Chain wear measurement

Jay Beattie writes:

Perching a ruling on a dirty chain is hardly accurate either -- I
wonder what the error rate is in the usual $.39 wood ruler. Not
that I am sold on the expensive Park tool, either -- although it is
kind of fun and looks more technical than a ruler (even my steel
ruler).

http://www.teamendorphin.com/jpg/newchain.jpg

That's a lot more accurate than pushing a chain-gauge into the chain,
as those who have sampled the process report. The measurement
involves seeing the difference of a line width of a wooden freebee
yardstick on a chain. You can readily see whether the pin at 12"
downstream is one, two, or three line widths from where it should be.
Even more accurate is laying a worn chain next to a new one and seeing
how many fractions of a link (1/2") the old chain is longer than the
new one. Neither of these methods produce a go-and-no go result which
I suppose is what bothers folks because they must make a decision and
that's difficult.

Getting back to Jim Beams comment, why does roller wear not
count, but pin and bushing wear does? The rings haning on nails
analogy does not work for me, since some rings may be worn more
than others -- and in my case, even broken. -- Jay Beattie.

How do you wear some rollers more than others? I mean wear
differences that make a meaningful difference in pitch. I think you
should take a chain apart and see if you can measure any variation in
wear on any of the parts. The continuous randomness of loading the
chain guarantees as uniform a process as one can invent. Your
suggestion implies there are certain rollers that wear faster. I
propose that this is far fetched to the extreme.

Jobst Brandt

wrote:
Jay Beattie writes:

Getting back to Jim Beams comment, why does roller wear not
count, but pin and bushing wear does? The rings haning on nails
analogy does not work for me, since some rings may be worn more
than others -- and in my case, even broken. -- Jay Beattie.

How do you wear some rollers more than others? I mean wear
differences that make a meaningful difference in pitch. I think you
should take a chain apart and see if you can measure any variation in
wear on any of the parts. The continuous randomness of loading the
chain guarantees as uniform a process as one can invent. Your
suggestion implies there are certain rollers that wear faster. I
propose that this is far fetched to the extreme.

I can't resist playing devil's advocate and pointing out that the OP is
reporting exactly that - different readings from different parts a worn
chain. Have fun :-)

wrote in message
...
Jay Beattie writes:

Perching a ruling on a dirty chain is hardly accurate
either -- I
wonder what the error rate is in the usual $.39 wood ruler.
Not
that I am sold on the expensive Park tool, either -- although
it is
kind of fun and looks more technical than a ruler (even my
steel
ruler).

http://www.teamendorphin.com/jpg/newchain.jpg

That's a lot more accurate than pushing a chain-gauge into the
chain,
as those who have sampled the process report. The measurement
involves seeing the difference of a line width of a wooden
freebee
yardstick on a chain. You can readily see whether the pin at
12"
downstream is one, two, or three line widths from where it
should be.
Even more accurate is laying a worn chain next to a new one and
seeing
how many fractions of a link (1/2") the old chain is longer
than the
new one. Neither of these methods produce a go-and-no go
result which
I suppose is what bothers folks because they must make a
decision and
that's difficult.


Getting back to Jim Beams comment, why does roller wear not
count, but pin and bushing wear does? The rings haning on
nails
analogy does not work for me, since some rings may be worn
more
than others -- and in my case, even broken. -- Jay Beattie.

How do you wear some rollers more than others? I mean wear
differences that make a meaningful difference in pitch. I
think you
should take a chain apart and see if you can measure any
variation in
wear on any of the parts. The continuous randomness of loading
the
chain guarantees as uniform a process as one can invent. Your
suggestion implies there are certain rollers that wear faster.
I
propose that this is far fetched to the extreme.

Hardly far fetched. I smashed several rollers on the chain on my
touring bike, and broke at least one. Obviously, this problem was
not subtle and did not require a tool or even a ruler to detect,
but it suggests to me that rollers can wear differentially,
especially if wear means deformation.

Anyway, since the rollers actually contact the sprockets,
wouldn't roller wear change the "effective" pitch of the chain
and have some effect on the alignment of the chain and sprockets.
If this is so, wouldn't a measurement that accounts for roller
wear be helpful in determining remaining chain life? Even if all
roller wear is equal, doesn't it come into the chain life
equation somewhere? -- Jay Beattie.

wrote:
Even more accurate is laying a worn chain next to a new one and seeing
how many fractions of a link (1/2") the old chain is longer than the
new one.

I have a friend who has a new cheap chain hanging on a finishing nail
in the garage. When he cleans the chain, he puts it on the finishing
nail on top of the new chain. Easy to see that the old chain has
"stretched". Very dramatic when you look at the total "stretch" over a
4.5 foot chain. New chains are cheaper than chain checking tools.
$7.99 for a cheap chain at Nashbar compared to $9.95 for the Park chain
checking tool. And of course the truely frugal could just have his
replacement chain hanging in the garage and use it to check for chain
wear. And then install it when the old one is "stretched". Then get a
new replacement chain and repeat. No need to squander money on a $0.39
wood ruler.

On Mon, 15 Aug 2005 15:08:11 +0100, "Pete Biggs"
wrote:

[snip]

Park CC-2 can hardly be inserted into a new Campagnolo C9 chain to get a
zero reading but reads 0.25% for a new SRAM PC59. Does the average tech
realise that? Do you call that consistent?

~PB

Dear Pete,

It sounds as if the tool is consistent, while the two brands
of chain are surprisingly different in how they would fit on
a sprocket.

http://www.parktool.com/tools/CC_2BIG.shtml

Do these chains fit sprockets very differently? Are their
rollers different sizes?

Curiously,

Carl Fogel

wrote:
On Mon, 15 Aug 2005 15:08:11 +0100, "Pete Biggs"
wrote:
Park CC-2 can hardly be inserted into a new Campagnolo C9 chain to
get a zero reading but reads 0.25% for a new SRAM PC59. Does the
average tech realise that? Do you call that consistent?

Dear Pete,

It sounds as if the tool is consistent, while the two brands
of chain are surprisingly different in how they would fit on
a sprocket.

http://www.parktool.com/tools/CC_2BIG.shtml

Do these chains fit sprockets very differently? Are their
rollers different sizes?

Dear Carl, Just measured, both rollers look like 7.8mm to me. Perhaps
SRAM chains are supplied pre-stretched! ;-)

I did once fit a new SRAM chain that worked with worn sprockets on which a
new Campagnolo chain skipped.

~PB

wrote:

I have a friend who has a new cheap chain hanging on a finishing nail
in the garage. When he cleans the chain, he puts it on the finishing
nail on top of the new chain. Easy to see that the old chain has
"stretched". Very dramatic when you look at the total "stretch" over a
4.5 foot chain. New chains are cheaper than chain checking tools.
$7.99 for a cheap chain at Nashbar compared to $9.95 for the Park chain
checking tool. And of course the truely frugal could just have his
replacement chain hanging in the garage and use it to check for chain
wear. And then install it when the old one is "stretched". Then get a
new replacement chain and repeat. No need to squander money on a $0.39
wood ruler.

This is one of the more interesting and useful suggestions I've seen
lately. Think it may be catching?

Jay Beattie writes:

Getting back to Jim Beams comment, why does roller wear not count,
but pin and bushing wear does? The rings haning on nails analogy
does not work for me, since some rings may be worn more than
others -- and in my case, even broken.

How do you wear some rollers more than others? I mean wear
differences that make a meaningful difference in pitch. I think
you should take a chain apart and see if you can measure any
variation in wear on any of the parts. The continuous randomness
of loading the chain guarantees as uniform a process as one can
invent. Your suggestion implies there are certain rollers that
wear faster. I propose that this is far fetched to the extreme.

Hardly far fetched. I smashed several rollers on the chain on my
touring bike, and broke at least one. Obviously, this problem was
not subtle and did not require a tool or even a ruler to detect, but
it suggests to me that rollers can wear differentially, especially
if wear means deformation.

I have a bunch of chains here and a typical SRAM chain has 0.00675"
radial clearance on its sleeve, the upset collar of the inner plate.
The chain has a 0.5" pitch that is not affected by the roller
clearance, whether new, as above, or worn even more. The rings on a
pegboard is the best example I can invent at the moment and you can
see that no matter how you rotate the peg board around the peg axis,
the ring (chain roller) spacing remains unchanged.

Anyway, since the rollers actually contact the sprockets,
wouldn't roller wear change the "effective" pitch of the chain
and have some effect on the alignment of the chain and sprockets.
If this is so, wouldn't a measurement that accounts for roller
wear be helpful in determining remaining chain life? Even if all
roller wear is equal, doesn't it come into the chain life
equation somewhere?

Not for sprocket wear and probably not for the user, but that doesn't
happen since the wear in the critical pin to sleeve (0.1425" dia) is
at least double what it is between roller and shoulder. The roller
doesn't roll much while the pin at its smaller diameter turns under
full load. That the roller doesn't turn much was apparent back in the
early days of (block) chains when there were no rollers.

Jobst Brandt

Pat Lamb wrote:
wrote:

I have a friend who has a new cheap chain hanging on a finishing nail
in the garage. When he cleans the chain, he puts it on the finishing
nail on top of the new chain. Easy to see that the old chain has
"stretched". Very dramatic when you look at the total "stretch"
over a 4.5 foot chain. New chains are cheaper than chain checking tools.
$7.99 for a cheap chain at Nashbar compared to $9.95 for the Park
chain checking tool. And of course the truely frugal could just
have his replacement chain hanging in the garage and use it to check
for chain wear. And then install it when the old one is
"stretched". Then get a new replacement chain and repeat. No need
to squander money on a $0.39 wood ruler.

This is one of the more interesting and useful suggestions I've seen
lately. Think it may be catching?

Well, if you walk by the nail while wearing a Rayon shirt...

Bill "still slow afternoon" S.

Carl Fogel writes:

Park CC-2 can hardly be inserted into a new Campagnolo C9 chain to
get a zero reading but reads 0.25% for a new SRAM PC59. Does the
average tech realise that? Do you call that consistent?

It sounds as if the tool is consistent, while the two brands of
chain are surprisingly different in how they would fit on a
sprocket.

As far as I have seen all these chains have 0.302" diameter rollers,
it's the width between side plates and overall width that is tuned to
how close the sprockets are in the gear cluster.

http://www.parktool.com/tools/CC_2BIG.shtml

Do these chains fit sprockets very differently? Are their rollers
different sizes?

All the ones I measured have 0.302" diameter rollers. That includes
ancient ones and current chains.

Jobst Brandt