Why a worn chain ruins sprockets
jobst....@stanfordalumni.org
looking at mechanics of it might resolve some some of the problems.
Sprockets have a pitch diameter that exactly meshes with a half inch
pitch chain in which the gaps between teeth are larger than the chain
rollers to make noise free running when used with a worn (elongated
chain) and to allow the chain to exit at an angle, as with crossing
from a right sprocket to the left chainwheel. Cross-over from one
extreme to the other accelerates chain and sprocket wear. The chain
to enter rear sprockets at an angle is taken care of by the derailleur
which accepts incoming chain at angles on its plastic idlers. Large
chain angles cause more rapid wear here.
As a chain wears, its pitch elongates, the length increase occurring
between inner links, the ones that carry two rollers between side
plates and remain at a half inch regardless of wear. It is the pins
and sleeves that wear and cause pitch elongation between inner links.
The chain DOES NOT STRETCH from tension, but rather gets longer as its
pivots wear.
With increasing pitch (spacing between roller centers) the chain
describes a larger pitch diameter as it rides higher on sprocket teeth
than when new. Besides riding higher, so that you can see light under
the chain when it is pulling, it is loading every second tooth because
only every second roller pair is out of pitch.
Pockets are worn in the sprocket from chain contact, not at the root
diameter for half inch pitch chain, but at a higher pitch diameter.
This does not come to bear immediately because the case hardening of
the sprocket wears slowly, but once that is gone, these pockets that
match the worn chain and lie on a larger pitch diameter develop
rapidly.
When a new chain with correct pitch is used, it idles nicely until it
transmits force against the worn teeth. At that time it rolls up the
last tooth of engagement, there where the chain departs and travels
straight to the chainwheel, and seats in the worn pocket that is at a
larger pitch diameter. Tooth pockets are often not readily
discernible because the sprocket form is circular and deviation from
that circle is hard to detect.
A new, in-pitch chain, rides high on the exit cog of the sprocket so
that the incoming roller of the chain cannot enter because it is ahead
of the cog that it should engage. It remains not engaged, riding on
top of that tooth, which makes certain that subsequent teeth cannot
begin to enter from the exaggerated pitch diameter of riding over the
tops of teeth.
Under high chain load, even small wear pockets made by an out-of-pitch
chain will prevent engagement. Regularly replacing the chain can
prevent this because the wear pockets will be at the correct pitch
diameter, but this can become expensive. The alternative is to
replace the chain at 1/16" per foot (which is in effect 1/8" since
only every second link wears) and replace only those sprockets that
won't mesh.
In earlier times, when sprockets were separate and symmetrical, one
could turn a worn sprocket around and use its unworn flank and get
twice the mileage. With teeth shaved and ramped to make shifting
under load possible, most sprockets don't work well when reversed.
Also multi sprocket units cannot be reversed.
Worn chains do not significantly affect shifting or engagement with
chainwheels, chainwheels having forced engagement and don't care
whether the chain rides high or out of engagement on exit. One thing
that occurs with worn chainwheels is that a new chain can cause "chain
suck" because the chain rides in hooked chainwheel teeth that do not
want to release. This is especially so with small chainwheels.
Loss of exact and quick shifting can occur because a strongly worn
chain can bend sideways and defy being pushed to an adjacent sprocket
by the derailleur, but this requires substantial wear and too long a
path between derailleur and sprocket cluster.
Jobst Brandt
Brian Huntley
least needs a tighter edit. It's far too run-on for on-line reading.
But while that may take time, can we get a quick fix for:
"The chain to enter rear sprockets at an angle is taken care of by the
derailleur
which accepts incoming chain at angles on its plastic idlers"
Should that be "The chain *flex required* to enter" or *displacement*
perhaps?
"The chain, the rear sprockets at an angle to enter, the incoming at
angles accepting chain derailleur plastic idlers uses."
Hmm - ought to be able to squeeze out a surpurfluous "chain" or "angle"
there.
:0)
Nate Knutson
jobst....@stanfordalumni.org wrote:
> Worn chains do not significantly affect shifting or engagement with
> chainwheels, chainwheels having forced engagement and don't care
> whether the chain rides high or out of engagement on exit.
I thought that worn chains accelerate wear on chainrings; you are
saying that a worn chain will ride high on the chainring teeth, but
engagement will occur with the same number of roller/tooth interfaces
as if the chain were new, correct?
Worn chainrings do accelerate wear to new chains by limiting the amount
of engagement, right?
jobst....@stanfordalumni.org
> Hmm - ought to be able to squeeze out a superfluous "chain" or
> "angle" there.
Thanks, it was getting late and the last cut and paste didn't work well.
Jobst Brandt
jobst....@stanfordalumni.org
>> Worn chains do not significantly affect shifting or engagement with
>> chainwheels, chainwheels having forced engagement and don't care
>> whether the chain rides high or out of engagement on exit.
> I thought that worn chains accelerate wear on chainrings; you are
> saying that a worn chain will ride high on the chainring teeth, but
> engagement will occur with the same number of roller/tooth
> interfaces as if the chain were new, correct?
Chain engagement is not affected by a worn chain coming into a new or
used chainwheel although it affects chainwheel wear.
> Worn chainrings do accelerate wear to new chains by limiting the amount
> of engagement, right?
No. chain wear is entirely dependent on articulation angle and grime.
Chains go though an angle of about 7 degrees entering a 50t chainwheel
under load while on a 13t sprocket the angle is 27.7 degrees. That's
where the ware occurs.
Jobst Brandt
Zog the Undeniable
> Worn chainrings do accelerate wear to new chains by limiting the amount
> of engagement, right?
>
An alu chainwheel isn't going to do much to a hardened steel chain.
Martin Borsje
No, but the sand between will....
--
Posted by news://news.nb.nu
Michael Nitabach
news:450cc400$0$34551$742e...@news.sonic.net:
> Effects of a worm chain on sprockets comes up often and I think
> looking at mechanics of it might resolve some some of the
> problems.
<snip>
How would you apply this analysis to a fixed-wheel?
I recently changed a worn chain on my fixed-wheel, without changing the
cog. With the new-chain/worn-cog combo, whenever I put a lot of force
on the pedals, a "sproinging" sound emanated from what I presume to be
the chain/cog interface. This was a powerful incentive to change the
cog, after which the sound went away.
What do you think was the mechanical origin of the "sproinging" sound?
--
Mike Nitabach
jobst....@stanfordalumni.org
>> Effects of a worm chain on sprockets comes up often and I think
>> looking at mechanics of it might resolve some some of the problems.
> <snip>
> How would you apply this analysis to a fixed-wheel?
> I recently changed a worn chain on my fixed-wheel, without changing
> the cog. With the new-chain/worn-cog combo, whenever I put a lot of
> force on the pedals, a "sproinging" sound emanated from what I
> presume to be the chain/cog interface. This was a powerful incentive
> to change the cog, after which the sound went away.
> What do you think was the mechanical origin of the "sproinging"
> sound?
Forced engagement at the rear sprocket, that effect that prevents a
spring loaded (derailleur) chain from engaging.
Jobst Brandt
jobst....@stanfordalumni.org
problems.
New sprockets have a pitch diameter that exactly meshes with a half
inch pitch chain, the chain commonly used on most bicycles. Gaps
between sprocket teeth are larger than chain rollers (0.3055" dia.) in
order to facilitate noise free running when used with a worn
(elongated chain) and to allow the chain to exit at an angle, when its
chain line makes a diagonal from sprocket to chainwheel.
Entering rear sprockets the chain is not under load and goes straight
in because, after making two U-turns around the idler wheels that are
aligned with the sprocket. Derailleur idler wheels are shaped to
accept incoming chain at angles, however, large chain angles increase
wear of these wheels (sprockets) as well as sprockets and chainwheels
because they load only one edge.
Chain wear is mainly caused by articulation angle and grime when
loaded in tension. For instance, chains go though an angle of about 7
degrees entering a 50t chainwheel while exiting a 13t sprocket the
angle is 27.7 degrees. That's where the ware occurs. The U-turns
made through two idler wheels of a derailleur, although having a
fairly large reversing angle (10t = as much as 36 degrees), probably
have relatively little effect because they occur under light load.
As a chain wears, its pitch elongates, as the link pins and the
sleeves in which they run wear. This occurs between pairs of inner
links, whose two rollers remain at a half inch spacing regardless of
wear. Pitch elongation occurs from link pin and sleeve wear between
these links. The chain DOES NOT STRETCH from tension, but rather
becomes longer as its pivots wear.
With increasing pitch (spacing between roller pair centers) the chain
describes a larger pitch diameter, as it rides higher on sprocket
teeth, than it would when new. Besides riding higher (you can see
light under the chain when it is pulling), it loads every second tooth
because only every second roller pair is out of pitch.
Pockets formed in the sprocket from chain contact, form not at the
root diameter for half inch pitch chain, but at a higher diameter.
This does not come to bear immediately because case hardening of the
sprocket wears slowly, but once that is gone, pockets made by the worn
chain develop rapidly and on a larger pitch diameter. This in
contrast to a new chain that loads several teeth of a sprocket with a
distributed load.
When a new chain with correct pitch is used, it meshes nicely until it
transmits force against the worn teeth. At that time it rolls up the
teeth to seat in the wear pocket at the larger pitch diameter of the
sprocket so that the incoming roller of the chain cannot enter because
it is ahead of the cog that it should engage. It remains disengaged,
riding on top of that tooth, which makes certain that subsequent teeth
cannot enter from the exaggerated pitch diameter of the rollers riding
over the tops of teeth.
Under high chain load, even small wear pockets made by an out-of-pitch
chain will prevent engagement. Regularly replacing the chain can
prevent this because the wear pockets will occur at the correct pitch
diameter, but this can become expensive. The alternative is to
replace the chain at 1/16" per foot (which is in effect 1/8" since
only every second link wears) and replace only those sprockets that
won't mesh.
In earlier times, when sprockets were separate and symmetrical, one
could turn a worn sprocket around and use its unworn flank to get
twice the mileage. With teeth shaved and ramped to make shifting
under load possible, most sprockets don't work well when reversed.
Also multi sprocket units cannot be reversed.
Worn chains do not significantly affect shifting or engagement with
chainwheels, chainwheels having forced engagement and don't care
whether the chain rides high or out of engagement on exit. One thing
that occurs with worn chainwheels is that a new chain can cause "chain
suck" because the chain rides in hooked chainwheel teeth that do not
want to release. This is especially so with small chainwheels.
Loss of exact and quick shifting can occur because a strongly worn
chain can bend sideways and defy being pushed to an adjacent sprocket
by the derailleur, but this requires substantial wear and a long path
Michael Nitabach
news:450dfb3a$0$34520$742e...@news.sonic.net:
Could you explain what you mean by "forced engagement", what piece
of metal is vibrating to make the sproinging sound, and what exactly
is causing that piece of metal to vibrate?
I am not an engineer, so if you could limit and/or explain the use
of engineering jargon, that would help.
--
Mike Nitabach
Johan Bornman
jobst....@stanfordalumni.org wrote:
> Effects of a worm chain on sprockets comes up often and I think
> looking at the mechanics of it might resolve some some of the
> problems.
>
>
Park and Rohloff chain wear indicators are useless and a foot-long
measurement is better?
Also JB
jobst....@stanfordalumni.org
>>>> Effects of a worm chain on sprockets comes up often and I think
>>>> looking at mechanics of it might resolve some some of the
>>>> problems.
>>> <snip>
>>> How would you apply this analysis to a fixed-wheel?
>>> I recently changed a worn chain on my fixed-wheel, without
>>> changing the cog. With the new-chain/worn-cog combo, whenever I
>>> put a lot of force on the pedals, a "sproinging" sound emanated
>>> from what I presume to be the chain/cog interface. This was a
>>> powerful incentive to change the cog, after which the sound went
>>> away.
>>> What do you think was the mechanical origin of the "sproinging"
>>> sound?
>> Forced engagement at the rear sprocket, that effect that prevents a
>> spring loaded (derailleur) chain from engaging.
> Could you explain what you mean by "forced engagement", what piece
> of metal is vibrating to make the sproinging sound, and what exactly
> is causing that piece of metal to vibrate?
Because the chain has no spring loaded slack loop as a derailleur
chain, it must engage and is forced to do so. That causes the lower
run of the chain to twang as a guitar string as it pops into
engagement. First it rides as high as it can before snapping into
engagement. With a derailleur, it never engages and skips over a
tooth to try again.
> I am not an engineer, so if you could limit and/or explain the use
> of engineering jargon, that would help.
I wasn't aware that there was any engineer jargon in it.
Jobst Brandt
jobst....@stanfordalumni.org
>> Effects of a worm chain on sprockets comes up often and I think
>> looking at the mechanics of it might resolve some some of the
>> problems.
> Nice explanation Jobst. Could you please also explain why those
> little Park and Rohloff chain wear indicators are useless and a
> foot-long measurement is better?
http://yarchive.net/bike/chain_wear.html
Jobst Brandt
Michael Nitabach
$742e...@news.sonic.net:
So under small loads, the chain just rides high and never snaps into
engagement, thus making no noise?
--
Mike Nitabach
jobst....@stanfordalumni.org
>>>>> <snip>
It's the opposite. When loose, the chain isn't bearing on the
pressure face of sprocket teeth and just falls into place freely
between them, the sprocket teeth being spaced the same as the chain
rollers. When the chain is pulled into the wear pockets of the
sprocket, it rides up the teeth so that the next tooth comes along
sooner than the next roller and blocks it from entering. These are
small dimensions that are not easily seen, but are enough to make
getting the chain to mesh a press fit. That is what you are hearing.
Jobst Brandt
Diablo Scott
> Johan Bornman writes:
>>Nice explanation Jobst. Could you please also explain why those
>>little Park and Rohloff chain wear indicators are useless and a
>>foot-long measurement is better?
> http://yarchive.net/bike/chain_wear.html
"The Rholoff gauge measures between two
rollers a several links apart. Because rollers always have clearance,
they can be displaced longitudinally on the chain."
True, but you're only adding the roller clearance on two rollers, not
every roller in the span of the measurement; it's a very small number.
"A ruler measuring over 12 inches, even non contacting,
is more accurate and easier to use than this tool. Until you've tried
the ruler method, it is not obvious how easy it is to see a 1/16 inch
wear in a foot by holding a yardstick next to the bottom run of the
chain."
The Park tool is very simple to use: you can take the measurement and
remove the tool to read it in better light, you can do it with one hand,
you can do it wihout your glasses on, you can do it while your bike's
hanging on the wall, you can take five readings in ten seconds. It
measures in "percent elongation": my new chains typically measure 0.25%
elongation or 1/32" over 12 inches, and I start having problems at about
0.75% elongation or 3/32" over 12 inches. So it may not make the
measurement you'd like it to, but it's very useful as an indicator of
when to replace a chain, especially after you've got some history with it.
carl...@comcast.net
Dear Diablo,
My aging eyes like what you say.
My experience with measuring wear variation with a micromeeter in
6-inch sections of a chain also likes what you say.
Several quick and easy to read measurements in my garage sounds nice.
Keep talking like that, you smooth-talkin' devil, and I'm going to
blame you for making me buy one.
Cheers,
Carl Fogel
jobst....@stanfordalumni.org
>>> Nice explanation Jobst. Could you please also explain why those
>>> little Park and Rohloff chain wear indicators are useless and a
>>> foot-long measurement is better?
http://yarchive.net/bike/chain_wear.html
> "The Rohloff gauge measures between two rollers, several links
> apart. Because rollers always have clearance, they can be displaced
> longitudinally on the chain."
> True, but you're only adding the roller clearance on two rollers, not
> every roller in the span of the measurement; it's a very small number.
For the short (4# section over which is is measuring, 1/3 as much
elongation than the yard stick method. the common wear criterion of
1/16" (0.0625") is therefore 0.208" when divided by 3. Roller
clearance on a typical new SRAM chain is 0.0125" taken twice since
this involves two rollers pushed apart, gives 0.025". That is already
40% of the dimension to be measured. Because roller clearance varies
for different models and brands of chain, this is an unacceptable
method. Roller clearance on some chains increases faster than pitch
elongation, the dimension of interest, so you cannot be sure of what
the chain elongation is with this instrument.
> "A ruler measuring over 12 inches, even non contacting, is more
> accurate and easier to use than this tool. Until you've tried the
> ruler method, it is not obvious how easy it is to see a 1/16 inch
> wear in a foot by holding a yardstick next to the bottom run of the
> chain."
> The Park tool is very simple to use: you can take the measurement
> and remove the tool to read it in better light, you can do it with
> one hand, you can do it without your glasses on, you can do it while
> your bike's hanging on the wall, you can take five readings in ten
> seconds.
Why take your glasses off if you need them to see a yard stick. It
seems you have not tried the yard stick method, or you have already
spent the money to buy such a gauge and feel compelled to defend it.
Besides, why would you want to take five readings?
> It measures in "percent elongation": my new chains typically measure
> 0.25% elongation or 1/32" over 12 inches, and I start having
> problems at about 0.75% elongation or 3/32" over 12 inches. So it
> may not make the measurement you'd like it to, but it's very useful
> as an indicator of when to replace a chain, especially after you've
> got some history with it.
You can also see how much air you get between chainwheel and chain in
a side view to get an "estimate" but a 1/16" per foot is unambiguous
with a hardware store give-away (free) wooden yardstick.
Jobst Brandt
jobst....@stanfordalumni.org
>>> Nice explanation Jobst. Could you please also explain why those
>>> little Park and Rohloff chain wear indicators are useless and a
>>> foot-long measurement is better?
http://yarchive.net/bike/chain_wear.html
> "The Rohloff gauge measures between two rollers, several links
> apart. Because rollers always have clearance, they can be displaced
> longitudinally on the chain."
> True, but you're only adding the roller clearance on two rollers, not
> every roller in the span of the measurement; it's a very small number.
For the short (4" section over which is is measuring, 1/3 as much
elongation occurs than in one foot with the yard stick method. The
common wear criterion of 1/16" (0.0625") is therefore 0.208" when
divided by 3. Roller clearance on a typical new SRAM chain is 0.0125"
taken twice since this involves two rollers pushed apart, gives
0.025". That is already 40% of the dimension to be measured. Because
roller clearance varies for different models and brands of chain, this
is an unacceptable method. Roller clearance on some chains increases
faster than pitch elongation, the dimension of interest, so you cannot
be sure of what the chain elongation is with this instrument.
> "A ruler measuring over 12 inches, even non contacting, is more
> accurate and easier to use than this tool. Until you've tried the
> ruler method, it is not obvious how easy it is to see a 1/16 inch
> wear in a foot by holding a yardstick next to the bottom run of the
> chain."
> The Park tool is very simple to use: you can take the measurement
> and remove the tool to read it in better light, you can do it with
> one hand, you can do it without your glasses on, you can do it while
> your bike's hanging on the wall, you can take five readings in ten
> seconds.
Why take your glasses off if you need them to see a yard stick. It
seems you have not tried the yard stick method, or you have already
spent the money to buy such a gauge and feel compelled to defend it.
Besides, why would you want to take five readings?
> It measures in "percent elongation": my new chains typically measure
> 0.25% elongation or 1/32" over 12 inches, and I start having
> problems at about 0.75% elongation or 3/32" over 12 inches. So it
> may not make the measurement you'd like it to, but it's very useful
> as an indicator of when to replace a chain, especially after you've
> got some history with it.
You can also see how much air you get between chainwheel and chain in
Johnny Sunset aka Tom Sherman
Diablo Scott wrote:
> jobst....@stanfordalumni.org wrote:
>
> > Johan Bornman writes:
>
> >>Nice explanation Jobst. Could you please also explain why those
> >>little Park and Rohloff chain wear indicators are useless and a
> >>foot-long measurement is better?
>
>
> > http://yarchive.net/bike/chain_wear.html
>
> "The Rholoff gauge measures between two
> rollers a several links apart. Because rollers always have clearance,
> they can be displaced longitudinally on the chain."
>
> True, but you're only adding the roller clearance on two rollers, not
> every roller in the span of the measurement; it's a very small number.
>
> "A ruler measuring over 12 inches, even non contacting,
> is more accurate and easier to use than this tool. Until you've tried
> the ruler method, it is not obvious how easy it is to see a 1/16 inch
> wear in a foot by holding a yardstick next to the bottom run of the
An architect's scale will also work well here.
--
Tom Sherman - Here, not there.
Diablo Scott
> Diablo Scott writes:
> Because roller clearance varies for different models and brands of chain,
> this is an unacceptable method.
The worst case is I throw away a chain that still had a few miles on it.
It's gotten a lot of use, but not quite as much as the ones in the
Brandt garbage can. In practice though, whenever I do notice chain
skipping, the Chain Checker consistently reads 0.75% - SRAM and Shimano
chains. So whenever I measure more than 0.75% I replace the chain even
if I haven't noticed skipping. It's acceptable.
jobst....@stanfordalumni.org
Chains don't skip when run on the same sprockets since the sprockets
wear in accordance with the chain pitch. Worn chains will run on new
sprockets without problem.
That you say you ride them until they skip, makes we wonder what you
are riding. The scenario you present cannot happen. Even if you have
other wheels with new sprockets, they will run fine with a worn chain.
Jobst Brandt
Michael Nitabach
news:450f0a5a$0$34553$742e...@news.sonic.net:
Got it. Thanks for the detailed explanation.
--
Mike Nitabach