measuring chain wear (again)

[NATURAL ANKLING]

Hi,
On a new chain it can have negative stretch, whether this is by design or the gaps are full of lube
You will need to tension chain some to get an accurate measure.
Cheap chains I have demolished to over 1 % in 300 miles, I am lucky to get to 1000 before I take it off.
A decent chain which you havent stressed to much will probably fair well
Especially if your chain wheel is large

The chain I have on the skip trainer at the mo is a shimano hg 50 which I think I took of my MTB after about 1000, but its still below 0.75 % on the skip trainer after several hundred miles.
The MTB had a non compact chainwheel (46) and on this wheel the chain rides up (visually) but is fine on the compact skip trainer (34 - 42).
Max out of a very good second hand chain is about 600 miles (skip trainer, 20 % off road) on the cheapest KMC that is shipped with throwaway bikes.

[NATURAL ANKLING]

Hi,
I use this, it saves all that farting around with rules.
1% on this and I reckon the chain is stuffed.
Cheap at £ 2.39.......
http://www.ebay.co.uk/itm/SILVERLINE-BI ... 2073wt_952

[beardy]

I have always found that the best way of measuring chain wear ("stretch") is to hang it through the open link pin holes from a suitable nail. This is easily done now that we have quick release links. Any stretch above half a link from its length as new (about 0.4% - 0.5%) should put it in line for changing. There is no need to be meticulously accurate. I have a nail and mark at 110 links on my garage door post. Chains are cheap expendable items and changing them regularly saves more than their cost in the life of rings and sprockets.

Yes, Luddites of the world unite.

A pin-nail near the top of the door and two pencil marks near the bottom.

[mercalia]

At the moment I dont have much confidence in any of these things. The digital vernier - unless the internal probes are very thin getting them in one link is prone to error as they may catch the upper suface of the roller to an unknowable degree so when zeroed give a low stretch value? My vernier may fall into that group. I thought a way around this is to measure 4 links and set to zero and then measure 5 to get what 1 link stretch is? My inner probes will fit in the links thus. But even then it is possible to catch the chain on differnt parts of the probes if the guage not kept parallel to the chain and get differnt reading ?But I still get rather incredible low stretch. I still dont know how any one can do the ruler thing unless they have it all nailed to a wall.And then I wonder if chains really obey the 1" rule especially cheap ones

[Brucey]

I'm not sure what you mean when you say that the verniers 'catch the upper surface of the roller to an unknowable degree'... the vernier jaws sit at a tangent to the roller; they can't do much else; you can see where they sit so you know what you are measuring. If you can't see that very well maybe you need a magnifier or better reading glasses or something...?

BTW trying to measure one link is not a good idea. To get a meaningful measurement of chain stretch, measuring multiple links in one go is the best scheme, as per Bob's original suggestion.

cheers

[mercalia]

I'm not sure what you mean when you say that the verniers 'catch the upper surface of the roller to an unknowable degree'... the vernier jaws sit at a tangent to the roller; they can't do much else; you can see where they sit so you know what you are measuring. If you can't see that very well maybe you need a magnifier or better reading glasses or something...?

BTW trying to measure one link is not a good idea. To get a meaningful measurement of chain stretch, measuring multiple links in one go is the best scheme, as per Bob's original suggestion.

cheers

well my inner jaws dont fit very well in between 2 rollers to do the zeroing - not sure where the surfaces are touching - it could be just the tips as the jaws dont go all the way thru.

well i did this calc

5x + a = measurement1 5 links
4x +a = measurement2 4 links
"a" is the between link that would normally be zeroed out?
subtract and the "a" is removed & u get x = measurement 1- measurement 2
if you zero the 4 link measurement u get what 1 link is when u do the 5 link measurement? but not by measureing 1 link. And the jaws of my vernier will go all the way thru the chain links ( just not together)

[Brucey]

you only need to measure using the vernier jaw tips if that is all that will go in. That is why the jaws are shaped like that, and ground to be parallel.

If you are trying to maximise your errors, taking the difference between two (large) measurements will do it pretty well.

Use Bob's method.

cheers

[mercalia]

you only need to measure using the vernier jaw tips if that is all that will go in. That is why the jaws are shaped like that, and ground to be parallel.

If you are trying to maximise your errors, taking the difference between two (large) measurements will do it pretty well.

Use Bob's method.

cheers

not sure if I get the matter re errors unless these verniers have a percentage error which would affect small differences a great deal. I thought they claimed an accuracy/error what in effect would be to certain decimal places not significant figures?

[Brucey]

if I understand you correctly, what you are doing is fundamentally a bad idea; no scientist or engineer worth their salt would choose this method if there was an alternative. Measuring one link at a time for wear doesn't necessarily tell you much in relation to the rest of the chain, either.

It is like measuring a mile by going ten miles out and nine miles back again, if you like. And then using that 'mile' to infer an exact distance from London to Edinburgh.

cheers

[Mick F]

Mr Tullio Campagnolo and his mates suggest this:

[Brucey]

now that is like measuring the distance between Watford and Musselburgh, and then assuming that you somehow know the distance between London and Edinburgh.

Even so it may be meaningful for campag chains, but not others I'd expect.

cheers

[Mick F]

Brand new chain?

Not saying it's a good system, but if you were to use a digi caliper and measure as per Mr Campag's system, would your brand new chain pass the test?

If you can, please check as per the above.

[Brucey]

if it is a campag one, yes, but otherwise no, not necessarily. The clearances on the roller bushings vary with chain design, and the campag measurement contains 1x roller bushing slack.

cheers

[MikeF]

Hi,
I use this, it saves all that farting around with rules.
1% on this and I reckon the chain is stuffed.
Cheap at £ 2.39.......
http://www.ebay.co.uk/itm/SILVERLINE-BI ... 2073wt_952

I deduce that if you use one of those tools, you may change your chain sooner than necessary, because you are measuring roller wear as well. However I have to agree about the simplicity. I don't know how accurate a cheapo Silverline one is compared with say a Park one.

[NATURAL ANKLING]

Hi,

Hi,
I use this, it saves all that farting around with rules.
1% on this and I reckon the chain is stuffed.
Cheap at £ 2.39.......
http://www.ebay.co.uk/itm/SILVERLINE-BI ... 2073wt_952

I deduce that if you use one of those tools, you may change your chain sooner than necessary, because you are measuring roller wear as well. However I have to agree about the simplicity. I don't know how accurate a cheapo Silverline one is compared with say a Park one.

As said its not simple as all chains are minutely different.
So one tool will not suit all chains..........but remember its the chain sprocket that the worn or not worn chain will run on

Do we have different chains for different sprockets........................No.
What we are (ideally) looking at is the pitch of a chain over several links, not the individual roller or clearance etc etc.

I will have a go at "Reverse Engineering" the Silverline tool, and see what one of my models brings up on this very controversial subject.........