontrager sure seems to know his stuff.
-- Chuck
--
Chuck Fry -- Jack of all trades, master of none
chu...@chucko.com (text only please), chuc...@home.com (MIME enabled),
chu...@gateway.idiom.com (SPAM ONLY)
Mark Atanowicz
chu...@best.com (Chuck Fry) writes:
> According to some books on fasteners for non-engineers that I have read,
> what holds threaded fasteners (spokes and nipples) tight is tension in
> the threads. The tension causes the threads to deform minutely; the
> local deformation is what locks the nut (nipple) in place on the bolt
> (spoke).
This local deformation is elastic, meaning that as soon as the load is
removed the threads return to their unloaded state. The "secret" to
keeping spokes from loosening is to keep them tensioned.
> So for this scheme to work, you would need to use a smaller thread size
> on the non-drive side. Just using a smaller cross section is not
> enough.
For the reason stated above, you need a thinner spoke, not a smaller
thread. Thinner spokes help normalize elongation WRT to the drive
side.
> I'd like to try building a rear wheel with 14/15/14 gauge drive side
> and 15/16/15 gauge non-drive side spokes. My guess is it will hold its
> trueness better than a wheel built with all the same size spokes, or one
> built with a mix of butted and non-butted spokes of the same thread
> size.
Try 14/17/14 spokes. They work well on the NDS.
>
> OTOH, maybe I'll just build one with a trick asymmetrical rim instead...
> Keith Bontrager sure seems to know his stuff.
Wheels with asymmetrical rims *and* thin spokes on the NDS would be
nice.
Mark Atanowicz
Chuck Fry
Mark Atanowicz <MAtan...@aol.com> wrote:
>In article <6ntjrl$b8d$1...@shell5.ba.best.com>
>chu...@best.com (Chuck Fry) writes:
>> So for this scheme to work, you would need to use a smaller thread size
>> on the non-drive side. Just using a smaller cross section is not
>> enough.
>
>For the reason stated above, you need a thinner spoke, not a smaller
>thread. Thinner spokes help normalize elongation WRT to the drive
>side.
Right, but it seems to me (remember, pure speculation!) that you need to
increase the unit stress in the threads as well, thus I suspect a
smaller diameter thread (say, 15/16/15) would better withstand the
stress cycles in actual use than the 14/17/14 you suggest.
>> OTOH, maybe I'll just build one with a trick asymmetrical rim instead...
>> Keith Bontrager sure seems to know his stuff.
>
>Wheels with asymmetrical rims *and* thin spokes on the NDS would be
>nice.
Are both necessary?
Mark Atanowicz
chu...@best.com (Chuck Fry) writes:
> In article <6nto3h$m78$1...@newsgate.sps.mot.com>,
> Mark Atanowicz <MAtan...@aol.com> wrote:
>
> >For the reason stated above, you need a thinner spoke, not a smaller
> >thread. Thinner spokes help normalize elongation WRT to the drive
> >side.
>
> Right, but it seems to me (remember, pure speculation!) that you need to
> increase the unit stress in the threads as well, thus I suspect a
> smaller diameter thread (say, 15/16/15) would better withstand the
> stress cycles in actual use than the 14/17/14 you suggest.
Not sure what you mean by "better withstand stress cycles". Smaller
threads would deform more, but the extra elongation is trivial compared
to that of the rest of the spoke. Also, this deformation only occurs
if there is tension on the spoke; as soon as the load is removed the
threads return to their unstressed condition. In short, the diamater
of the thread has virtually nothing to do with whether or not the
spokes loosen.
>
> >> OTOH, maybe I'll just build one with a trick asymmetrical rim instead...
> >> Keith Bontrager sure seems to know his stuff.
> >
> >Wheels with asymmetrical rims *and* thin spokes on the NDS would be
> >nice.
>
> Are both necessary?
You can't get there by using just an asymmetrical rim. You could get
real close by using a conventional rim and 14/15 and 14/17 spokes and
actually have slightly more NDS elongation by using these spokes and an
asym rim.
Mark Atanowicz
Rick Denney
>>
>>Wheels with asymmetrical rims *and* thin spokes on the NDS would be
>>nice.
>
>Are both necessary?
>
Neither are necessary to make an adequately strong and durable wheel.
Both contribute to making the strongest and most durable possible
wheel. So, necessity depends on your requirements.
Rick Denney
Take what you want and leave the rest.
Mark Atanowicz
ppwr...@aol.com (PPWrench) writes:
> Yikes-'ask a simple question....'
Whatsa matter Peter, can't throw the jargon with the big dogs? ;-)
Mark Atanowicz
Nick Payne
tension between the drive and non-drive side spokes will be the same no
matter what guage of spokes are used.
Erik Voldengen wrote:
>
> I've been told that using lighter guage spokes on the non drive
> side of the rear wheel will increase the tension on that side,
> reducing the difference in tension between the two sides.
>
PPWrench
Not an engineer-don' wanna be but I wonder how many of these other guys are
either....
"a little knowledge is a dangerous thing"
Peter
Rick Denney
wrote:
>This is nonsense. For a given amount of dish in a wheel, the relative
>tension between the drive and non-drive side spokes will be the same no
>matter what guage of spokes are used.
>
You are correct. But the discussion revolves around deflection, not
tension. The narrower spoke will deflect more at the same tension, and
will therefore maintain its tension over a large deflection of the rim
under load.
Mark McMaster
>
> MA=Mark Atanowicz <MAtan...@aol.com>
> EV=bbc...@user1.teleport.com (Erik Voldengen)
>
> EV> I've been told that using lighter guage spokes on the non drive
> EV> side of the rear wheel will increase the tension on that side,
> EV> reducing the difference in tension between the two sides.
>
> EV> BS? I've read "the book," and nothing, not even a mention of
> EV> folklore, was written.
>
> MA> The use of thinner spokes on the NDS doesn not increase the
> MA> tension on those spokes, but rather the elongation, which is a
> MA> function of tension *and* X-sectional area. This increases the
> MA> rim deformation at which the spokes go slack, making the wheel
> MA> more durable. While Jobst has never acknowledged my repeated
> MA> claims of this, his non-flaming signals tacit approval. ;-)
>
> Why not use the thinner spokes on both sides? Flipping your terms, I
> don't see how using thicker spokes on the drive side increases the elongation
> of spokes on the non-drive side.
I think he means that elongations become more uniform by using spoke
thicknesses more proportional to the static tensions on either side.
Yes, you could increase elongation of the drive side spokes by using
thinner spokes; but if the spokes are too thin you have more wind-up
problems (and possibly torsional failure) on the more highly tensioned
drive-side spokes. Since the drive-side spokes are under much higher
static tension, slackening of these spokes becomes less of an issue.
Mark McMaster
MMc...@ix.netcom.com
John Getsoian
>Why not use the thinner spokes on both sides? Flipping your terms, I
>don't see how using thicker spokes on the drive side increases the elongation
>of spokes on the non-drive side
The thicker spokes on the drive side unload with less deformation of
the rim. Thinner DS spokes would allow the rim to deform right past
the relaxation point of the lower tension NDS spikes (relatively
speaking) - right back to where you started from with 14s on both
sides. IAC, I think I like the asymmetrical rim solution better. Is
the Bontrager (or any other 'bare') asym rim available in 700C? I've
only seen a 26" version so far.
regards,
john getsoian
(jget...@csi.com)
Bruce Jackson
EV> I've been told that using lighter guage spokes on the non drive
EV> side of the rear wheel will increase the tension on that side,
EV> reducing the difference in tension between the two sides.
EV> BS? I've read "the book," and nothing, not even a mention of
EV> folklore, was written.
MA> The use of thinner spokes on the NDS doesn not increase the
MA> tension on those spokes, but rather the elongation, which is a
MA> function of tension *and* X-sectional area. This increases the
MA> rim deformation at which the spokes go slack, making the wheel
MA> more durable. While Jobst has never acknowledged my repeated
MA> claims of this, his non-flaming signals tacit approval. ;-)
Why not use the thinner spokes on both sides? Flipping your terms, I
don't see how using thicker spokes on the drive side increases the elongation
of spokes on the non-drive side.
--
Bruce Jackson - Sr. UNIX Systems Administrator - The M/A/R/C Group
b.a.j...@ieee.org
Mark Atanowicz
"John Getsoian" <jget...@csi.com> writes:
> On 10 Jul 1998 02:26:50 GMT, Bruce Jackson wrote:
>
> >Why not use the thinner spokes on both sides? Flipping your terms, I
> >don't see how using thicker spokes on the drive side increases the elongation
> >of spokes on the non-drive side
>
> The thicker spokes on the drive side unload with less deformation of
> the rim.
Not true, assuming the same spoke tension.
> Thinner DS spokes would allow the rim to deform right past
> the relaxation point of the lower tension NDS spikes (relatively
> speaking) - right back to where you started from with 14s on both
> sides.
The error in your logic is assuming that spoke elongation, not spoke
tension, determines when the spokes go slack for a given load. Wheels
laced with thick DS spokes at 100 lbs. of tension will go slack at
exactly the same loading as a wheel laced with thin DS spokes at 100
lbs. tension. NDS spokes are a different issue because the bracing
angle is different.
> Is the Bontrager (or any other 'bare') asym rim available in 700C? I've
> only seen a 26" version so far.
No, perhaps because the amount of offset available isn't significant.
Mark Atanowicz
Mark Atanowicz
jac...@silo.csci.unt.edu (Bruce Jackson) writes:
> MA=Mark Atanowicz <MAtan...@aol.com>
>
> MA> The use of thinner spokes on the NDS doesn not increase the
> MA> tension on those spokes, but rather the elongation, which is a
> MA> function of tension *and* X-sectional area. This increases the
> MA> rim deformation at which the spokes go slack, making the wheel
> MA> more durable. While Jobst has never acknowledged my repeated
> MA> claims of this, his non-flaming signals tacit approval. ;-)
>
> Why not use the thinner spokes on both sides?
Because the elongation on the DS will still be double.
> Flipping your terms, I
> don't see how using thicker spokes on the drive side increases the elongation
> of spokes on the non-drive side.
It's all relative. If spokes with 1/2 the X-section area are use on
the NDS, which have 1/2 the tension, the elongation is exactly the
same.
Mark Atanowicz
Mark McMaster
>
> In article <ysooiigciwgj...@news.compuserve.com>
> "John Getsoian" <jget...@csi.com> writes:
>
> > On 10 Jul 1998 02:26:50 GMT, Bruce Jackson wrote:
> >
> > >don't see how using thicker spokes on the drive side increases the elongation
> > >of spokes on the non-drive side
> >
> > the rim.
>
> Not true, assuming the same spoke tension.
I think what Bruce is getting at is that if the drive side spokes are
thicker, they will compress less under the same load (higher spring
constant). See below.
> > Thinner DS spokes would allow the rim to deform right past
> > the relaxation point of the lower tension NDS spikes (relatively
> > speaking) - right back to where you started from with 14s on both
> > sides.
>
> The error in your logic is assuming that spoke elongation, not spoke
> tension, determines when the spokes go slack for a given load. Wheels
> laced with thick DS spokes at 100 lbs. of tension will go slack at
> exactly the same loading as a wheel laced with thin DS spokes at 100
> lbs. tension. NDS spokes are a different issue because the bracing
> angle is different.
But in the case of a wheel, in which there is more than one spoke at a
time in the Load Effected Zone, the relative stiffnesses of the spoke
determines how much of the load is taken up by each spoke, and whether
the entire load is taken up by stiffer spokes before less tensioned
spokes become slack.
By way of example. Say the left spokes have a stiffness of K(l) and
static tension of T(l) and the right spokes have a stiffness of K(r) and
a static tension of T(r). The elongation of the left spokes will be
E(l) = T(l)/K(l) and likewise the elongation of the right spokes will be
E(r) = T(r)/K(r). If there were two spokes in the LAZ, one right and
one left, the combined stiffness of the two spokes would be K(tot.) =
K(l)+K(r). Under a load P, the deflection D of the rim would be D =
P/K(tot.) = P/{K(l)+K(r)}. The left spokes will slacken at the point at
which deflection D equals their static elongation E(l), so D = E(l) =
P/{K(l)+K(r)}. The laod required to slacken the left spokes is thus P =
E(l){K(l)+K(r)}. Substituting T(l)/K(l) in for E(l) we can see how much
wheel load is required to slacken the left spokes as a function of their
static preload and the spoke stiffnesses:
P = {T(l)/K(l)}{K(l)+K(r)} = T(l)*{1+K(r)/K(l)}
From this we see that the load required to slacken the left spokes can
be increased by either decreasing the stiffness (thickness) of the lefts
spokes, or increasing the stiffness (thickness) of the right spokes. If
the spokes were the same thickness, the left spokes would support half
the load, but if the left spokes are thinner, they will support less
than half the load, and so would require a higher total wheel load to
slacken.
Of course, it isn't quite this simple, because the number of spokes and
the relative stiffness of the spokes vs. the rim will determine how many
spokes are in the LAZ. Also since the thicker spokes take more of the
load, that means that the spoke holes in the rim are not evenly loaded,
with the right spoke holes taking higher loads than the left. But, even
with the uneven spoke and rim loading, I tend to think that using
slightly thinner spokes on the left builds a more durable wheel overall.
> > Is the Bontrager (or any other 'bare') asym rim available in 700C? I've
> > only seen a 26" version so far.
>
> No, perhaps because the amount of offset available isn't significant.
Ritchey is presently selling their OCR road rim.
Mark McMaster
MMc...@ix.netcom.com
Mark Hickey
require a Cray to solve) is to assume....
1) Every spoke has a range of "ideal tension"
2) Thinner spokes have a lower range of "ideal tension"
3) Given the difference in tension between DS and NDS spokes, it's
possible to get all the spokes closer to their "ideal tension" by
using lighter gauge spokes on the NDS (a method I've used for a long
time).
Mark Hickey
Habanero Cycles
http://www.cynetfl.com/habanero/
Home of the $695 ti frame
Bruce Jackson
Mark Hickey <mhi...@cynetfl.com> wrote:
> Perhaps the easiest way to approach this issue (and one that doesn't
> require a Cray to solve) is to assume....
> 1) Every spoke has a range of "ideal tension"
> 2) Thinner spokes have a lower range of "ideal tension"
> 3) Given the difference in tension between DS and NDS spokes, it's
> possible to get all the spokes closer to their "ideal tension" by
> using lighter gauge spokes on the NDS (a method I've used for a long
> time).
The third statement does not logically follow the first two.
Let's assume that as you claim that there is an ideal tension where
spokes work best. Let's further assume that thin spokes have a lower
ideal tension than thick spokes.
Without having a way to calculate the ideal spoke tension we can not tell
if the ideal tension of thin spokes lies in the tension range of non-drive
side spokes or the ideal tension of thicker spokes lies in the range of drive
side spokes. If the ideal tension of all spokes is above the tension range
of both sides of the wheel than we would be better off using the thinnest
practical spokes on both sides. If the ideal tension of all spokes lies below
the tension range of both sides than we would be better off using the thickest
practical spokes on both sides.
Andy Dingley
wrote:
>Without having a way to calculate the ideal spoke tension
We don't actually need this for this example, just the ratio of IST
for thick & thin spokes. As you say though, we don't know that either.
So, does anyone know what it is ? Is the ideal tension proportional
to the spoke cross section (assuming spokes have homogeneous
metallurgy), or the spoke diamter (assuming that all load is
concentrated in a thin surface shell), or is it something else
entirely ?
I do like the simplicity of Mark's explanation though.
--
Smert' Spamionem
Chuck Fry
Andy Dingley <din...@codesmiths.com> wrote:
>So, does anyone know what it is ? Is the ideal tension proportional
>to the spoke cross section (assuming spokes have homogeneous
>metallurgy), or the spoke diamter (assuming that all load is
>concentrated in a thin surface shell), or is it something else
>entirely ?
What little literature I have read on threaded fasteners in tension
applications (of which spokes are an example) says that the ideal
tension is 70-85% of the yield strength of the material. This in turn
is determined by the cross section area.
You could take a look at the graphs in Jobst's book for an idea what
that value is for some particular spokes.
>I do like the simplicity of Mark's explanation though.
Indeed. Mark is one of the few posters whose posts I always read.
Mark Hickey
>I do like the simplicity of Mark's explanation though.
Simple is as simple does.... ;-)
Mark "life is a bowl of alloy nipples" Hickey
Bikefixr
will lessen the tension differential between the 2 side of a rear wheel. He
claims that the miimal dish of a rear wheel has no severe tension differential.
He goes on to say that: ...the relative tension between the drive and
non-drive side spokes will be the same no matter what gauge of spokes are
used."
Well, Nick, I have to tell 'ya! You are sooooo full of crap I can smell it
from here! It's pretty obvious that you aren't a skilled wheelbuilder. If I was
one of your clients-I'd be scared! After the thousands of wheels i have built,
I'm here to tell you and Mr. Voldengen, the original inquiree, that it
absolutely DOES make a difference-a huge one. With a given drive side tension,
the difference between a 14/15 db non drive and a 15/16 is approx 35% according
to my Hozan spoke tensiometer. With some of the new offset ferruled rims and a
really thin 15/17 spoke I can build a wheel whereby the non-drive is under a
HIGHER tension than the drive side. Simple physics-a certain amount of work is
necessary to hold the rim in center of the axle lock nuts based upon the work
being done by the drive spokes. This given amount of work to hold the rim is a
constant. A thinner spoke has less material-hence it must work harder. So the
tension per unit of material is higher. Sorry-but I would suggest not giving
advice about things you know nothing about.
Mark Atanowicz
bike...@aol.com (Bikefixr) writes:
> Nick Payne writes that it is "nonsense" that using lighter non-drive spokes
> will lessen the tension differential between the 2 side of a rear wheel. He
> claims that the miimal dish of a rear wheel has no severe tension differential.
> He goes on to say that: ...the relative tension between the drive and
> non-drive side spokes will be the same no matter what gauge of spokes are
> used."
> Well, Nick, I have to tell 'ya! You are sooooo full of crap I can smell it
> from here!
Oh, the irony...
> It's pretty obvious that you aren't a skilled wheelbuilder. If I was
> one of your clients-I'd be scared! After the thousands of wheels i have built,
> I'm here to tell you and Mr. Voldengen, the original inquiree, that it
> absolutely DOES make a difference-a huge one. With a given drive side tension,
> the difference between a 14/15 db non drive and a 15/16 is approx 35% according
> to my Hozan spoke tensiometer.
If you knew how most tensiometers worked, you'd know why. They measure
tension based on deflection and most do not "zero out" when they
contact the spoke; a thinner spoke will indicate a higher tension, even
if they are exactly the same. Try substituting a 15/16 spoke on a
14/15 wheel and measuring the tension. Gee, they're different!
> With some of the new offset ferruled rims and a
> really thin 15/17 spoke I can build a wheel whereby the non-drive is under a
> HIGHER tension than the drive side. Simple physics-a certain amount of work is
> necessary to hold the rim in center of the axle lock nuts based upon the work
> being done by the drive spokes.
Simple physics shows that if the rim is in equilibrium, the NDS spokes
*have* to be at a lower tension because of their higher bracing angle.
An FYI, no "work" is being done.
> This given amount of work to hold the rim is a
> constant. A thinner spoke has less material-hence it must work harder. So the
> tension per unit of material is higher. Sorry-but I would suggest not giving
> advice about things you know nothing about.
Good advice. I suggest taking it.
Mark Atanowicz
Brian Nystrom
Bikefixr wrote:
> Nick Payne writes that it is "nonsense" that using lighter non-drive spokes
> will lessen the tension differential between the 2 side of a rear wheel.
He's right. The amount of tension required to hold the rim in place is the same
regardless of the size of the spokes. However, lighter gauge spokes are more
elastic and will absorb the differences in tension that occur as the wheel rotates.
This reduces fatigue failures when lighter spokes are used on the NDS.
> He
> claims that the miimal dish of a rear wheel has no severe tension differential.
Dead wrong, unless he's building 5-speed wheels.
> He goes on to say that: ...the relative tension between the drive and
> non-drive side spokes will be the same no matter what gauge of spokes are
> used."
Absolutely correct.
> Well, Nick, I have to tell 'ya! You are sooooo full of crap I can smell it
> from here! It's pretty obvious that you aren't a skilled wheelbuilder. If I was
> one of your clients-I'd be scared! After the thousands of wheels i have built,
> I'm here to tell you and Mr. Voldengen, the original inquiree, that it
> absolutely DOES make a difference-a huge one. With a given drive side tension,
> the difference between a 14/15 db non drive and a 15/16 is approx 35% according
> to my Hozan spoke tensiometer. With some of the new offset ferruled rims and a
> really thin 15/17 spoke I can build a wheel whereby the non-drive is under a
> HIGHER tension than the drive side.
This is absolutely impossible, unless you dramatically reduce the spoke count on
the NDS. If you had 18 spokes on the drive side and 6 on the NDS, this might
possibly be the case.
> Simple physics-a certain amount of work is
> necessary to hold the rim in center of the axle lock nuts based upon the work
> being done by the drive spokes. This given amount of work to hold the rim is a
> constant. A thinner spoke has less material-hence it must work harder. So the
> tension per unit of material is higher. Sorry-but I would suggest not giving
> advice about things you know nothing about.
Perhaps your physics are for another universe? They sure don't work here!
--
TO REPLY: Remove the *** from my email address. Sorry for the inconvenience, but I
have to fight spammers somehow!
Al Williams
> Well, Nick, I have to tell 'ya! You are sooooo full of crap I can smell it
> from here! It's pretty obvious that you aren't a skilled wheelbuilder.
If I was
> one of your clients-I'd be scared! After the thousands of wheels i have
built,
> I'm here to tell you and Mr. Voldengen, the original inquiree, that it
> absolutely DOES make a difference-a huge one. With a given drive side tension,
> the difference between a 14/15 db non drive and a 15/16 is approx 35%
according
> to my Hozan spoke tensiometer.
You received a calibration chart in the box with your Hozan. It has
different curves for different size spokes. Are you using the right curve
for each size spoke? A given reading on the Hozan corresponds to a higher
tension on a thin spoke than on a thick spoke. If you refer to the same
curve for readings with different sized spokes, you will think the thin
spokes have higher tension when they don't.
With some of the new offset ferruled rims and a
> really thin 15/17 spoke I can build a wheel whereby the non-drive is under a
> HIGHER tension than the drive side.
My meter is about 15 years old. There's no curve for spokes lighter than
15/16 gauge on my calibration chart. Does yours have a curve for 15/17?
> Simple physics-a certain amount of work is
> necessary to hold the rim in center of the axle lock nuts based upon the work
> being done by the drive spokes. This given amount of work to hold the rim is a
> constant. A thinner spoke has less material-hence it must work harder. So the
> tension per unit of material is higher.
Oh, boy. Work is defined as force multiplied by the distance over which
it is applied. The spoke has tension (force). Where's the distance, in a
spoked wheel?
>Sorry-but I would suggest not giving advice about things you know nothing
about.
That may be the only true statement in your post! Maybe you should think
about it some more :)
Two things that you might consider doing:
1: Try building a standard dimension rear wheel that has half of the left
side spokes one size and half another. Get all the right-side tensions
the same with your meter. Tell us what the tensions are on the left-side
spokes, when the wheel is true. (I am confident that you will find the
tensions to be the same, if you refer to the correct calibration curve for
each measurement.)
2: Check out the "Resolution of Forces" chapter in any high school
physics text (or other applicable reference). Better yet, check several
references. Then explain to us how left side spoke tension could vary
with size of spokes, in the context of the information in the books.
Al Williams
San Jose CA
ILitig8
thread....
ive always used lighter spokes on the non-drive side on my MTB wheels heres why
(and it doesnt require a degree)
1. less rotational mass = faster acceleration with less energy used
2. lighter spokes= less rotational mass
3. faster acceleration=speed
4. speed is good
5.lighter spokes = speed
6. drive side spokes = close proximity to chain
7. chain has desire to eat spokes
8. thicker spokes = better ability to resist chains desire to eat them
so light spokes for non-drive=speed
and heavy spokes for drive side = the spokes stay in one piece
all in all light where i can heavy where i need
Ride hard and fall Harder!!!!!
There is a fine line between self-assurance and arrogance, one that becomes
more obscure when juxtaposed against seemingly inhuman feats.
Vandy
Stergios Papadakis
Bikefixr wrote:
> Nick Payne writes that it is "nonsense" that using lighter non-drive spokes
> will lessen the tension differential between the 2 side of a rear wheel. He
> claims that the miimal dish of a rear wheel has no severe tension differential.
> He goes on to say that: ...the relative tension between the drive and
> non-drive side spokes will be the same no matter what gauge of spokes are
> used."
> Well, Nick, I have to tell 'ya! You are sooooo full of crap I can smell it
> from here!
It pays to use a softer tone in your posts. Then you don't look so bad when you
screw up.
> It's pretty obvious that you aren't a skilled wheelbuilder. If I was
> one of your clients-I'd be scared! After the thousands of wheels i have built,
> I'm here to tell you and Mr. Voldengen, the original inquiree, that it
> absolutely DOES make a difference-a huge one. With a given drive side tension,
> the difference between a 14/15 db non drive and a 15/16 is approx 35% according
> to my Hozan spoke tensiometer. With some of the new offset ferruled rims and a
> really thin 15/17 spoke I can build a wheel whereby the non-drive is under a
> HIGHER tension than the drive side. Simple physics-a certain amount of work is
> necessary to hold the rim in center of the axle lock nuts based upon the work
> being done by the drive spokes. This given amount of work to hold the rim is a
> constant. A thinner spoke has less material-hence it must work harder. So the
> tension per unit of material is higher. Sorry-but I would suggest not giving
> advice about things you know nothing about.
The drive spokes must be under higher tension. Nick is right. You were probably
headed in the right direction when you said "the tension per unit of material is
higher." With thinner NDS spokes, you may be able to build a wheel where the
stress on the NDS spokes is larger than the stress on the drive side, but the
relative tensions depend only on the angles of the spokes.
Stress, by the way, is tension/(cross sectional area).
Stergios
DEMPSEY
The force exerted by any thickness of NDS spoke is the same, however the
tension in the thinned section of spoke (force per unit cross sectional area)
is higher.
Use different thicknesses of spokes in wheels of differring bracing angles
if you want to achieve similar spoke tensions in the middle of the spokes
on each side.
-Brian
--
DEMPSEY,BRIAN DAVID
Georgia Institute of Technology, Atlanta Georgia, 30332
uucp: ...!{decvax,hplabs,ncar,purdue,rutgers}!gatech!prism!ch89abe
Internet: ch8...@prism.gatech.edu
Mark McMaster
>
> Get it right, guys.
> The force exerted by any thickness of NDS spoke is the same, however the
> tension in the thinned section of spoke (force per unit cross sectional area)
> is higher.
> Use different thicknesses of spokes in wheels of differring bracing angles
> if you want to achieve similar spoke tensions in the middle of the spokes
> on each side.
Close, but not quite there yet. The term "tension" is usually used as a
measure of force. A smaller cross section under the same force (or
tension) will have a higher "stress" (force per unit area) - I believe
this is the word you were looking for. To get even closer, I think what
you are really trying to achieve by using different thickness spokes on
either side is similar elongations (length increase under load), to
prevent the spokes from slackening under load; elongation is
proportional to stress, so similar stresses will produce similar
elongations.
Mark McMaster
MMc...@ix.netcom.com
David Casseres
(DEMPSEY) wrote:
>Get it right, guys.
>The force exerted by any thickness of NDS spoke is the same, however the
>tension in the thinned section of spoke (force per unit cross sectional area)
>is higher.
>Use different thicknesses of spokes in wheels of differring bracing angles
>if you want to achieve similar spoke tensions in the middle of the spokes
>on each side.
Last time I heard about it (in college physics) tension was the pulling
force, which is the same along the entire length of the spoke. Stress is
tension divided by cross-sectional area.
There is no way to have the same tension in the spokes on both sides of an
unsymmetrical wheel. You can have the same stress, with different spoke
gauges, but I don't know why you'd want to do that.
--
David Casseres
Exclaimer: Hey!
Chuck Fry
David Casseres <cass...@apple.com> wrote:
>Last time I heard about it (in college physics) tension was the pulling
>force, which is the same along the entire length of the spoke. Stress is
>tension divided by cross-sectional area.
>
>There is no way to have the same tension in the spokes on both sides of an
>unsymmetrical wheel. You can have the same stress, with different spoke
>gauges, but I don't know why you'd want to do that.
Why? So that both sides respond to changing loads in the same
proportions. More evenly distributed stresses are less likely
to put the wheel out of true, IMHO.
Of course I haven't tried this myself... yet. But as often as my
current MTB rear wheel requires truing, I'm seriously considering a
rebuild with an asymmetrical rim and lighter gauge spokes on the
non-drive side (since even asym rims don't eliminate the imbalance).
David Casseres
>In article <casseres-300...@cassda.apple.com>,
>David Casseres <cass...@apple.com> wrote:
>>Last time I heard about it (in college physics) tension was the pulling
>>force, which is the same along the entire length of the spoke. Stress is
>>tension divided by cross-sectional area.
>>
>>There is no way to have the same tension in the spokes on both sides of an
>>unsymmetrical wheel. You can have the same stress, with different spoke
>>gauges, but I don't know why you'd want to do that.
>
>Why? So that both sides respond to changing loads in the same
>proportions. More evenly distributed stresses are less likely
>to put the wheel out of true, IMHO.
But wheels aren't pulled out of true by unequal loads. They go out of
true because spokes loosen, and spokes loosen because they aren't tight
enough in the first place. I recommend a thorough reading of "The Bicycle
Wheel" by Jobst Brandt.
John Getsoian
>But wheels aren't pulled out of true by unequal loads. They go out of
>true because spokes loosen, and spokes loosen because they aren't tight
>enough in the first place.
David;
But a fat NDS spoke at a low stress will be at lower elongation and
will slacken and thus be a candidate for nipple rotation sooner than
its high stress/high elongation mates on the drive side that are
actually controlling the deformation of the rim.
IOW, the elongation change on the DS controls the deformation of the
rim (size of load affected zone), since most of the net tension in
the wheel in on the DS and the system reaches equilibrium when the
total tension reduction of the spokes in the load-affected zone is
equal to the applied load at the axle. If the NDS/DS tension ratio is
low enough, the low stress NDS spokes will go to a state of zero load
at a point when the DS spokes have absorbed the load but are still
under stress and still in their elastic range.
If you match the diameter ratios to the tension ratios, all spokes
will be at equal elongation and both sides will support the applied
load at that same ratio thoughout their elastic ranges - which will
now be of equal displacement. This is probably the better argument
for this technique, since the wheel stays in lateral force balance as
it is vertically loaded, which cannot be not true of a laterally
asymmetrical
wheel (i.e. dished) built with identical number and gauge spokes on
both sides. It's probably only because the bracing angle on the DS of
an 8/9sp hub is already so near zero that there isn't a noticeable
squim of the wheel to the right in the load affected zone.
regards,
john getsoian
(jget...@csi.com)
Snoopy
wrote:
>
>IOW, the elongation change on the DS controls the deformation of the
>rim (size of load affected zone), since most of the net tension in
>the wheel in on the DS and the system reaches equilibrium when the
>total tension reduction of the spokes in the load-affected zone is
>
How extensive is this 'load affected zone'? SNOOPY
---------------------------------------------------
Message posted by SNOOPY
using Forte Free Agent Version 1.11/32
Rick Denney
On Fri, 31 Jul 1998 11:32:57 GMT, tenn...@caverock.net.nz (Snoopy)
wrote:
Mark Atanowicz
"John Getsoian" <jget...@csi.com> writes:
> On Thu, 30 Jul 1998 14:59:55 -0700, David Casseres wrote:
>
> >But wheels aren't pulled out of true by unequal loads. They go out of
> >true because spokes loosen, and spokes loosen because they aren't tight
> >enough in the first place.
>
> David;
>
> But a fat NDS spoke at a low stress will be at lower elongation and
> will slacken and thus be a candidate for nipple rotation sooner than
> its high stress/high elongation mates on the drive side that are
> actually controlling the deformation of the rim.
>
> IOW, the elongation change on the DS controls the deformation of the
> rim (size of load affected zone), since most of the net tension in
> the wheel in on the DS and the system reaches equilibrium when the
> total tension reduction of the spokes in the load-affected zone is
> equal to the applied load at the axle. If the NDS/DS tension ratio is
> low enough, the low stress NDS spokes will go to a state of zero load
> at a point when the DS spokes have absorbed the load but are still
> under stress and still in their elastic range.
Not true, John. As the NDS spokes near zero tension, the DS spokes
will pull the rim to maintain static equilibrium as you point out
below. I'm in agreement, however, that attempting to equalize spoke
elongation is good.
>
> If you match the diameter ratios to the tension ratios, all spokes
> will be at equal elongation and both sides will support the applied
> load at that same ratio thoughout their elastic ranges - which will
> now be of equal displacement. This is probably the better argument
> for this technique, since the wheel stays in lateral force balance as
> it is vertically loaded, which cannot be not true of a laterally
> asymmetrical
> wheel (i.e. dished) built with identical number and gauge spokes on
> both sides. It's probably only because the bracing angle on the DS of
> an 8/9sp hub is already so near zero that there isn't a noticeable
> squim of the wheel to the right in the load affected zone.
>
>
>
> regards,
> john getsoian
> (jget...@csi.com)
>
>
>
Mark Atanowicz
Andy Dingley
<papa...@princeton.edu> wrote:
> The drive spokes must be under higher tension.
Does this always have to be true ? Strictly it's the axial component
of the combined tensions which must be equal.
Going from that statement to the DS spokes always having the higher
tension requires we make the assumptions of the spoke numbers being
equal, and that the lacing patterns are the same.
The equal numbers assumption is reasonable for a bike wheel, although
it's not always true for a car wheel (some wide rims have spokes
arranged in 3 groups).
The lacing pattern assumption is questionable, although generally
true. If the NDS lacing is radial and the DS crossed, then this will
keep the axial components equal but the different angles between the
spokes mean that the spoke tension is no longer such a a simple
matter. I haven't calculated it personally, and my trig is too rusty
to do it in my head right now (manyana), but is it possible that a
suitable geometry might allow asymmetric lacing to equalise the spoke
tensions ?
--
Smert' Spamionem
Mark McMaster
>
> On Sat, 25 Jul 1998 20:11:56 -0400, Stergios Papadakis
> <papa...@princeton.edu> wrote:
>
> > The drive spokes must be under higher tension.
>
>
> true. If the NDS lacing is radial and the DS crossed, then this will
> keep the axial components equal but the different angles between the
> spokes mean that the spoke tension is no longer such a a simple
> matter. I haven't calculated it personally, and my trig is too rusty
> to do it in my head right now (manyana), but is it possible that a
> suitable geometry might allow asymmetric lacing to equalise the spoke
> tensions ?
On a modern 8 or 9 speed wheel, the non-driveside flange is offset from
the center of the hub almost twice as much as the drive side flange
(typically 36mm vs. 19mm). A small change in where the spokes are laced
from the non-driveside flange will do little change to the spoke angles
and relative right/left tension. The only way to get the tensions even
close to equal are either to offset the spoke holes on the rim (but road
rims are so narrow they can only be offset a few millimeters at best),
or move the non-driveside flange inward (making the wheel laterally weak
in both directions, instead of just one).
In regard to radial lacing: If you radially lace the spokes heads out
(pulling from the inner face of the flange), it may seem that you will
have a smaller bracing angle than when crossing the spokes (crossed
spokes effectively pull from the center of the flange). But when you
look closer, you see that although the effective flange offset has been
decreased by about half the flange width (about 1.5mm, or about 5%), a
radial laced spoke is also shorter than a crossed laced spoke (by about
12mm or about 5%). You don't have to remember much trig to know that if
you shorten two sides of a right triangle by the same proportion, the
angles remain the same; there is no change in the bracing angle of the
spokes for radial vs. crossed lacing.
Mark McMaster
MMc...@ix.netcom.com
Mark Atanowicz
din...@codesmiths.com (Andy Dingley) writes:
> On Sat, 25 Jul 1998 20:11:56 -0400, Stergios Papadakis
> <papa...@princeton.edu> wrote:
>
> > The drive spokes must be under higher tension.
>
> Does this always have to be true ? Strictly it's the axial component
> of the combined tensions which must be equal.
>
> Going from that statement to the DS spokes always having the higher
> tension requires we make the assumptions of the spoke numbers being
> equal, and that the lacing patterns are the same.
>
> The equal numbers assumption is reasonable for a bike wheel, although
> it's not always true for a car wheel (some wide rims have spokes
> arranged in 3 groups).
>
> The lacing pattern assumption is questionable, although generally
> true. If the NDS lacing is radial and the DS crossed, then this will
> keep the axial components equal but the different angles between the
> spokes mean that the spoke tension is no longer such a a simple
> matter. I haven't calculated it personally, and my trig is too rusty
> to do it in my head right now (manyana), but is it possible that a
> suitable geometry might allow asymmetric lacing to equalise the spoke
> tensions ?
A 3X pattern requires only about 5% more tension. This is
insignificant compared to the 100% more required by dishing.
Mark Atanowicz
John Getsoian
>As the NDS spokes near zero tension, the DS spokes
>will pull the rim to maintain static equilibrium as you point out
>below.
Yes, true enough if all the dynamics are considered. OTOH, this is
not a very desireable result either!
regards,
john getsoian
(jget...@csi.com)
Snoopy
Denney) wrote:
>
>>How extensive is this 'load affected zone'? SNOOPY
>>
>
Intuitively I would have thought that all spokes in the 'bottom half'
of the wheel would support the load and so therefore have the tension
affected by the load.
Tension in an instantaneously vertical spoke would certainly be
entirely dependent on the load, whereas the spoke one removed from
this (i.e. ten degrees away) would be affected by an amount
(load x cos10degrees)
The next spoke up the rim would be affected by an amount
(load x cos20)
and so on up the rim until you came to the instantaneously horizontal
spoke which would be affected by
(load x cos90) , which is zero, since cos90 is zero.
At any given instant it is the spokes near the bottom of the rim which
ar subject to the most compression (load x cos'theta') imposed over
the existing spoke tension, so maybe in practical terms your answer of
five spokes is right? SNOOPY
Snoopy
Dingley) wrote:
>
>> The drive spokes must be under higher tension.
>
>Does this always have to be true ? Strictly it's the axial component
>of the combined tensions which must be equal.
>
>Going from that statement to the DS spokes always having the higher
>tension requires we make the assumptions of the spoke numbers being
>equal, and that the lacing patterns are the same.
>
>The equal numbers assumption is reasonable for a bike wheel, although
>it's not always true for a car wheel (some wide rims have spokes
>arranged in 3 groups).
>
a bike wheel either. Check out this site.
http://www.tenthouse.com/daveswheels
I have been running just such a wheel for over a year now (4000km),
although it wasn't built by Dave. SNOOPY
Mark McMaster
All this would be correct _if_ the rim where infinitely rigid; but it is
not. Radially, the rim is much less rigid than the spokes. At the
bottom of the wheel, the rim flexes inward in a small region near the
ground contact point (the LAZ or Load Affected Zone). Because spokes
are so rigid radially, the spokes in this small region only need to
compress (detension) a small amount to take up the entire wheel load, so
the rim transmits virtually no load to spokes outside this region.
Mark McMaster
MMc...@ix.netcom.com
Rick Denney
stiff, then what you say would be true. But it ain't. It's actually
less stiff than the spokes.
On Sat, 01 Aug 1998 10:26:02 GMT, tenn...@caverock.net.nz (Snoopy)
wrote:
>On Fri, 31 Jul 1998 13:20:43 GMT, rde...@mail.viggen.com (Rick
>Denney) wrote:
>>
>>>How extensive is this 'load affected zone'? SNOOPY
>>>
>>About the bottom five spokes on a 36-spoke wheel.
>>
>For a wheel 360degrees round, 5 spokes enclose a 40degree arc.
>Intuitively I would have thought that all spokes in the 'bottom half'
>of the wheel would support the load and so therefore have the tension
>affected by the load.
>
>Tension in an instantaneously vertical spoke would certainly be
>entirely dependent on the load, whereas the spoke one removed from
>this (i.e. ten degrees away) would be affected by an amount
> (load x cos10degrees)
>The next spoke up the rim would be affected by an amount
> (load x cos20)
>and so on up the rim until you came to the instantaneously horizontal
>spoke which would be affected by
> (load x cos90) , which is zero, since cos90 is zero.
>
>At any given instant it is the spokes near the bottom of the rim which
>ar subject to the most compression (load x cos'theta') imposed over
>the existing spoke tension, so maybe in practical terms your answer of
>five spokes is right? SNOOPY
>
>---------------------------------------------------
>Message posted by SNOOPY
>using Forte Free Agent Version 1.11/32
Snoopy
Denney) wrote:
>>>
>>>>How extensive is this 'load affected zone'? SNOOPY
>>>>
>>>About the bottom five spokes on a 36-spoke wheel.
>>>
>>For a wheel 360degrees round, 5 spokes enclose a 40degree arc.
>>Intuitively I would have thought that all spokes in the 'bottom half'
>>of the wheel would support the load and so therefore have the tension
>>affected by the load.
>>
>>Tension in an instantaneously vertical spoke would certainly be
>>entirely dependent on the load, whereas the spoke one removed from
>>this (i.e. ten degrees away) would be affected by an amount
>
>Except that the rim deflects in compression. If the rim was infinitely
>stiff, then what you say would be true. But it ain't. It's actually
>less stiff than the spokes.
>
post as I wanted to see if anyone else would come up with the same
idea independently.
However if the rim is to deflect across a 40degree arc, does not that
also imply the spokes enclosed in that 40degree arc deflect also,
which in turn implies the spokes instantaneously go slack, through
that arc? Or would it be more correct to say that if the spokes are
sufficiently pre-tensioned the only deflection is in the rim between
spokes? SNOOPY
Stergios Papadakis
Snoopy wrote:
In simple terms, the spokes are stretched when the wheel is built. They
are stretched enough that they don't go completely slack even when the part
of the wheel in that 40 degree arc deflects vertically a little bit. The
tension on the spokes is reduced some, though. You can hear it if you
pluck the spokes when someone is sitting on the bike.
Stergios
Snoopy
<papa...@princeton.edu> wrote:
>
>> However if the rim is to deflect across a 40degree arc, does not that
>> also imply the spokes enclosed in that 40degree arc deflect also,
>> which in turn implies the spokes instantaneously go slack, through
>> that arc? Or would it be more correct to say that if the spokes are
>> sufficiently pre-tensioned the only deflection is in the rim between
>> spokes? SNOOPY
>>
>In simple terms, the spokes are stretched when the wheel is built. They
>are stretched enough that they don't go completely slack even when the part
>of the wheel in that 40 degree arc deflects vertically a little bit. The
>tension on the spokes is reduced some, though. You can hear it if you
>pluck the spokes when someone is sitting on the bike.
>
instantaneously reduce tension on the spokes within the ground to
wheel 'contact zone.'
If the spokes were not tensioned enough this could indeed cause the
spokes to go 'completely slack'.
However any reputably built wheel will be sufficiently pre-tensioned
so that spoke tension in the 'contact zone' will not drop to zero when
the bike is being ridden laden. But if there is an (albeit reduced)
tension in the spokes this is only possible if the spokes are still in
contact with the rim.
I.e. there is no deflection of the rim relative to the spokes. This
means that the rim can only deflect between spokes. Therefore the
contact zone is not 40 degrees as assumed above, but the angle between
single spokes which for a 36 spoke rim is 10 degrees.
Have I got it right, or is there some obvious flaw in my argument?
Stergios Papadakis
Snoopy wrote:
> On Mon, 03 Aug 1998 18:03:37 -0400, Stergios Papadakis
> <papa...@princeton.edu> wrote:
> >In simple terms, the spokes are stretched when the wheel is built. They
> >are stretched enough that they don't go completely slack even when the part
> >of the wheel in that 40 degree arc deflects vertically a little bit. The
> >tension on the spokes is reduced some, though. You can hear it if you
> >pluck the spokes when someone is sitting on the bike.
> >
> So I think we are agreed that any weight on a bike will
> instantaneously reduce tension on the spokes within the ground to
> wheel 'contact zone.'
> If the spokes were not tensioned enough this could indeed cause the
> spokes to go 'completely slack'.
Yes, this we agree on.
> However any reputably built wheel will be sufficiently pre-tensioned
> so that spoke tension in the 'contact zone' will not drop to zero when
> the bike is being ridden laden. But if there is an (albeit reduced)
> tension in the spokes this is only possible if the spokes are still in
> contact with the rim.
>
> I.e. there is no deflection of the rim relative to the spokes. This
> means that the rim can only deflect between spokes. Therefore the
> contact zone is not 40 degrees as assumed above, but the angle between
> single spokes which for a 36 spoke rim is 10 degrees.
>
> Have I got it right, or is there some obvious flaw in my argument?
> SNOOPY
In the same way that rubber bands stretch, spokes stretch (only spokes stretch a
lot less). The spokes in a wheel after it is built are actually longer than
they were when they were sitting on the table. So, strictly speaking, there is
no deflection of the rim relative to the outer ends of the spokes, but there is
deflection of the rim relative to the inner (hub) ends of the spokes. So, the
rim deflects over some arc, roughly 40 degrees. The tension in the spokes in
that region is reduced, and the spokes get shorter to soak up the deflection.
The nipples in that arc remain in contact with the rim, because the spokes were
stretched enough when the wheel was built that they are STILL stretched
(although less so) when the rim deflects.
Stergios
Snoopy
<papa...@princeton.edu> wrote:
>
>In the same way that rubber bands stretch, spokes stretch (only spokes stretch a
>lot less). The spokes in a wheel after it is built are actually longer than
>they were when they were sitting on the table. So, strictly speaking, there is
>no deflection of the rim relative to the outer ends of the spokes, but there is
>deflection of the rim relative to the inner (hub) ends of the spokes.
>
stand, the rim will be permanently compressed, by an amount equal to
twice the radial component of the tension strain on each spoke.
For example on a radially spoked wheel the amount of elongation in
each spoke represents half the decrease in the diameter of the rim,
but in a spoke that is not radial it is only the component of the
spoke tension in the diametral direction that affects the overall
diameter of the rim.
>
>So, the
>rim deflects over some arc, roughly 40 degrees. The tension in the spokes in
>that region is reduced, and the spokes get shorter to soak up the deflection.
>The nipples in that arc remain in contact with the rim, because the spokes were
>stretched enough when the wheel was built that they are STILL stretched
>(although less so) when the rim deflects.
>
Returning to the case of a ,say, 36 spoke drive wheel being ridden
(i.e. the subject of this thread), and to keep the maths simple:
(i) the case where the tension in the 18 drive side spokes is exactly
twice that of the 18 non drive side
(ii) There are equal numbers of spokes, of equal gauge, equally
crossed on each side.
If the rim deflects radially while being ridden over a 40 degree arc,
then the instantaneous change in tension of the spokes in the contact
affected zone will be unbalanced.
I.e. because there are 5 spokes in the 40 degree arc we will
instantaneously reduce the tension in either:
a/ three drive side spokes and two offside spokes OR
b/ three off side spokes and two drive side spokes.
In both these cases there should be a measurable effect caused by the
sideways component of the instantaneous release in spoke tension.
**************************************************************
Please bear with me while I define the problem in mathematical terms
If you are not into maths, then skip this next bit.
**************************************************************
If we look at the bike being ridden head on towards us (bike & rider
in a vertical plane), and consider the rear wheel, then the offside
side spokes will appear at a small angle to the vertical (lets call
this angle 'omega'), and the drive side spokes will be at an even
smaller angle (let's call this 'delta').
In this example delta is exactly half of omega.
Where T is the tension in the offside spokes and hence 2T is by
implication the tension in the drive side spokes.
The instantaneous sideways out of balance force will be in case (a),
as above:
3[2Tsin(delta)] - 2[Tsin(omega) ]
but since for static balance of sideways forces
Tsin(omega)=2Tsin(delta), then this out of balance force reduces to
3[Tsin(omega)] - 2[Tsin(omega)] = Tsin(omega)
Similarly we can calculate the instantaneous out of balance force in
case (b) above as:
2[2Tsin(delta)] - 3[Tsin(omega)] = -Tsin(omega), which is equal and
opposite to case (a), which is what you might intuitively expect.
**********************************************************************
End of mathematical bit
**********************************************************************
I am saying here that there is a constantly rotating out of balance
force scenario which is caused by there being an uneven number of
spokes in the assumed 40 degree contact zone. If there were only
four spokes in the contact zone it would not happen. Four spokes
in a 40 degree arc implies a spacing of 13.333 degrees per spoke or 27
spokes per wheel for 'dynamic' wheel balance.
Assuming that anyone has read this far do you agree with me, or is my
reasoning fatally flawed? SNOOPY
Mark Atanowicz
tenn...@caverock.net.nz (Snoopy) writes:
> On Tue, 04 Aug 1998 12:38:32 -0400, Stergios Papadakis
> <papa...@princeton.edu> wrote:
> >
> >In the same way that rubber bands stretch, spokes stretch (only spokes stretch a
> >lot less). The spokes in a wheel after it is built are actually longer than
> >they were when they were sitting on the table. So, strictly speaking, there is
> >no deflection of the rim relative to the outer ends of the spokes, but there is
> >deflection of the rim relative to the inner (hub) ends of the spokes.
> >
> So in a uniformly tensioned cycle wheel, as it sits in a trueing
> stand, the rim will be permanently compressed, by an amount equal to
> twice the radial component of the tension strain on each spoke.
>
> For example on a radially spoked wheel the amount of elongation in
> each spoke represents half the decrease in the diameter of the rim,
> but in a spoke that is not radial it is only the component of the
> spoke tension in the diametral direction that affects the overall
> diameter of the rim.
If a rim is evenly tensioned, the diameter of the rim will stay the
same. The only way a spoke can elongate is if the ends of the spoke
are fixed.
> >So, the
> >rim deflects over some arc, roughly 40 degrees. The tension in the spokes in
> >that region is reduced, and the spokes get shorter to soak up the deflection.
> >The nipples in that arc remain in contact with the rim, because the spokes were
> >stretched enough when the wheel was built that they are STILL stretched
> >(although less so) when the rim deflects.
> >
> Returning to the case of a ,say, 36 spoke drive wheel being ridden
> (i.e. the subject of this thread), and to keep the maths simple:
> (i) the case where the tension in the 18 drive side spokes is exactly
> twice that of the 18 non drive side
> (ii) There are equal numbers of spokes, of equal gauge, equally
> crossed on each side.
>
> If the rim deflects radially while being ridden over a 40 degree arc,
> then the instantaneous change in tension of the spokes in the contact
> affected zone will be unbalanced.
> I.e. because there are 5 spokes in the 40 degree arc we will
> instantaneously reduce the tension in either:
> a/ three drive side spokes and two offside spokes OR
> b/ three off side spokes and two drive side spokes.
Your reasoning assumes the LAZ is *exactly* 40 degrees. In reality it
is less cleary defined; maximum rim deflection occurs the the point of
contact and gradually decreases from there. In reality there will
always by some imbalance regardless of the number of spokes used.
> In both these cases there should be a measurable effect caused by the
> sideways component of the instantaneous release in spoke tension.
This is true not only because the unequal number of spokes in the LAZ,
but also because the drive and non-drive side spoke strains are not
equal. As the NDS spokes near zero tension (well before the DS does),
the rim will actually deflect towards the DS.
Mark Atanowicz
Mark McMaster
>
> In article <35c7ac78...@news.caverock.net.nz>
> tenn...@caverock.net.nz (Snoopy) writes:
>
> > On Tue, 04 Aug 1998 12:38:32 -0400, Stergios Papadakis
> > <papa...@princeton.edu> wrote:
> > >
> > >In the same way that rubber bands stretch, spokes stretch (only spokes stretch a
> > >lot less). The spokes in a wheel after it is built are actually longer than
> > >they were when they were sitting on the table. So, strictly speaking, there is
> > >no deflection of the rim relative to the outer ends of the spokes, but there is
> > >deflection of the rim relative to the inner (hub) ends of the spokes.
> > >
> > So in a uniformly tensioned cycle wheel, as it sits in a trueing
> > stand, the rim will be permanently compressed, by an amount equal to
> > twice the radial component of the tension strain on each spoke.
> >
> > For example on a radially spoked wheel the amount of elongation in
> > each spoke represents half the decrease in the diameter of the rim,
> > but in a spoke that is not radial it is only the component of the
> > spoke tension in the diametral direction that affects the overall
> > diameter of the rim.
>
> If a rim is evenly tensioned, the diameter of the rim will stay the
> same. The only way a spoke can elongate is if the ends of the spoke
> are fixed.
I'm not sure what you mean by this. The rim in a tension spoke wheel
will be under circumferential compression, so of course it's diameter
will shrink. According to my calculations, a 450 gram constant cross
section aluminum rim with 36 spokes tensioned to 200 lb. each will have
its diameter compressed aprox. 0.02" under spoke tension.
The ends of the spokes don't need to be fixed to elongate, they ends
just have to change their distance from one another (basically the
definition of elongation). Since the spokes are made of an elastic
material, they will always elongage under tension.
Under 200 lb of tension a 1.8mm (15 gauge) spoke will elongate about
0.02". So, we have the rim shrinking by 0.02" in diameter, and the
spokes elongating by 0.02" on opposite sides. Balancing out the
dimensional difference, it looks like the rim only compressess half as
much as two spokes on opposite sides of the wheel (0.02" for the rim vs.
0.02"+0.02" = 0.04" for opposite spokes). How can this be? Simple -
turning the nipples to adjust the spoke tension varies the free length
of the spoke, so that rim compression balances out the spoke elongations
plus the change in free length, until the desired spoke tension is
achieved.
Mark McMaster
MMc...@ix.netcom.com
Mark Atanowicz
Mark McMaster <MMc...@ix.netcom.com> writes:
> I'm not sure what you mean by this. The rim in a tension spoke wheel
> will be under circumferential compression, so of course it's diameter
> will shrink. According to my calculations, a 450 gram constant cross
> section aluminum rim with 36 spokes tensioned to 200 lb. each will have
> its diameter compressed aprox. 0.02" under spoke tension.
>
> The ends of the spokes don't need to be fixed to elongate, they ends
> just have to change their distance from one another (basically the
> definition of elongation). Since the spokes are made of an elastic
> material, they will always elongage under tension.
>
> Under 200 lb of tension a 1.8mm (15 gauge) spoke will elongate about
> 0.02". So, we have the rim shrinking by 0.02" in diameter, and the
> spokes elongating by 0.02" on opposite sides. Balancing out the
> dimensional difference, it looks like the rim only compressess half as
> much as two spokes on opposite sides of the wheel (0.02" for the rim vs.
> 0.02"+0.02" = 0.04" for opposite spokes). How can this be? Simple -
> turning the nipples to adjust the spoke tension varies the free length
> of the spoke, so that rim compression balances out the spoke elongations
> plus the change in free length, until the desired spoke tension is
> achieved.
>
> Mark McMaster
> MMc...@ix.netcom.com
Thanks Mark.
Mark Atanowicz
Nick Kies
wrote:
>On Tue, 04 Aug 1998 12:38:32 -0400, Stergios Papadakis
><papa...@princeton.edu> wrote:
>>
>>In the same way that rubber bands stretch, spokes stretch (only spokes stretch a
>>lot less). The spokes in a wheel after it is built are actually longer than
>>they were when they were sitting on the table. So, strictly speaking, there is
>>no deflection of the rim relative to the outer ends of the spokes, but there is
>>deflection of the rim relative to the inner.....
Boy fellows, now what happens if wheel is at an angle in a corner or
because the guy pedaling is standing on it or what if the ground is
uneven or different tires or low pressure tires or, or, or.
Actually I can see some sense in all this, but it's like waking up
from a dream fantasy.
Snoopy
>>
>> If the rim deflects radially while being ridden over a 40 degree arc,
>> then the instantaneous change in tension of the spokes in the contact
>> affected zone will be unbalanced.
>> I.e. because there are 5 spokes in the 40 degree arc we will
>> instantaneously reduce the tension in either:
>> a/ three drive side spokes and two offside spokes OR
>> b/ three off side spokes and two drive side spokes.
>
>Your reasoning assumes the LAZ is *exactly* 40 degrees. In reality it
>is less cleary defined; maximum rim deflection occurs the the point of
>contact and gradually decreases from there. In reality there will
>always by some imbalance regardless of the number of spokes used.
>
think we are agreed that instantaneously at the bottom of the wheel
there is a LAZ even when there is a spoke directly at the bottom of
the wheel to instantaneously reinforce that point.
I am assuming here that such a spoke will instantaneously buckle to
create the LAZ.
Perhaps the spokes up to two away from the contact point will also
instantaneously buckle, if we are to believe that for a 36 spoke wheel
the LAZ is contained within a 40 degree arc.
But at some point I theorize that the triangulated structure of the
wheel is rigid enough to resist the LAZ effect. I suggest to you
that this cut off point will be where there is a step change in the
stiffness of the structure.
So it may be the LAZ extends one spoke either side of the contact
point (a 20 degree LAZ), or two spokes either side of the contact
point (a 40 degree LAZ) or even three spokes either side of the
contact point (a 60 degree LAZ). Rim stiffness, and applied load
will be a factor in deciding which of these three it will be.
But I submit it WILL be one of these three and not a figure in
between. SNOOPY
Snoopy
<MMc...@ix.netcom.com> wrote:
>>>
>>> For example on a radially spoked wheel the amount of elongation in
>>> each spoke represents half the decrease in the diameter of the rim,
>>> but in a spoke that is not radial it is only the component of the
>>> spoke tension in the diametral direction that affects the overall
>>
>> If a rim is evenly tensioned, the diameter of the rim will stay the
>> same. The only way a spoke can elongate is if the ends of the spoke
>> are fixed.
>
>will be under circumferential compression, so of course it's diameter
>will shrink. According to my calculations, a 450 gram constant cross
>section aluminum rim with 36 spokes tensioned to 200 lb. each will have
>its diameter compressed aprox. 0.02" under spoke tension.
>
>just have to change their distance from one another (basically the
>definition of elongation). Since the spokes are made of an elastic
>material, they will always elongage under tension.
>
>Under 200 lb of tension a 1.8mm (15 gauge) spoke will elongate about
>0.02". So, we have the rim shrinking by 0.02" in diameter, and the
>spokes elongating by 0.02" on opposite sides. Balancing out the
>dimensional difference, it looks like the rim only compressess half as
>much as two spokes on opposite sides of the wheel (0.02" for the rim vs.
>0.02"+0.02" = 0.04" for opposite spokes). How can this be? Simple -
>turning the nipples to adjust the spoke tension varies the free length
>of the spoke, so that rim compression balances out the spoke elongations
>plus the change in free length, until the desired spoke tension is
>achieved.
>
length of a test piece (read spoke here), divided by the original
length of the spoke, expressed as a percentage.
This means that a 300mm spoke which is stretched 3mm under a certain
test load ends up 303mm long and 1% [ (3/300)*100] elongated (i.e
elongation has no units of measurement- it is just a percentage and
the terms 'elongation' and 'percentage elongation' are often
technically understood to mean the same thing).
But Mark (McM), you are talking about 'elongation' here to mean the
amount that the spoke has stretched (in mm or in) are you not?
You are saying here that, for a radial spoke, indeed it is the
tension in the spokes that is responsible for 'shrinking' the wheel.
But you cannot apply tension without stretching the spoke which (by
turning the spoke nipple) decreases the tensioned length (or increases
the free length- the bit that sticks into the rim and punctures your
inner tube) at the same time.
So I put it to you that as all spokes are 'shortened' by movement of
the spoke nipple down the spoke, this causes extra tension which will
stretch the spoke. These two effects will tend to cancel each other
out which means that Mark (Atz)'s assertion that there is no change in
the diameter of the wheel could be correct(?).
What has all this got to do with the subject of this thread? Quite a
lot I think.
If we have a light gauge spoke on the non drive side which uses the
same spoke nipple and nipple thread pitch as on the drive side AND
both sides have the same cross pattern AND both sides spokes are near
enough the same length....
Lets start with a perfectly balanced wheel built as above and then
give all the spoke nipples a half turn tweak. If the gauge of all the
spokes were the same, then there would be no difference in the
geometry of the wheel. But if the spokes on the non-drive side were
lighter gauge, they would elongate more for a given amount of nipple
tweak.
Thus doing up all the spokes 'evenly' on a wheel which has lighter
gauge spokes on the non drive side, will cause the whole wheel to move
in a direction towards the drive end of the drive wheel axle (?)
John Getsoian
>These two effects will tend to cancel each other
>out which means that Mark (Atz)'s assertion that there is no change in
>the diameter of the wheel could be correct(?).
A wheel does shrink a bit when it's built. Consider the stress/strain relationship
for the RIM. The net result of all the spoke tension is to produce substantial
circumferential compression in the rim. Take a free body diagram of one-half a
wheel cut on a horizontal section line. At the ends of the rim section where you
have cut through there must be a compressive force equal to the sum of the
vertical components of spoke tension of all the spokes in that half of the wheel.
By symmetry this same compressive force exists everywhere within the rim.
Thus the rims experiences a circumferential compressive strain in response - i.e.,
it gets smaller in circumference and thus diameter.
> But if the spokes on the non-drive side were
>lighter gauge, they would elongate more for a given amount of nipple
>tweak.
>
>Thus doing up all the spokes 'evenly' on a wheel which has lighter
>gauge spokes on the non drive side, will cause the whole wheel to move
>in a direction towards the drive end of the drive wheel axle (?)
>SNOOPY
Yes. If you start in an undished state, as you tighten the spokes the rim will
compress in response to the net load produced by both sides. This will be
balanced by corresponding tensions and elongations in the spokes. But a for
given elongation, each heavy spoke pulls harder than each light one. Since the
lateral force is the tension times the cosine of the bracing angle, the heavy
spoke side generates more lateral force and the wheel dishes until equilibrium is
acheived between all the radial and lateral forces. Conversely, if the spokes are
all the same thickness, balance in the dished state can only occur with less
elongation of the NDS spokes - QED the original proposition.
regards,
john getsoian
Mark McMaster
We're starting to get into semantics here, but yes, the spoke will
stretch, or strain, or elongate, or whatever term you want to use to
describe the increase in length due to tension.
> You are saying here that, for a radial spoke, indeed it is the
> tension in the spokes that is responsible for 'shrinking' the wheel.
> But you cannot apply tension without stretching the spoke which (by
> turning the spoke nipple) decreases the tensioned length (or increases
> the free length- the bit that sticks into the rim and punctures your
> inner tube) at the same time.
>
> So I put it to you that as all spokes are 'shortened' by movement of
> the spoke nipple down the spoke, this causes extra tension which will
> stretch the spoke. These two effects will tend to cancel each other
> out which means that Mark (Atz)'s assertion that there is no change in
> the diameter of the wheel could be correct(?).
Sorry, but when the spokes are tensioned, the rim must be in net
circumferential compression. For a 36 spoke wheel with 200lb. of
tension per spoke, the rim will be in approx. 1150 lb. of compressive
force. The aluminum of the rim, being an elastic material, will
compress slightly, and it's diameter will shrink. The only force acting
on the rim of an unloaded wheel is the spoke tension, so there is no
restoring force to keep the rim from shrinking.
> What has all this got to do with the subject of this thread? Quite a
> lot I think.
>
> If we have a light gauge spoke on the non drive side which uses the
> same spoke nipple and nipple thread pitch as on the drive side AND
> both sides have the same cross pattern AND both sides spokes are near
> enough the same length....
>
> Lets start with a perfectly balanced wheel built as above and then
> give all the spoke nipples a half turn tweak. If the gauge of all the
> spokes were the same, then there would be no difference in the
> geometry of the wheel. But if the spokes on the non-drive side were
> lighter gauge, they would elongate more for a given amount of nipple
> tweak.
>
> Thus doing up all the spokes 'evenly' on a wheel which has lighter
> gauge spokes on the non drive side, will cause the whole wheel to move
> in a direction towards the drive end of the drive wheel axle (?)
When you say "doing up all the spokes 'evenly'", do you mean that the
nipples are screwed in equal amounts? Why would you want to do that? I
don't know about you, but I screw in the nipples on either side as
necessary to center the wheel on the hub. Even if you did "do up all
the spokes evenly", on a dished wheel with equal size spokes the rim
would be pulled to the left (greater bracing angle so more lateral
pull), so using lighter gauge spokes on the left side would help even it
out.
Besides, the difference in size between the thickest spokes (2.0mm dia.,
3.14mm^2 area) and the thinnest spokes (1.5mm dia., 1.77mm^2 area) is
about a 2:1 ratio of cross-section. The right/left tension differences
on a standard 8/9spd dished wheel is also about a 2:1 ratio. "doing up
all the spokes evenly" with the maximum difference in spoke size will
not pull the rim far to one side or the other, and any lesser difference
in spoke size will still pull the rim to the left.
Mark McMaster
MMc...@ix.netcom.com
Mark McMaster
>
> On 5 Aug 1998 22:40:30 GMT, MAtan...@aol.com (Mark Atanowicz) wrote:
> >>
> >> If the rim deflects radially while being ridden over a 40 degree arc,
> >> then the instantaneous change in tension of the spokes in the contact
> >> affected zone will be unbalanced.
> >> I.e. because there are 5 spokes in the 40 degree arc we will
> >> instantaneously reduce the tension in either:
> >> a/ three drive side spokes and two offside spokes OR
> >> b/ three off side spokes and two drive side spokes.
> >
> >Your reasoning assumes the LAZ is *exactly* 40 degrees. In reality it
> >is less cleary defined; maximum rim deflection occurs the the point of
> >contact and gradually decreases from there. In reality there will
> >always by some imbalance regardless of the number of spokes used.
> >
> I want to explore this concept of the Load Affected Zone further. I
> think we are agreed that instantaneously at the bottom of the wheel
> there is a LAZ even when there is a spoke directly at the bottom of
> the wheel to instantaneously reinforce that point.
>
> I am assuming here that such a spoke will instantaneously buckle to
> create the LAZ.
If the spokes are adequately pre-tensioned, they will not buckle, merely
compress (de-tension) elastically. The wheel and all of its components
will behave as a linearly elastic structure.
> Perhaps the spokes up to two away from the contact point will also
> instantaneously buckle, if we are to believe that for a 36 spoke wheel
> the LAZ is contained within a 40 degree arc.
> But at some point I theorize that the triangulated structure of the
> wheel is rigid enough to resist the LAZ effect. I suggest to you
> that this cut off point will be where there is a step change in the
> stiffness of the structure.
> So it may be the LAZ extends one spoke either side of the contact
> point (a 20 degree LAZ), or two spokes either side of the contact
> point (a 40 degree LAZ) or even three spokes either side of the
> contact point (a 60 degree LAZ). Rim stiffness, and applied load
> will be a factor in deciding which of these three it will be.
> But I submit it WILL be one of these three and not a figure in
> between. SNOOPY
The problems with your theories are that they treat the rim as having
discontinuous "hinge points". It does not. It has continuously varying
arc, much like a spline. Depending on the stiffness of the rim and the
number and stiffnesses of the spokes, the size of the LAZ will vary,
with a typical wheel having a LAZ of about 40 deg. Decreasing the
number of spokes, decreasing their stiffnesses, or increasing the rim
stiffness will increase the length of the LAZ, while changes in the
opposite direction will decrease it.
You don't need to theorize, as the analysis to understand the workings
of the tension spoke wheel are already well known. A good place to
start in looking up this information is the "The Bicycle Wheel" by Jobst
Brandt. This book includes discussions of the mechanics of spoked
wheels, including finite element analyses and diagrams of spoke tensions
and deflections under a variety of loadings. Independent analyses (both
numerical and experimental) have confirmed the theories in this book.
Mark McMaster
MMc...@ix.netcom.com
Damon Rinard
>
> In metallurgy the term 'elongation' generally means the change in the
> length of a test piece (read spoke here),
... exactly ...
> divided by the original
> length of the spoke, expressed as a percentage.
This is "percent elongation", not "elongation". Perhaps inexact use
shortens it to just "elongation", but that is innacurate. Elongation is
the change in length.
> This means that a 300mm spoke which is stretched 3mm under a certain
> test load ends up 303mm long and 1% [ (3/300)*100] elongated
Right. "1% elongated" is the "percent elongation" (note presence of the
percent sign after the number). However, the elongation is 3mm (note
absence of the word percent).
> the terms 'elongation' and 'percentage elongation' are often
> technically understood to mean the same thing).
I would say "contextually understood to mean the same thing."
Interchanging the phrases is technically incorrect.
> Lets start with a perfectly balanced [rear] wheel built as above and then
> give all the spoke nipples a half turn tweak. If the gauge of all the
> spokes were the same, then there would be no difference in the
> geometry of the wheel.
Nope. Try it. In fact, the rim moves left, toward the non-drive side.
--
Damon Rinard
Damon Rinard's Bicycle Tech Site:
http://home.earthlink.net/~rinard/
Snoopy
<MMc...@ix.netcom.com> wrote:
>Snoopy wrote:
>>>Under 200 lb of tension a 1.8mm (15 gauge) spoke will elongate about
>>>0.02". So, we have the rim shrinking by 0.02" in diameter, and the
>>>spokes elongating by 0.02" on opposite sides. Balancing out the
>>>dimensional difference, it looks like the rim only compressess half as
>>>much as two spokes on opposite sides of the wheel (0.02" for the rim vs.
>>>0.02"+0.02" = 0.04" for opposite spokes). How can this be? Simple -
>>>turning the nipples to adjust the spoke tension varies the free length
>>>of the spoke, so that rim compression balances out the spoke elongations
>>>plus the change in free length, until the desired spoke tension is
>>>achieved.
>
>> You are saying here that, for a radial spoke, indeed it is the
>> tension in the spokes that is responsible for 'shrinking' the wheel.
>> But you cannot apply tension without stretching the spoke which (by
>> turning the spoke nipple) decreases the tensioned length (or increases
>> the free length- the bit that sticks into the rim and punctures your
>> inner tube) at the same time.
>>
>> So I put it to you that as all spokes are 'shortened' by movement of
>> the spoke nipple down the spoke, this causes extra tension which will
>> stretch the spoke. These two effects will tend to cancel each other
>> out which means that Mark (Atz)'s assertion that there is no change in
>> the diameter of the wheel could be correct(?).
>
>Sorry, but when the spokes are tensioned, the rim must be in net
>circumferential compression. For a 36 spoke wheel with 200lb. of
>tension per spoke, the rim will be in approx. 1150 lb. of compressive
>force. The aluminum of the rim, being an elastic material, will
>compress slightly, and it's diameter will shrink. The only force acting
>on the rim of an unloaded wheel is the spoke tension, so there is no
>restoring force to keep the rim from shrinking.
>
O.K., I accept that the rim is under tension form all the spokes, and
so therefore must shrink.
But do you agree that turning a spoke nipple through one turn on a
thin gauge spoke will increase the tension less than turning that same
spoke nipple one turn on a thicker gauge spoke?
(Assuming that in each case all spokes are the same gauge, and in the
same relative position between the hub and the rim!) SNOOPY
Snoopy
<rin...@earthlink.net> wrote:
>
>> Lets start with a perfectly balanced [rear] wheel built as above and then
>> give all the spoke nipples a half turn tweak. If the gauge of all the
>> spokes were the same, then there would be no difference in the
>> geometry of the wheel.
>
>Nope. Try it. In fact, the rim moves left, toward the non-drive side.
>
to all spokes on a conventionally spoked balanced rear wheel the rim
will indeed move towards the non-drive side.
This is because the spokes on the off side are at a greater angle,
which means that the sideways component of the tension increment will
be greater toward the off side. SNOOPY
PS However, I stand by my original comment for my own rear wheel,
which is of course dished, but has 24 spokes on the drive side and 12
on the non drive side.
Snoopy
<MMc...@ix.netcom.com> wrote:
>
>If the spokes are adequately pre-tensioned, they will not buckle, merely
>compress (de-tension) elastically. The wheel and all of its components
>will behave as a linearly elastic structure.
>
>discontinuous "hinge points". It does not. It has continuously varying
>arc, much like a spline. Depending on the stiffness of the rim and the
>number and stiffnesses of the spokes, the size of the LAZ will vary,
>with a typical wheel having a LAZ of about 40 deg.
>
are enough spokes to keep the basic geometry of the wheel then what is
wrong with running an 18 spoke wheel (one spoke every 20 degrees)?
Or is there some trade off which means that higher tension is required
to keep a wheel's form with fewer spokes, and with only 18 spokes the
tension per spoke is just too great?
> Decreasing the
>number of spokes, decreasing their stiffnesses, or increasing the rim
>stiffness will increase the length of the LAZ, while changes in the
>opposite direction will decrease it.
>
Agreed, except did you mean to say decreasing the rim stiffness will
increase the length of the LAZ? Or were you saying that a less
stiff rim will have a smaller LAZ, but implying the actual deformation
will be greater within that smaller LAZ?
>
>You don't need to theorize, as the analysis to understand the workings
>of the tension spoke wheel are already well known. A good place to
>start in looking up this information is the "The Bicycle Wheel" by Jobst
>Brandt.
>
Have seen a copy of this years ago, but must get a copy for myself.
Might be a good excuse for an expedition up the Amazon (dotcom).
SNOOPY
Mark McMaster
>
> O.K., I accept that the rim is under tension form all the spokes, and
> so therefore must shrink.
> But do you agree that turning a spoke nipple through one turn on a
> thin gauge spoke will increase the tension less than turning that same
> spoke nipple one turn on a thicker gauge spoke?
> (Assuming that in each case all spokes are the same gauge, and in the
> same relative position between the hub and the rim!) SNOOPY
Yes, I agree that there will be less tension change with each turn of
the nipple on a thinner spoke, i.e. less elastic force for a given
amount of change in spoke length. This was why it was originally
asserted that lighter gauge spokes should be used on the non-drive side
- so that as the spokes are compressed at the bottom of the wheel in the
Load Affected Zone, the non-drive side spokes (which have lower static
tension) will have less change in their tension, and are less likely to
completely de-tension.
Mark McMaster
MMc...@ix.netcom.com
Mark McMaster
>
> On Sat, 08 Aug 1998 14:12:53 -0700, Mark McMaster
> <MMc...@ix.netcom.com> wrote:
> >
> >If the spokes are adequately pre-tensioned, they will not buckle, merely
> >compress (de-tension) elastically. The wheel and all of its components
> >will behave as a linearly elastic structure.
> >
> >The problems with your theories are that they treat the rim as having
> >discontinuous "hinge points". It does not. It has continuously varying
> >arc, much like a spline. Depending on the stiffness of the rim and the
> >number and stiffnesses of the spokes, the size of the LAZ will vary,
> >with a typical wheel having a LAZ of about 40 deg.
> >
> If the rim has a 40deg LAZ and no 'hinge points' then as long as there
> are enough spokes to keep the basic geometry of the wheel then what is
> wrong with running an 18 spoke wheel (one spoke every 20 degrees)?
> Or is there some trade off which means that higher tension is required
> to keep a wheel's form with fewer spokes, and with only 18 spokes the
> tension per spoke is just too great?
With greater gaps between spokes, there are fewer spokes in the LAZ, so
each spoke must carry a greater percentage of the load, and subsequently
their fatigue lives will be shorter. Aerowheels with deep cross section
rims can be made with fewer spokes because their deep cross-section rims
are very stiff which increases the length of the LAZ, so even with
widely spaced spokes there will still be enough spokes in the LAZ for
good distribution of the load.
> > Decreasing the
> >number of spokes, decreasing their stiffnesses, or increasing the rim
> >stiffness will increase the length of the LAZ, while changes in the
> >opposite direction will decrease it.
> >
> Agreed, except did you mean to say decreasing the rim stiffness will
> increase the length of the LAZ? Or were you saying that a less
> stiff rim will have a smaller LAZ, but implying the actual deformation
> will be greater within that smaller LAZ?
The second one. A less stiff rim will deform more in a smaller region,
causing greater detensioning of the spokes in a smaller more
concentrated region, increasing the load each of these fewer spokes must
bear. A stiffer rim will bend less, and acts as an arch to distribute
the load over a wider region.
Mark McMaster
MMc...@ix.netcom.com
Snoopy
Dingley) wrote:
>
>>Without having a way to calculate the ideal spoke tension
>
>So, does anyone know what it is ? Is the ideal tension proportional
>to the spoke cross section (assuming spokes have homogeneous
>metallurgy), or the spoke diameter (assuming that all load is
>concentrated in a thin surface shell), or is it something else
>entirely ?
>
I think it may be helpful to look at the extremes when considering
what spoke tension is 'ideal'.
If the spoke tension is too loose, the spokes will lose all tension in
the load affected zone while in use.
If the spoke tension is too tight the spoke or the rim my permanently
deform as the wheel is being built.
IMO spoke tension could be 'too low' for a given rider and gear, but
perhaps fine for a lighter rider. If the spoke tension is 'too
high', then that is a metallurgical consideration.
In practical terms, the maximum stress a spoke can take will be
directly proportional to it's minimum cross sectional area, and there
is definitely no 'thin surface shell' affect as is speculated above.
Snoopy
Dingley) wrote:
>On 12 Jul 1998 00:30:44 GMT, jac...@silo.csci.unt.edu (Bruce Jackson)
>wrote:
>
>>Without having a way to calculate the ideal spoke tension
>
>We don't actually need this for this example, just the ratio of IST
>for thick & thin spokes. As you say though, we don't know that either.
>
The load on a wheel will cause a decrease in the tension of spokes in
that pretensioned wheel in the LAZ. If the spokes lose tension for a
particular rider/load combination we can say the tension is 'less than
ideal' for that particular rider/load.
The solution is to tighten the spokes, and for high tensile steel
spokes I can't see why they can't be tightened to up to 100% of the
spoke steel yield point (as any wheel loads during riding will only
reduce stresses in the already prestressed wheel).
The maximum spoke stress (spoke tension/minimum spoke Xsection area),
should be able to be calculated from the material properties of the
spoke.
So I would argue that the 'ideal spoke tension range' depends on the
vertical load on the wheel as a lower bound and maximum tension in the
spoke (depends on spoke properties, and the extent of the LAZ) as an
upper bound, and so the 'ideal spoke tension' can in fact be
calculated in a straightforward manner.
But for a 'thin' spoke the upper bound will be lower. SNOOPY
Snoopy
<MMc...@ix.netcom.com> wrote:
>> If the rim has a 40deg LAZ and no 'hinge points' then as long as there
>> are enough spokes to keep the basic geometry of the wheel then what is
>> wrong with running an 18 spoke wheel (one spoke every 20 degrees)?
>> Or is there some trade off which means that higher tension is required
>> to keep a wheel's form with fewer spokes, and with only 18 spokes the
>> tension per spoke is just too great?
>
>With greater gaps between spokes, there are fewer spokes in the LAZ, so
>each spoke must carry a greater percentage of the load, and subsequently
>their fatigue lives will be shorter. Aerowheels with deep cross section
>rims can be made with fewer spokes because their deep cross-section rims
>are very stiff which increases the length of the LAZ, so even with
>widely spaced spokes there will still be enough spokes in the LAZ for
>good distribution of the load.
>
Fatigue occurs in a spoke because of a cyclic change in spoke loading
which is superimposed over the overall spoke tension. The amount of
this fluctuation depends only on the rider load on the wheel.
However, for a highly tensioned spoked wheel, the change in tension
(relaxation) will be less for a given rider/load, which means the
fatigue life of a highly tensioned wheel should be better than a
lower tensioned wheel carrying the same rider/load.
I submit that the fatigue life of a spoke depends on this factor
alone, and not the Load Affected Zone.
Or would I be more correct to say that the the tighter the spokes, the
more extensive the LAZ, so what I am saying is actually in agreement
with what you have written above? SNOOPY
Jobst Brandt
> So, does anyone know what it is ? Is the ideal tension proportional
> to the spoke cross section (assuming spokes have homogeneous
> metallurgy), or the spoke diameter (assuming that all load is
> concentrated in a thin surface shell), or is it something else
> entirely ?
Unless you have a non standard wheel, spoke tension is prescribed by
what the rim can take and not limited by the spokes. The rims of most
28 or more spoked wheels can be collapsed by spoke tension before the
spokes reach half of their yield tension.
Jobst Brandt <jbr...@hpl.hp.com>
Mark McMaster
>
> On Mon, 10 Aug 1998 22:21:01 -0700, Mark McMaster
> <MMc...@ix.netcom.com> wrote:
>
> >> If the rim has a 40deg LAZ and no 'hinge points' then as long as there
> >> are enough spokes to keep the basic geometry of the wheel then what is
> >> wrong with running an 18 spoke wheel (one spoke every 20 degrees)?
> >> Or is there some trade off which means that higher tension is required
> >> to keep a wheel's form with fewer spokes, and with only 18 spokes the
> >> tension per spoke is just too great?
> >
> >With greater gaps between spokes, there are fewer spokes in the LAZ, so
> >each spoke must carry a greater percentage of the load, and subsequently
> >their fatigue lives will be shorter. Aerowheels with deep cross section
> >rims can be made with fewer spokes because their deep cross-section rims
> >are very stiff which increases the length of the LAZ, so even with
> >widely spaced spokes there will still be enough spokes in the LAZ for
> >good distribution of the load.
> >
> Fatigue occurs in a spoke because of a cyclic change in spoke loading
> which is superimposed over the overall spoke tension. The amount of
> this fluctuation depends only on the rider load on the wheel.
The magnitude of the loading cycle on the individual spokes also depends
on how many spokes are supporting the load. With more spokes in the
LAZ, each spoke takes a smaller percentage of the load, so the magnitude
of the loading cycles is smaller for each spoke.
> However, for a highly tensioned spoked wheel, the change in tension
> (relaxation) will be less for a given rider/load, which means the
> fatigue life of a highly tensioned wheel should be better than a
> lower tensioned wheel carrying the same rider/load.
This is incorrect. As long as there is enough static tension to prevent
any of the spokes from completely slackening, the wheel will behave in a
linearly elastic way. The static tension has no effect on the dynamic
loading cycles, since the dynamic load will be super-imposed over the
static load. Static loads (mean stresses) can actually reduce fatigue
life, but since spokes are tensioned well below their yield strength,
static spoke loads have little effect on their fatigue life.
> I submit that the fatigue life of a spoke depends on this factor
> alone, and not the Load Affected Zone.
> Or would I be more correct to say that the the tighter the spokes, the
> more extensive the LAZ, so what I am saying is actually in agreement
> with what you have written above?
No, since the wheel is a linearly elastic structure, the size of the LAZ
is determined by the relative stiffnesses of the components (spokes and
rim), and not on their static pre-loads.
Mark McMaster
MMc...@ix.netcom.com
Mark McMaster
>
> The load on a wheel will cause a decrease in the tension of spokes in
> that pretensioned wheel in the LAZ. If the spokes lose tension for a
> particular rider/load combination we can say the tension is 'less than
> ideal' for that particular rider/load.
> The solution is to tighten the spokes, and for high tensile steel
> spokes I can't see why they can't be tightened to up to 100% of the
> spoke steel yield point (as any wheel loads during riding will only
> reduce stresses in the already prestressed wheel).
Except that for most wheels, you will exceed the yeild point of the
rim
before you reach the yield point of the rim. However, the yield point
of
the rim (minus some factor of safety) is an acceptable maximum limit
for
spoke tension.
> The maximum spoke stress (spoke tension/minimum spoke Xsection area),
> should be able to be calculated from the material properties of the
> spoke.
Or can be taken from published data for major manufacturer's spokes.
> So I would argue that the 'ideal spoke tension range' depends on the
> vertical load on the wheel as a lower bound and maximum tension in the
> spoke (depends on spoke properties, and the extent of the LAZ) as an
> upper bound, and so the 'ideal spoke tension' can in fact be
> calculated in a straightforward manner.
Wheel loads are highly dynamic, when you take into consideration
hitting
potholes and such. Using the rim yield point as the limit of spoke
tension will maximize the strenghth of the wheel, so this can be
considered the optimum tension.
> But for a 'thin' spoke the upper bound will be lower.
Since spoke tension in a typical wheel is limited by rim strength, not
spoke strength, the optimum tension will be the same regardless of
spoke
thickness.
Mark McMaster
MMc...@ix.netcom.com
tku...@diabloresearch.com
MMc...@ix.netcom.com wrote:
> Snoopy wrote:
> >
> > I submit that the fatigue life of a spoke depends on this factor
> > alone, and not the Load Affected Zone.
> > Or would I be more correct to say that the the tighter the spokes, the
> > more extensive the LAZ, so what I am saying is actually in agreement
> > with what you have written above?
>
> No, since the wheel is a linearly elastic structure, the size of the LAZ
> is determined by the relative stiffnesses of the components (spokes and
> rim), and not on their static pre-loads.
Hmm, I don't think that we need be pedantic about spoke strength. In fact,
the rim stiffness is the major contributor to the LAZ limits. I think that
we agree that if the rim is rigid enough that the LAZ is essentially the
entire structure.
This would indicate that there would be few failures of spokes on super
aero rims and I think that that is the case.
-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
Mark McMaster
>
> In article <35D241...@ix.netcom.com>,
> MMc...@ix.netcom.com wrote:
> > Snoopy wrote:
> > >
> > > I submit that the fatigue life of a spoke depends on this factor
> > > alone, and not the Load Affected Zone.
> > > Or would I be more correct to say that the the tighter the spokes, the
> > > more extensive the LAZ, so what I am saying is actually in agreement
> > > with what you have written above?
> >
> > No, since the wheel is a linearly elastic structure, the size of the LAZ
> > is determined by the relative stiffnesses of the components (spokes and
> > rim), and not on their static pre-loads.
>
> Hmm, I don't think that we need be pedantic about spoke strength. In fact,
> the rim stiffness is the major contributor to the LAZ limits. I think that
> we agree that if the rim is rigid enough that the LAZ is essentially the
> entire structure.
Certainly, rim stiffness is a major contributor, but don't overlook
spoke stiffness effecting the size of the LAZ. This is the reason that
swaged (butted spokes) are quite effective; by being more elastic than
straight spokes they increase the size of the LAZ to allow more spokes
to share the load.
My examinations of deep rim aerowheels indicate that there is still a
distinct load effected zone - for Campagnolo 12 and 16 spoke Shamal
wheels it is about 3 to 5 spokes long (although that does incompass
about 1/4 of the circumference of the wheel).
> This would indicate that there would be few failures of spokes on super
> aero rims and I think that that is the case.
I've seen enough spoke broken in deep-rimmed aero wheels to know that
they are not completely immune to spoke breakage - afterall, Campagnolo
increase the number of spokes in their Shamal rear wheel from 12 (in
'97) to 16 in (in '98) while using the same rim due to spoke breakage
problems.
Mark McMaster
MMc...@ix.netcom.com
Snoopy
<MMc...@ix.netcom.com> wrote:
>> The solution is to tighten the spokes, and for high tensile steel
>> spokes I can't see why they can't be tightened to up to 100% of the
>> spoke steel yield point (as any wheel loads during riding will only
>> reduce stresses in the already prestressed wheel).
>
>Except that for most wheels, you will exceed the yield point of the rim
>before you reach the yield point of the spoke. However, the yield point of
>the rim (minus some factor of safety) is an acceptable maximum limit for
>spoke tension.
>
I see two reasons a 'factor of safety' when building a wheel.
a/ Uncertainty as to the manufactured strength of the rim. I
presume the weakest point will be the join, and also presume the rim
maker will be aware of this. So any welding will be under closely
controlled conditions and any claimed 'rim strength' will be quoted so
that 99% of the rims off the production line meet that strength. So
in practice do you really need a factor of safety for this?
Curiously I looked in the 1998 Mavic catalogue but couldn't see any
claims for rim strength. Is there any technical rim strength info on
the net?
b/ Uncertainty as to the load placed on the wheel. It is certainly
possible to build wheels to a known tension. If you superimpose the
load of a rider, or the inertial reaction of a sharp bump these forces
are acting in the opposite direction to the tension in the spoke and
should instantaneously REDUCE spoke tension (and the associated rim
tension). All events that I can forsee will reduce the stress on a
wheel, so do you really need a factor of safety for this?
I repeat my original assertion that it makes sense to build a such
that the weakest link (the rim), is stressed right up to it's yield
point.
>
>Wheel loads are highly dynamic, when you take into consideration
>hitting potholes and such.
>
True, but I can't see how that fact ties in with the next sentence in
your paragraph (below). Hitting a pothole is another example of
instanttaneously REDUCED spoke tension. SNOOPY
>Using the rim yield point as the limit of spoke
>tension will maximize the strength of the wheel, so this can be
>considered the optimum tension.
>
Snoopy
>
>Unless you have a non standard wheel, spoke tension is prescribed by
>what the rim can take and not limited by the spokes. The rims of most
>28 or more spoked wheels can be collapsed by spoke tension before the
>spokes reach half of their yield tension.
>
ability of the rim to resist collapse under very high stress the issue
here?
IMO fatigue life of the wheel would be more important. And since
aluminium (rims) have a much lower fatigue life than steel (spokes)--
(or to be more technically exact aluminium has a lower 'endurance
limit' than high tensile steel), is the reason that a rim is more
likely to fail simply because it is usually made from aluminium?
SNOOPY
Snoopy
<MMc...@ix.netcom.com> wrote:
>>
>> Fatigue occurs in a spoke because of a cyclic change in spoke loading
>> which is superimposed over the overall spoke tension. The amount of
>> this fluctuation depends only on the rider load on the wheel.
>
>The magnitude of the loading cycle on the individual spokes also depends
>on how many spokes are supporting the load. With more spokes in the
>LAZ, each spoke takes a smaller percentage of the load, so the magnitude
>of the loading cycles is smaller for each spoke.
>
>
>> However, for a highly tensioned spoked wheel, the change in tension
>> (relaxation) will be less for a given rider/load, which means the
>> fatigue life of a highly tensioned wheel should be better than a
>> lower tensioned wheel carrying the same rider/load.
>
now, like you, believe it is incorrect also, but I take issue with
some of your explanation as to why this is so, below.
Thinking about it more, the higher the tension in the spokes means
that spoke relaxation when going over a pothole will be less. But
unless you live in a street of potholes your wheel will not hit a pot
hole on every revolution.
It is obviously very important that a wheel is built with enough
tension to withstand potholes. You could argue that going over many
potholes over the life of the bike, is what subjects the wheel to
fatigue loading (true).
But far more frequent (albeit much less in magnitude) is the fatigue
loading caused by just normal riding on a flat surface (the on again
off again weight of the rider which the wheel spokes and rim see as
the wheel rolls along the road).
To really understand the fatigue life of a wheel you have to combine
the effects of (a) potholes and (b) flat road riding.
However, I am now claiming that if you design a wheel to resist type
(a) fatigue, that wheel may not be the best in resisting type (b)
fatigue and vica versa (see explanation further on).
>This is incorrect. As long as there is enough static tension to prevent
>any of the spokes from completely slackening, the wheel will behave in a
>linearly elastic way.
Are you saying you cannot suffer fatigue failure in a wheel that
behaves in a linearly elastic way? If so, I believe you are wrong.
>The static tension has no effect on the dynamic
>loading cycles, since the dynamic load will be super-imposed over the
>static load.
And since it is the dynamic load (in this case the on again off again
weight of the rider that each spoke in the wheel 'sees' as it
rotates), which causes wheel fatigue, I don't think your next sentence
follows.
>Static loads (mean stresses) can actually reduce fatigue life,
However, I do agree with your statement immediately above, that
static loads (from the point of view of the wheel, that is the tension
in the spokes), does affect the fatigue life. I think it was fatigue
test results collated by Soderberg which first documented this
phenomenom. Soderberg showed that test piece pre-loading can affect
the size of a dynamic load that a test piece can stand.
To paraphrase Soderberg's results in 'spoke' terms:
If a highly tensioned spoke (high static load), will only be able to
withstand a small rider (low dynamic load), because of a projected low
fatigue life
THEN if the overall wheel spoke tension is reduced (lower static
load), the the fatigue life of the spokes will increase, which will
suit a larger rider (higher dynamic load).
Now this doesn't seem too intuitive and clearly if you are building a
wheel to minimize 'road-riding' wheel fatigue, you should not reduce
spoke tension too much as over the first decent pot hole the wheel
could collapse. But neverthless reducing overall wheel tension to
minimize 'road riding' wheel fatigue failure is consistant with
Soderbergs experiments. SNOOPY
Jobst Brandt
>> Except that for most wheels, you will exceed the yield point of the
>> rim before you reach the yield point of the spoke. However, the
>> yield point of the rim (minus some factor of safety) is an
>> acceptable maximum limit for spoke tension.
> I see two reasons a 'factor of safety' when building a wheel. a/
> Uncertainty as to the manufactured strength of the rim. I presume
> the weakest point will be the join, and also presume the rim maker
> will be aware of this.
I assume you mean the rim joint. I think you don't have a good
visualization of the loads in the rim. Unless the wheel strikes a
concentrated obstacle, the rim is always in compression. That is why
riveting or welding the rim joint is a wasted effort. If there were
tensile loads in that area, rims would fail there regularly, but in
fact, rims fail there only in destructive crashes where the wheel is
folded anyway.
> So any welding will be under closely controlled conditions and any
> claimed 'rim strength' will be quoted so that 99% of the rims off
> the production line meet that strength. So in practice do you
> really need a factor of safety for this?
I don't know what you are proposing here. You seem to be convinced of
something manufacturers are not nor is there any evidence of it.
> Curiously I looked in the 1998 Mavic catalogue but couldn't see any
> claims for rim strength. Is there any technical rim strength info on
> the net?
In what units do you expect to see that? At best, there could be a
maximum spoke tension given, possibly in total tension, because these
rims are offered in several drillings.
> Uncertainty as to the load placed on the wheel. It is certainly
> possible to build wheels to a known tension. If you superimpose the
> load of a rider, or the inertial reaction of a sharp bump these
> forces are acting in the opposite direction to the tension in the
> spoke and should instantaneously REDUCE spoke tension (and the
> associated rim tension). All events that I can forsee will reduce
> the stress on a wheel, so do you really need a factor of safety for
> this?
I suggest you read up on this and not retrace well trodden territory.
The safety factor is discussed and explained. Rim stress increases
over the static stress in some instances that occur often.
> I repeat my original assertion that it makes sense to build a such
> that the weakest link (the rim), is stressed right up to it's yield
> point.
I think you missed some points in your assessment of the loads
involved.
> Message posted by SNOOPY
So who is Snoopy and why are you afraid to reveal your identity?
Jobst Brandt <jbr...@hpl.hp.com>
Jobst Brandt
> Thinking about it more, the higher the tension in the spokes means
> that spoke relaxation when going over a pothole will be less. But
> unless you live in a street of potholes your wheel will not hit a pot
> hole on every revolution.
Pot holes are terrible examples because their main impact is the same
as running into a square curb, it is a rim denting event, not an
example of high loading. A far better example is a hand patched
asphalt road or washboard, both of which are relatively continuous and
not dent inducing.
Riding over such uneven surfaces or dropping off a high curb causes
loads great enough to slacken spokes. This does not materially
increase tension in any spokes although it increases rim stress
through bending. The inside of the bend increases tension (decreases
compression) while the outside increases compression, adding to the
compression caused by spoke tension.
> It is obviously very important that a wheel is built with enough
> tension to withstand potholes. You could argue that going over many
> potholes over the life of the bike, is what subjects the wheel to
> fatigue loading (true).
It is not obvious and if it is, please explain why.
> But far more frequent (albeit much less in magnitude) is the fatigue
> loading caused by just normal riding on a flat surface (the on again
> off again weight of the rider which the wheel spokes and rim see as
> the wheel rolls along the road).
I think you are just mouthing jargon you picked up here on the net.
How about thinking this out on a scratch pad before spreading this
stuff on the net. This sort of (un)thinking on-line is like a chain
letter, it propagates itself into urban legend by repetition.
> To really understand the fatigue life of a wheel you have to combine
> the effects of (a) potholes and (b) flat road riding.
> However, I am now claiming that if you design a wheel to resist type
> (a) fatigue, that wheel may not be the best in resisting type (b)
> fatigue and vica versa (see explanation further on).
Oh drive. Cut it out.
> Are you saying... suffer fatigue failure... linearly elastic...
> static tension... dynamic loading cycles... dynamic load super-imposed
> on again off again weight of the rider...
Sounds like science.
> To paraphrase Soderberg's results in 'spoke' terms:
Oh? A little misplaced information can cause large misinterpretations.
Jobst Brandt <jbr...@hpl.hp.com>
Rick Denney
wrote:
>
>Are you saying you cannot suffer fatigue failure in a wheel that
>behaves in a linearly elastic way? If so, I believe you are wrong.
Spokes that fail from fatigue had some portion of the spoke that was
yielded as a result internal stresses induced during the formation of
the spoke. The fatigue crack initiates in these yielded areas, and
then the point of the crack is such a stress riser that it yields, and
so on, until the crack propagates far enough such that the remaining
material cannot hold the tension of the spoke. In spokes that break,
the portion that ultimately ruptures is surprisingly small compared to
the portion that cracks from fatigue.
>
>>The static tension has no effect on the dynamic
>>loading cycles, since the dynamic load will be super-imposed over the
>>static load.
>
>And since it is the dynamic load (in this case the on again off again
>weight of the rider that each spoke in the wheel 'sees' as it
>rotates), which causes wheel fatigue, I don't think your next sentence
>follows.
>
>>Static loads (mean stresses) can actually reduce fatigue life,
>
>However, I do agree with your statement immediately above, that
>static loads (from the point of view of the wheel, that is the tension
>in the spokes), does affect the fatigue life. I think it was fatigue
>test results collated by Soderberg which first documented this
>phenomenom. Soderberg showed that test piece pre-loading can affect
>the size of a dynamic load that a test piece can stand.
>
>To paraphrase Soderberg's results in 'spoke' terms:
>If a highly tensioned spoke (high static load), will only be able to
>withstand a small rider (low dynamic load), because of a projected low
>fatigue life
>THEN if the overall wheel spoke tension is reduced (lower static
>load), the the fatigue life of the spokes will increase, which will
>suit a larger rider (higher dynamic load).
>
>Now this doesn't seem too intuitive and clearly if you are building a
>wheel to minimize 'road-riding' wheel fatigue, you should not reduce
>spoke tension too much as over the first decent pot hole the wheel
>could collapse. But neverthless reducing overall wheel tension to
>minimize 'road riding' wheel fatigue failure is consistant with
>Soderbergs experiments. SNOOPY
>
Remember, however, that the static load in a spoke causes the greatest
stress the spoke normally sees, and the dynamic load reduces the
stress. The dynamic load is a compressive load that is superposed on a
static tensile load. If the static load is below the fatigue limit for
the spoke, then dynamic loads should have no effect. Spokes that do
not have internal stresses above the fatigue limit do not fail in
practice. Potholes do not cause spoke fatigue, but they do cause rim
failures. Potholes increase the dynamic range of loads, but that range
is extended at the low end, with the high end of the load range
represented by static tensile load.
Rims fail at potholes because the beam strength of the rim between the
spokes is inadequate and they buckle, or because the loading on the
wheel increases to the point where the spokes go slack, which takes
the lateral support away from the rim, allowing a collapse.
Rick Denney
Take what you want and leave the rest.
Tom Kunich
>
> I assume you mean the rim joint. I think you don't have a good
> visualization of the loads in the rim. Unless the wheel strikes a
> concentrated obstacle, the rim is always in compression. That is why
> riveting or welding the rim joint is a wasted effort. If there were
> tensile loads in that area, rims would fail there regularly, but in
> fact, rims fail there only in destructive crashes where the wheel is
> folded anyway.
I want to take exception to this statement because I think that it
needs a lot of discussion. I don't see it at ALL as obvious that
rims don't COMMONLY fail at the rim joint.
1) A tandem came into the shop a number of months (maybe 8) ago.
The owner is blind and stokes and he gets several different
captains but mostly one guy who is quite experienced with him.
His rear wheel was folded over at the joint at almost 90 degrees.
He said that they simply came to a stop and while they were sitting
at the stop sign the bike rim collapsed.
2) It happened to the same guy a second time. Possibly a third!
3) There is a picture in Wheelsmith upstairs that shows three
track guys sprinting and one of them has a wheel that is in
the process of folding over right at the rim joint.
Obviously all of this bothers the heck out of me. What's going
on?
Mark McMaster
>
> On 12 Aug 1998 16:37:21 GMT, jbr...@hpl.hp.com (Jobst Brandt) wrote:
> >
> >Unless you have a non standard wheel, spoke tension is prescribed by
> >what the rim can take and not limited by the spokes. The rims of most
> >28 or more spoked wheels can be collapsed by spoke tension before the
> >spokes reach half of their yield tension.
> >
> An interesting fact that I have no reason to doubt. However, is the
> ability of the rim to resist collapse under very high stress the issue
> here?
> IMO fatigue life of the wheel would be more important. And since
> aluminium (rims) have a much lower fatigue life than steel (spokes)--
> (or to be more technically exact aluminium has a lower 'endurance
> limit' than high tensile steel), is the reason that a rim is more
> likely to fail simply because it is usually made from aluminium?
Well, your knowledge of materials (or lack there of) is becoming
apparant - aluminum has no endurance limit.
The rim failure from excess spoke tension is a yield failure, not a
fatigue failure, and is related to the yield strength of the rim
material (which is usually, but not always, aluminum) and it's cross
sectional area.
Mark McMaster
MMc...@ix.netcom.com
Jobst Brandt
I don't believe a word of "sitting at the stop sign the bike rim
collapsed". I understand enough about wheel so know that a wheel that
can be ridden for a reasonable distance, accelerating and stopping,
cannot suddenly collapse while standing there. That you pass such
hearsay on as evidence od anything is like forwarding a chain letter.
You keep a false notion alive.
The Wheelsmith picture is a classic example of what I mentioned in the
response above. This is a destructive crash that folded the wheel.
At such moments there is no tension in the wheel and if you are lucky,
the joint will be in the bending zone where it will separate. If not
the rim will taco and the joint remains closed.
This changes nothing about that the rim is always in compression under
operating conditions. I get to testify in liability claims where
people believe as you that rims fail because they have a joint. None
of these cases of misplaced machismo has prevailed... they are, after
all exclusively men who bring these cases because they cannot admit
they fell from the bicycle through their own fault. It must have been
a material or design failure.
Jobst Brandt <jbr...@hpl.hp.com>
Snoopy
<MMc...@ix.netcom.com> wrote:
>> IMO fatigue life of the wheel would be more important. And since
>> aluminium (rims) have a much lower fatigue life than steel (spokes)--
>> (or to be more technically exact aluminium has a lower 'endurance
>> limit' than high tensile steel), is the reason that a rim is more
>> likely to fail simply because it is usually made from aluminium?
>
>Well, your knowledge of materials (or lack there of) is becoming
>apparant - aluminum has no endurance limit.
>
So what you are saying is that for an aluminium part under an on-again
off-again load (fatigue) there is actually no fatigue load low enough
that can be applied to the aluminium part without it failing
eventually? And so therefore, unlike mild steel, there is no
'endurance limit'?
Well that is correct- I did know that.
However, it is still true that if the aluminium part is subject to a
particular fatigue load it will last longer than that same part
subjected to a larger fatigue load- and that was my point here.
I have a design reference which quotes an 'endurance limit' for an
Aluminium Alloy after 500 million load applications. This is
technically nonsense, as you have pointed out, as this is when the
test piece may break, which would not happen if there was a genuine
'endurance limit'.
What I was suggesting was that the rim breaking first might suggest
that the endurance strength of the aluminium rim eyelet is less than
the endurance strength of the steel spoke nipple- that was the point
of the post.
Perhaps you should have also pointed out that high strength steel (the
145ksi/ 1000MPa strength) that spokes are made from does not have an
'endurance limit' either. A genuine 'endurance limit' is only found
in mild steels.
>The rim failure from excess spoke tension is a yield failure, not a
>fatigue failure, and is related to the yield strength of the rim
>material (which is usually, but not always, aluminum) and it's cross
>sectional area.
>
Fair enough, but while I don't doubt you could get a rim to fail this
way, could it ever happen while riding. Is the fatigue life of the
rim not more important in a real riding situation? SNOOPY
Snoopy
>Snoopy writes anonymously:
>
><message snipped>
>
>> Message posted by SNOOPY
>
>So who is Snoopy and why are you afraid to reveal your identity?
>
Snoopy flies through cyberspace.
Snoopy 'snoops' for information.
Snoopy is a cyclist.
Snoopy hasn't got your book yet Jobst, (but it is on my list!)
Snoopy knows enough about engineering to ask the curly questions.
Snoopy respects the collective wisdom of rec.bicycles.tech.
Snoopy respects the messangers on this group, but Snoopy may still
question any message posted.
Snoopy is the only Snoopy on rec.bicycles.tech
Jobst is the only Jobst on rec.bicycles.tech
So there is no confusion- nothing is hidden.
Snoopy doesn't always get things right.
Snoopy will be judged on his postings.
Snoopy does not hide. Snoopy is right here.
_
/ \_
/ (_)---_
( ( / {)
---_ \__ !_ _--/
-_ \/^\/^\__)_(
\-_-^--^ __}--O0
\-/) \__0O
/ (")
__<=>(___!!___
/______________\
/________________\
/________________ _\
!----------------- !
!X Red Baron ------!
!X( might be you ) !
!------------------!
!__________________!
Well you did ask.................SNOOPY
Snoopy
>I assume you mean the rim joint. I think you don't have a good
>visualization of the loads in the rim. Unless the wheel strikes a
>concentrated obstacle, the rim is always in compression.
>
>> So any welding will be under closely controlled conditions and any
>> claimed 'rim strength' will be quoted so that 99% of the rims off
>> the production line meet that strength. So in practice do you
>> really need a factor of safety for this?
>
>I don't know what you are proposing here.
>
I am proposing that most rim manufacturers know what they are doing
and it is unlikely that one rim off a production line will have a
claimed strength significantly less than another off the same
production line.
I am agreeing with you that the rim is always under compression.
I am asking if it is even possible for the rim to be under more
compression than static compression?
><cut and pasted from another Jobst post on this thread>
>Riding over such uneven surfaces (potholes) or dropping off a high curb
> causes loads great enough to slacken spokes. This does not materially
>increase tension in any spokes although it increases rim stress
>through bending. The inside of the bend increases tension (decreases
>compression) while the outside increases compression, adding to the
>compression caused by spoke tension.
>
If there is no possibility of increasing the compression on the rim
while riding, then why not apply maximum compression to the rim when
building it? Is there anything wrong with building a rim this way?
So Jobst is saying above that if you go over a pot hole or drop off a
high curb the compression in the rim can go above the material
compressive yield point. Fair enough.
But can someone please explain how this 'adds to the compression
caused by the spoke tension' if the spokes have gone slack?
>
>> Curiously I looked in the 1998 Mavic catalogue but couldn't see any
>> claims for rim strength. Is there any technical rim strength info on
>> the net?
>
>In what units do you expect to see that? At best, there could be a
>maximum spoke tension given, possibly in total tension, because these
>rims are offered in several drillings.
>
Yes, it might be useful to have information on the maximum total
tension that a rim can take. After all, the equivalent information
is available for spokes.
>
>> Uncertainty as to the load placed on the wheel. It is certainly
>> possible to build wheels to a known tension. If you superimpose the
>> load of a rider, or the inertial reaction of a sharp bump these
>> forces are acting in the opposite direction to the tension in the
>> spoke and should instantaneously REDUCE spoke tension (and the
>> associated rim tension). All events that I can forsee will reduce
>> the stress on a wheel, so do you really need a factor of safety for
>> this?
>
Sorry, the above sentance should read
REDUCE spoke tension (and the associated rim compression)
>I suggest you read up on this and not retrace well trodden territory.
Yes, I am planning to get your book, but the answer was not in the
rec.bicycles.tech FAQ.
>The safety factor is discussed and explained. Rim stress increases
>over the static stress in some instances that occur often.
>
If so then I accept a safety factor is needed. SNOOPY
Snoopy
>Snoopy writes anonymously:
>
>> Thinking about it more, the higher the tension in the spokes means
>> that spoke relaxation when going over a pothole will be less. But
>> unless you live in a street of potholes your wheel will not hit a pot
>> hole on every revolution.
>
>Pot holes are terrible examples because their main impact is the same
>as running into a square curb, it is a rim denting event, not an
>example of high loading. A far better example is a hand patched
>asphalt road or washboard, both of which are relatively continuous and
>not dent inducing.
>
I picked a pothole as an extreme example. There was no intention to
imply a rim denting event. As Jobst has noted above a washboard road
is a better example.
>
>> It is obviously very important that a wheel is built with enough
>> tension to withstand potholes. You could argue that going over many
>> potholes over the life of the bike, is what subjects the wheel to
>> fatigue loading (true).
>
>It is not obvious and if it is, please explain why.
>
Rephrase the above by replacing pothole with 'washboard road'.
What I am saying is that by building a wheel as a solid elastic
structure for smooth roads, will not guarantee it remains a solid
elastic structure if used on washboard roads.
>> But far more frequent (albeit much less in magnitude) is the fatigue
>> loading caused by just normal riding on a flat surface (the on again
>> off again weight of the rider which the wheel spokes and rim see as
>> the wheel rolls along the road).
>
>I think you are just mouthing jargon you picked up here on the net.
>How about thinking this out on a scratch pad before spreading this
>stuff on the net. This sort of (un)thinking on-line is like a chain
>letter, it propagates itself into urban legend by repetition.
>
No, it is up to people such as yourself to flame such 'urban legends',
if appropriate. But labelling an explanation as 'jargon' does not
refute the explanation. Perhaps you would like to be more specific
as to what part of my explanation you disagree with. Are you saying
there is no rim/spoke fatigue loading when riding along a smooth road?
Usenet is a discussion forum. I will admit I am wrong if proved to
be so, and I hope other contributors will do likewise. This is an
unmoderated forum and as such anyone believing all the 'facts'
espoused here would be a fool indeed.
>
>> To really understand the fatigue life of a wheel you have to combine
>> the effects of (a) potholes and (b) flat road riding.
>> However, I am now claiming that if you design a wheel to resist type
>> (a) fatigue, that wheel may not be the best in resisting type (b)
>> fatigue and vica versa (see explanation further on).
>
>Oh drivel. Cut it out.
>
Well replace the term pothole with washboard road.
Is it still drivel? Well how would you interpret wheel building in
terms of a Soderberg diagram?
Let me put the question another way. Will increasing your spoke
tension by 5% all round on an already well built wheel increase the
life of the wheel or not?
>
>> To paraphrase Soderberg's results in 'spoke' terms:
>
>Oh? A little misplaced information can cause large misinterpretations.
>
Indeed, but you haven't explained why the information IS misplaced.
Mark McMaster
> What I was suggesting was that the rim breaking first might suggest
> that the endurance strength of the aluminium rim eyelet is less than
> the endurance strength of the steel spoke nipple- that was the point
> of the post.
Although rim eyelets are know to fail through fatigue, spoke breakages
from fatigue are much more common. Often, there are other factors that
effect rim eyelet failure, such as a brittle hard anodize surface
treatment.
> Perhaps you should have also pointed out that high strength steel (the
> 145ksi/ 1000MPa strength) that spokes are made from does not have an
> 'endurance limit' either. A genuine 'endurance limit' is only found
> in mild steels.
While it may not be an absolutely perfect one, for all practical
purposes, the high strength stainless steels used in spokes _do_ have an
endurance limit.
> >The rim failure from excess spoke tension is a yield failure, not a
> >fatigue failure, and is related to the yield strength of the rim
> >material (which is usually, but not always, aluminum) and it's cross
> >sectional area.
> >
> Fair enough, but while I don't doubt you could get a rim to fail this
> way, could it ever happen while riding.
Yes, rims have been known to collapse during hard landings when jumping,
and rims on heavily bicycles have been known to collapse during hard
breaking.
> Is the fatigue life of the
> rim not more important in a real riding situation?
Generally not. Except for a few instances of spoke eyelet failure, most
rims fail either due to braking abrasion wearing the sidewalls down to
thin, or from bending from sharp impacts.
Also, there are often cases in which a too flexible rim with too few
spokes (or spokes that are too stiff) results in a high frequency of
spoke breakage, even though the rim itself remains intact.
Mark McMaster
MMc...@ix.netcom.com
Mark McMaster
>
> On 14 Aug 1998 16:12:54 GMT, jbr...@hpl.hp.com (Jobst Brandt) wrote:
>
> >I assume you mean the rim joint. I think you don't have a good
> >visualization of the loads in the rim. Unless the wheel strikes a
> >concentrated obstacle, the rim is always in compression.
> >
> >> So any welding will be under closely controlled conditions and any
> >> claimed 'rim strength' will be quoted so that 99% of the rims off
> >> the production line meet that strength. So in practice do you
> >> really need a factor of safety for this?
> >
> >I don't know what you are proposing here.
> >
> I am proposing that most rim manufacturers know what they are doing
> and it is unlikely that one rim off a production line will have a
> claimed strength significantly less than another off the same
> production line.
I think the point here that the since the rim joint is always under
compression in a built wheel, there is little danger of it coming apart
regardless of how it is joined. Welded rims have only become common
recently, mostly as a marketing gimmick. Rims joints that are simply
held in alignment by a sleeve slipped into the ends of the extrusion
have been holding up fine for decades.
> I am agreeing with you that the rim is always under compression.
> I am asking if it is even possible for the rim to be under more
> compression than static compression?
Yes, there are situations in which the compression in some parts of the
rim can be increased. See other postings in this thread.
> ><cut and pasted from another Jobst post on this thread>
> >Riding over such uneven surfaces (potholes) or dropping off a high curb
> > causes loads great enough to slacken spokes. This does not materially
> >increase tension in any spokes although it increases rim stress
> >through bending. The inside of the bend increases tension (decreases
> >compression) while the outside increases compression, adding to the
> >compression caused by spoke tension.
> >
> If there is no possibility of increasing the compression on the rim
> while riding, then why not apply maximum compression to the rim when
> building it? Is there anything wrong with building a rim this way?
>
> So Jobst is saying above that if you go over a pot hole or drop off a
> high curb the compression in the rim can go above the material
> compressive yield point. Fair enough.
> But can someone please explain how this 'adds to the compression
> caused by the spoke tension' if the spokes have gone slack?
Obviously, only the spokes in the LAZ have gone slack - the rest have
just as much tension as before, so there is still a static compression
on the rim. In impacts, the rim is under a bending load, which adds an
additional tensile load on side of the rim and an additional compressive
load on the other. This compressive load is super-imposed over the
static compression, and may exceed the yield stress. In addition, if
the spokes in the LAZ have slackened, the rim becomes laterally unstable
as well, and is easily buckled sideways.
> >
> >> Curiously I looked in the 1998 Mavic catalogue but couldn't see any
> >> claims for rim strength. Is there any technical rim strength info on
> >> the net?
> >
> >In what units do you expect to see that? At best, there could be a
> >maximum spoke tension given, possibly in total tension, because these
> >rims are offered in several drillings.
> >
> Yes, it might be useful to have information on the maximum total
> tension that a rim can take. After all, the equivalent information
> is available for spokes.
But in the case of the spokes, the numbers are given for advertising
only. The spoke strength can't be used for building the wheel, since
the rim will yield before the spoke tension strength is reached.
Mark McMaster
MMc...@ix.netcom.com
Tom Kunich
>
> Tom Kunich writes:
>
> > I want to take exception to this statement because I think that it
> > needs a lot of discussion. I don't see it at ALL as obvious that
> > rims don't COMMONLY fail at the rim joint.
>
> > 1) A tandem came into the shop a number of months (maybe 8) ago.
> > The owner is blind and stokes and he gets several different captains
> > but mostly one guy who is quite experienced with him. His rear
> > wheel was folded over at the joint at almost 90 degrees. He said
> > that they simply came to a stop and while they were sitting at the
> > stop sign the bike rim collapsed.
>
> > 2) It happened to the same guy a second time. Possibly a third!
>
> > 3) There is a picture in Wheelsmith upstairs that shows three
> > track guys sprinting and one of them has a wheel that is in
> > the process of folding over right at the rim joint.
>
> I don't believe a word of "sitting at the stop sign the bike rim
> collapsed". I understand enough about wheel so know that a wheel that
> can be ridden for a reasonable distance, accelerating and stopping,
> cannot suddenly collapse while standing there. That you pass such
> hearsay on as evidence od anything is like forwarding a chain letter.
> You keep a false notion alive.
Whether you are willing to accept it or not the two people that were
on the tandem reported the same story. While at a stop the rear
wheel just slumped over, bending at the joint. The rim was a standard
36 hole Mavic if memory serves.
After about the THIRD case they replaced the rear wheel with a 48
hole Rino which is heavy enough to withstand a tank.
These guys would gain absolutely nothing by lying about it. So either
they made a mistake (three of them) or the wheel actually was overloaded
and bent at the joint as they claimed.
Of what use is it to call them liars or insinuate that I am saying
something that wasn't reported in good faith? Do you feel so threatened
by information that you must stomp it out before anyone else see it?
> The Wheelsmith picture is a classic example of what I mentioned in the
> response above. This is a destructive crash that folded the wheel.
> At such moments there is no tension in the wheel and if you are lucky,
> the joint will be in the bending zone where it will separate. If not
> the rim will taco and the joint remains closed.
The photo is just a photo as far as I can see. It does NOT appeat to
have been a crash. The man with the bending rim is clear enough of
the other two that it doesn't seem likely that he was involved with
either of them.
You might have information from outside of the limits of that
photograph and know that there was some sort of crash. If so
why not simply explain it instead of implying that it is plain
that there was a crash?
Snoopy
<MMc...@ix.netcom.com> wrote:
>
>Although rim eyelets are know to fail through fatigue, spoke breakages
>from fatigue are much more common. Often, there are other factors that
>effect rim eyelet failure, such as a brittle hard anodize surface
>treatment.
>
>> rim not more important in a real riding situation?
>
>Generally not. Except for a few instances of spoke eyelet failure, most
>rims fail either due to braking abrasion wearing the sidewalls down to
>thin, or from bending from sharp impacts.
>
a/ If a spoke fails, it is most likely to be a fatigue failure.
b/ If a rim fails (ignoring braking abrasion) it will normally be due
to bending from sharp impacts.
c/ Rim failure is far more likely than spoke failure.
But really what we are discussing here is a/, because in rim failure
it usually doesn't matter what the spoking system is? SNOOPY
Snoopy
wrote:
>>
>> I don't believe a word of "sitting at the stop sign the bike rim
>> collapsed". I understand enough about wheel so know that a wheel that
>> can be ridden for a reasonable distance, accelerating and stopping,
>> cannot suddenly collapse while standing there.
>
>on the tandem reported the same story. While at a stop the rear
>wheel just slumped over, bending at the joint. The rim was a standard
>36 hole Mavic if memory serves.
>
indeed suffered riding induced damage, and the bike happened to park
such that the weak point on the rim was subject to maximum static
stress, which simply finished off the dynamic failure that was about
to happen. SNOOPY
Snoopy
Denney) wrote:
>On Fri, 14 Aug 1998 06:24:54 GMT, tenn...@caverock.net.nz (Snoopy)
>wrote:
>
>>
>>Are you saying you cannot suffer fatigue failure in a wheel that
>>behaves in a linearly elastic way? If so, I believe you are wrong.
>
>Spokes that fail from fatigue had some portion of the spoke that was
>yielded as a result internal stresses induced during the formation of
>the spoke. The fatigue crack initiates in these yielded areas, and
>then the point of the crack is such a stress riser that it yields, and
>so on, until the crack propagates far enough such that the remaining
>material cannot hold the tension of the spoke. In spokes that break,
>the portion that ultimately ruptures is surprisingly small compared to
>the portion that cracks from fatigue.
>
Good explanation of the way it works. The only thing I see missing
is that the 'initiating fatigue crack' does NOT need the spoke
stressed to it's yield point to get the crack started.
>
>Remember, however, that the static load in a spoke causes the greatest
>stress the spoke normally sees, and the dynamic load reduces the
>stress. The dynamic load is a compressive load that is superposed on a
>static tensile load. If the static load is below the fatigue limit for
>the spoke, then dynamic loads should have no effect. Spokes that do
>not have internal stresses above the fatigue limit do not fail in
>practice.
>
Well I can follow that, but given that the general consensus in this
discussion seems to me to be that when spokes do fail they do so in
fatigue, how likely is the above scenario? SNOOPY
Dave Blake
>What I was suggesting was that the rim breaking first might suggest
>that the endurance strength of the aluminium rim eyelet is less than
>the endurance strength of the steel spoke nipple- that was the point
>of the post.
The spoke nipple is either brass or aluminum unless you
made it yourself.
DB
--
---
Dave Blake
dbl...@phy.ucsf.edu
Henri P. Gavin
dozens of wheel designs as well as measurements of spoke stresses
during actual road-riding conditions, a statistical evaluation of
fatigue life, and fatigue data from actual spokes, please refer to:
Gavin, Henri P., ``Bicycle Wheel Spoke Patterns and Spoke Fatigue,''
ASCE J. of Engineering Mechanics, vol. 122, no. 8 (1996) pp. 736-742.
Jobst Brandt
That's an interesting title but spoke fatigue is not affected by the
spoke pattern, unless it involves bending spokes around one another
such as with a snowflake pattern.
Jobst Brandt <jbr...@hpl.hp.com>
Mark McMaster
>
> On Sun, 16 Aug 1998 21:50:06 -0700, Mark McMaster
> <MMc...@ix.netcom.com> wrote:
> >
> >Although rim eyelets are know to fail through fatigue, spoke breakages
> >from fatigue are much more common. Often, there are other factors that
> >effect rim eyelet failure, such as a brittle hard anodize surface
> >treatment.
> >
> >> Is the fatigue life of the
> >> rim not more important in a real riding situation?
> >
> >Generally not. Except for a few instances of spoke eyelet failure, most
> >rims fail either due to braking abrasion wearing the sidewalls down to
> >thin, or from bending from sharp impacts.
> >
> So your view is that:
> a/ If a spoke fails, it is most likely to be a fatigue failure.
Almost exclusively. On the rare occasion, something going into the
wheel and actually getting caught in the spokes can break them. But in
the absence of hitting the spokes directly, even in impacts which
severely bend or break the rim the nipples will be yanked through the
spoke holes (or the eyelets pulled out of the rim) before the spoke
breaks.
> b/ If a rim fails (ignoring braking abrasion) it will normally be due
> to bending from sharp impacts.
Most often, although as mentioned before, spoke holes can sometimes
crack, especially on hard anodized rims. Using spokes too thick (stiff)
for the rim can also help instigate spoke hole cracks.
> c/ Rim failure is far more likely than spoke failure.
In properly built wheels, yes.
> But really what we are discussing here is a/, because in rim failure
> it usually doesn't matter what the spoking system is?
See response to b/.
Mark McMaster
MMc...@ix.netcom.com
Snoopy
Well I now have that paper in front of me (good stuff Henri!)
For your information Jobst, there is a sentence in the introductory
paragraph which reads:
"Spoke strains due to radial loads are insensitive to spoke pattern".,
so it looks like Henri agrees with you.
For those who can't get this paper, here is a VERY brief summary, with
apologies for Henri Gavin for the obvious gross simplification .
***************************************************************************
Touring bike rear wheel is fitted with a strain gauge system to
measure strain of spokes in a road riding situation.
Wheel is 36 spoke, 309.5mm radius (is that a700C rim?), tyres at
700kpa (100psi), instruments indicate a LAZ of 40degrees. Total
weight of bike and rider was 930N i.e. roughly 93kg.
The largest variation in spoke stress was found to be 150MPa (22ksi),
going over a suspected (but not identified) pothole.
The experiment was then transferred to the lab, where different wheels
were subjected to fatigue variations of 175Mpa, 300MPa, 380Mpa and
460Mpa (i.e. all much higher than you would get on the road).
The spoke patterns tested were 2X, 3X and 4X and also tested were
different spoke diameters 1.6mm, 1.8mm and 2.0mm. No mention was
made of them being butted, so I assume that means they were straight.
(Is that correct Henri?)
Conclusion was:
"The fatigue resistance of spoked wheels to steady cycling loads is
high for most typical service conditions"
**********************************************************************
Also mentioned are wheel tests commissioned by Wheelsmith Inc at
Standford University in 1984-85. This has relevance to my assertion
which I cut and paste from another part of this thread below.
>>Static loads (mean stresses) can actually reduce fatigue life,
> I do agree with your statement immediately above, that
>static loads (from the point of view of the wheel, that is the tension
>in the spokes), does affect the fatigue life. I think it was fatigue
>test results collated by Soderberg which first documented this
>phenomenom. Soderberg showed that test piece pre-loading can affect
>the size of a dynamic load that a test piece can stand.
>To paraphrase Soderberg's results in 'spoke' terms:
>If a highly tensioned spoke (high static load), will only be able to
>withstand a small rider (low dynamic load), because of a projected low
>fatigue life
>THEN if the overall wheel spoke tension is reduced (lower static
>load), the the fatigue life of the spokes will increase, which will
>suit a larger rider (higher dynamic load).
> Reducing overall wheel tension to
>minimize 'road riding' wheel fatigue failure is consistant with
>Soderbergs experiments.
In the Wheelsmith tests, cycle wheels were tested with different
pretensions, namely 174MPa, 250MPa, 343MPa, 347Mpa, 424MPa and 501MPa.
The result was NO significant difference in fatigue life, so it would
appear that my own thought experiment above is wrong.
However, given that they went to the trouble of setting up the
experiment, I am guessing that the results that Wheelsmith got were
unexpected. Any comments Henri? SNOOPY
Jobst Brandt
>>> For a published discussion featuring extensive computer analysis of
>>> dozens of wheel designs as well as measurements of spoke stresses
>>> during actual road-riding conditions, a statistical evaluation of
>>> fatigue life, and fatigue data from actual spokes, please refer to:
>>> Gavin, Henri P., ``Bicycle Wheel Spoke Patterns and Spoke Fatigue,''
>>> ASCE J. of Engineering Mechanics, vol. 122, no. 8 (1996) pp. 736-742.
>> That's an interesting title but spoke fatigue is not affected by the
>> spoke pattern, unless it involves bending spokes around one another
>> such as with a snowflake pattern.
> Well I now have that paper in front of me (g!)
> For your information Jobst, there is a sentence in the introductory
> paragraph which reads:
> "Spoke strains due to radial loads are insensitive to spoke pattern".,
> so it looks like Henri agrees with you.
That doesn't excuse the title of his paper that is fodder for those
who purvey wacko spoke patterns, most people, reading no farther than
the title or having no access to the paper.
> For those who can't get this paper, here is a VERY brief summary, with
> apologies for Henri Gavin for the obvious gross simplification .
> **********************************************************************
> Touring bike rear wheel is fitted with a strain gauge system to
> measure strain of spokes in a road riding situation. Wheel is 36
> spoke, 309.5mm radius (is that a700C rim?), tyres at 700kpa
> (100psi), instruments indicate a LAZ of 40degrees. Total weight of
> bike and rider was 930N i.e. roughly 93kg.
> The largest variation in spoke stress was found to be 150MPa (22ksi),
> going over a suspected (but not identified) pothole.
This is highly misleading because it gives a stress in a spoke
implying that there was an increase in tension. Unless the wheel was
damaged by the "pothole", there would only be a reduction in tension
and a reduction in stress. What was measured is unclear here but it
definitely gives the wrong result.
> The experiment was then transferred to the lab, where different wheels
> were subjected to fatigue variations of 175Mpa, 300MPa, 380Mpa and
> 460Mpa (i.e. all much higher than you would get on the road).
> The spoke patterns tested were 2X, 3X and 4X and also tested were
> different spoke diameters 1.6mm, 1.8mm and 2.0mm. No mention was
> made of them being butted, so I assume that means they were straight.
> (Is that correct Henri?)
Whether the spokes were straight or swaged has no bearing on the loads
measured. All this testing different spoke patterns is like tensile
testing spokes painted different colors. It should be blatantly
obvious to a mechanical engineer that spoke patterns have no effect on
spoke stress for rolling loads. Of course there is prior literature
that analyses this in great detail that could have been reviewed.
> Conclusion was:
> "The fatigue resistance of spoked wheels to steady cycling loads is
> high for most typical service conditions"
What means this? I think you could replace the word "high" with "low"
and not change the accuracy of that statement because it doesn't
qualify it findings in relation to any other conditions.
> ********************************************************************
> Also mentioned are wheel tests commissioned by Wheelsmith Inc at
> Stanford University in 1984-85. This has relevance to my assertion
> which I cut and paste from another part of this thread below.
>>> Static loads (mean stresses) can actually reduce fatigue life,
>> I do agree with your statement immediately above, that static loads
>> (from the point of view of the wheel, that is the tension in the
>> spokes), does affect the fatigue life. I think it was fatigue test
>> results collated by Soderberg which first documented this
>> phenomenon. Soderberg showed that test piece pre-loading can
>> affect the size of a dynamic load that a test piece can stand.
This is true for certain materials that do not withstand stress
reversal in which the structural element repeatedly changes between
tension and compression. Taken out of context as above, this is also
highly misleading. If you look at the Soderberg diagram in "the
Bicycle Wheel" you will see they preloaded in either direction, the
closer the static stress approaches the yield point the smaller the
cyclic load can be maintained without early fatigue failure. The
point to this is that a spoke must be stress relieved, not that it
should be tensioned more or less because for most reasonable wheels,
tension does not bring it near the yield point, residual manufacturing
and lacing stresses do that.
>> To paraphrase Soderberg's results in 'spoke' terms: If a highly
>> tensioned spoke (high static load), will only be able to withstand
>> a small rider (low dynamic load), because of a projected low
>> fatigue life
>> THEN if the overall wheel spoke tension is reduced (lower static
>> load), the the fatigue life of the spokes will increase, which will
>> suit a larger rider (higher dynamic load).
>> Reducing overall wheel tension to minimize 'road riding' wheel
>> fatigue failure is consistent with Soderberg's experiments.
These are all misinterpretations that don't shouldn't be repeated.
> In the Wheelsmith tests, cycle wheels were tested with different
> pretensions, namely 174MPa, 250MPa, 343MPa, 347Mpa, 424MPa and
> 501MPa. The result was NO significant difference in fatigue life,
> so it would appear that my own thought experiment above is wrong.
What means "no significant difference"? I propose that there was NO
difference.
> However, given that they went to the trouble of setting up the
> experiment, I am guessing that the results that Wheelsmith got were
> unexpected. Any comments Henri? SNOOPY
The experiment was not set up with enough understanding of the
mechanics involved to arrive at reasonable results. Before performing
such tests, a theoretical model of the hardware must be analyzed to
assess what effects are at work. I don't believe this was done nor
was the process understood. That Wheelsmith undertook this project
makes it clear that they didn't understand the work I had done and
written about in "the Bicycle Wheel". They still do not understand
stress relieving and call it pre-stressing.
Jobst Brandt <jbr...@hpl.hp.com>