Chain wear

[JXV]

A few threads recently referenced one or other article that shows how the popular chain checkers are inaccurate.  I have the Park Tools CC-3.2 checker and up till recently, I have trusted this gadget to advise me when to replace my chain.   So I decided to calibrate it and see how bad it really is.

 

The best way to measure a chain is to remove from the bike, stretch it out under light tension and measure it with an accurate ruler against a brand new chain of the same type.  The longer the ruler, the better.   I have a 1m steel workshop ruler in my workshop at home but what a pain.....my chain checker is much more convenient so if I know how dishonest it is, I can still use it for a quick check. 

 

There are many articles on chain wear but I like this one the best : http://pardo.net/bike/pic/fail-004/000.html

It has diagrams, photos and references to engineering texts and is well written.

 

Summarizing this article  :

1) it is only pin-to-pin wear that matters, this is what causes chain length to increase resulting in wear of cassette and chainring teeth.

2) roller wear is practically a non-issue because it does not affect the pin spacing. 

3) all but one of the popular commercial chain checkers include roller wear in their measurement of chain length and thus tend to overestimate the actual extension in chain length.

 

I tend to use the same type of chain all the time (Shimano CN-HG95 HG-X 10 spd) so I figured that if I calibrate my chain checker to this one type, some truth might emerge.  

 

I measured a worn chain that I replaced at 0.5% according to my checker.   Pin to pin distance should be 12.7mm (1/2") on a new chain so 50 full links (101 pins) should measure 50" (1270mm).   My "0.5%- according-to-Park-Tools" worn chain actually measured 50+1/16 inches (1272mm) across 101 pins, corresponding to only 0.125% wear.   The chain currently on my bike is somewhere between the 0.5% and 0.75% mark on the chain checker and I measured this one at 0.25% wear.   So theoretically both these "worn" chains still have useable life.   I figure that my chain is still less than 0.5% stretched when my checker indicates 0.75%. so I am happy to use this  checker up to its maximum for this type of chain.

 

By replacing chains at "0.5%" as indicated by this checker I am probably changing chains at only 1/3 of their useable life and thus buying new chains three times more often than is necessary.   Something to think about..... 

[Odinson]

TL;DR

[johannrissik]

A few threads recently referenced one or other article that shows how the popular chain checkers are inaccurate.  I have the Park Tools CC-3.2 checker and up till recently, I have trusted this gadget to advise me when to replace my chain.   So I decided to calibrate it and see how bad it really is.

 

The best way to measure a chain is to remove from the bike, stretch it out under light tension and measure it with an accurate ruler against a brand new chain of the same type.  The longer the ruler, the better.   I have a 1m steel workshop ruler in my workshop at home but what a pain.....my chain checker is much more convenient so if I know how dishonest it is, I can still use it for a quick check. 

 

There are many articles on chain wear but I like this one the best : http://pardo.net/bike/pic/fail-004/000.html

It has diagrams, photos and references to engineering texts and is well written.

 

Summarizing this article  :

1) it is only pin-to-pin wear that matters, this is what causes chain length to increase resulting in wear of cassette and chainring teeth.

2) roller wear is practically a non-issue because it does not affect the pin spacing. 

3) all but one of the popular commercial chain checkers include roller wear in their measurement of chain length and thus tend to overestimate the actual extension in chain length.

 

I tend to use the same type of chain all the time (Shimano CN-HG95 HG-X 10 spd) so I figured that if I calibrate my chain checker to this one type, some truth might emerge.  

 

I measured a worn chain that I replaced at 0.5% according to my checker.   Pin to pin distance should be 12.7mm (1/2") on a new chain so 50 full links (101 pins) should measure 50" (1270mm).   My "0.5%- according-to-Park-Tools" worn chain actually measured 50+1/16 inches (1272mm) across 101 pins, corresponding to only 0.125% wear.   The chain currently on my bike is somewhere between the 0.5% and 0.75% mark on the chain checker and I measured this one at 0.25% wear.   So theoretically both these "worn" chains still have useable life.   I figure that my chain is still less than 0.5% stretched when my checker indicates 0.75%. so I am happy to use this  checker up to its maximum for this type of chain.

 

By replacing chains at "0.5%" as indicated by this checker I am probably changing chains at only 1/3 of their useable life and thus buying new chains three times more often than is necessary.   Something to think about..... 

Agreed, the short, 'push-in" chain checkers have a built-in error, but I guess that way they (the toolmakers) cover their rear ends.

 

Someone (and no, don't look at me) will no doubt one day develop an app where you plug on your chain type, cassette type, price of each, etc and the app will then tell you when to change your chain.....

 

Someone please do that.

 

Conventional wisdom (if ever there was a contradiciton in terms ) is that you replace it at 0.5% wear (properly measured), but this depends primarily on the cost ratio between chain (s) and cassette.

 

A bit like a VW Caddy I was in recently, the odometer over-reads by 8% ?? I'm sure VW have an explanation for that too.

[JXV]

Agreed, the short, 'push-in" chain checkers have a built-in error, but I guess that way they (the toolmakers) cover their rear ends.

 

Someone (and no, don't look at me) will no doubt one day develop an app where you plug on your chain type, cassette type, price of each, etc and the app will then tell you when to change your chain.....

 

Someone please do that.

 

Conventional wisdom (if ever there was a contradiciton in terms ) is that you replace it at 0.5% wear (properly measured), but this depends primarily on the cost ratio between chain (s) and cassette.

 

A bit like a VW Caddy I was in recently, the odometer over-reads by 8% ?? I'm sure VW have an explanation for that too.

I guess if you ride XX1 cassettes it makes sense to change the relatively inexpensive chain at much shorter intervals....and perhaps to run 3 or 4 chains and alternate regularly.

[raptor-22]

At 0.75% wear indicated on a push in chain checker the shifting of the chain across the sprockets is noticeably reduced in quality and accuracy. Drivetrain noise is increased as well.

 

In my mind milking a chain to its last mm is poor practice. The sacrifice in expensive chain ring life and sprocket life is not worth it.

 

Lastly if roller wear is not considered jmportant than you need to spend more time working with industrial machines which use chains to help you understand why roller wear is important.

Worn rollers don't roll they tend to slide and that sliding action wears the sprockets faster.

Pin wear does indeed alter the rollers pitch but pin wear affects the inner running surface of the roller which has a bigger effect on chain wear.

[JXV]

At 0.75% wear indicated on a push in chain checker the shifting of the chain across the sprockets is noticeably reduced in quality and accuracy. Drivetrain noise is increased as well.

 

In my mind milking a chain to its last mm is poor practice. The sacrifice in expensive chain ring life and sprocket life is not worth it.

 

Lastly if roller wear is not considered jmportant than you need to spend more time working with industrial machines which use chains to help you understand why roller wear is important.

Worn rollers don't roll they tend to slide and that sliding action wears the sprockets faster.

Pin wear does indeed alter the rollers pitch but pin wear affects the inner running surface of the roller which has a bigger effect on chain wear.

By the time roller wear becomes significant on a mtb chain the pin wear and chain length problem is already much worse - I think this is why the author of that article says roller wear can be disregarded in assessing bicycle chain wear.

 

I agree on the issue of not trying to eke every last km out of a chain...on our bikes the cassette and chainrings are worth preserving. The point of my post was that my chain checker significantly overstates the chain wear ( by a factor of 4x at 0.5%!!!) So it has been prompting me to change chains when they still have useable life even by conservative standards.

 

It is also better to buy several chains and alternate regularly on the same set of gears than to fit chains one after another as they wear out.

 

On my bike I have 2 chains worn to 0.13% and 0.25% on the same set of gears. I'm guessing that recycling the 0.13% chain will have less impact on the partially worn gears than a new chain would, so I'm going to run it for a while. Already got over 3000km on this drivetrain and still working ok.

 

I have some new chains, cassette and chainrings in waiting for when the current lot finally dies.

[raptor-22]

I've measured pins and rollers on my last cn hg 900 and compared it with a good one.

The inner diameter of the rollers were more worn than the pins.

This is reasonable since the pins appear to be much harder than the rollers.

 

Unless you can statistically show which component exhibits normally mire wear I assume that wear is consistent across pin and roller and therefore I am interested in both.

 

On alternating chains ;

I can't see the point as how do you track wear on the sprocket and chainring teeth. Unless you have measured easy tooth and compared it to the statistical production nork from the chain manufacturer you're just posturing. A wear control program needs to account for known starting points from all components not just one element.

[JXV]

I've measured pins and rollers on my last cn hg 900 and compared it with a good one.

The inner diameter of the rollers were more worn than the pins.

This is reasonable since the pins appear to be much harder than the rollers.

 

Unless you can statistically show which component exhibits normally mire wear I assume that wear is consistent across pin and roller and therefore I am interested in both.

 

On alternating chains ;

I can't see the point as how do you track wear on the sprocket and chainring teeth. Unless you have measured easy tooth and compared it to the statistical production nork from the chain manufacturer you're just posturing. A wear control program needs to account for known starting points from all components not just one element.

The bigger the disparity between pin spacing and gear tooth spacing the higher the chain rides in the teeth and the faster the teeth will wear. Hence the quite common advice to start a new set of gears with several new chains and swap out the chains regularly so the gears experience smaller incremental steps in chain growth between changes and hopefully wear more gradually as a result. This all assumes proper lubing and cleaning as well. We know how dirty a mtb chain gets so riding conditions will also affect wear even with a good cleaning regime. If we rode in diamond dust I think abrasive wear on the gears would totally overshadow chain stretch issues.

 

Even though the rollers wear, this does not affect the pin spacing (which the writer sees as the dominant influence on gear tooth wear in chains that are still functional). Yes, in a badly worn chain the rollers don't roll and then they eat the teeth even worse but we're talking about chains that get changed before this.

 

The most common advice I see written about chain wear is to change before 0.5% for 9 and 10spd drivetrains but in an era where nice wide range cassettes are costing upwards of R1500 perhaps we need to change even more often. The problem is measuring this wear. Even when you hang and measure a chain it is difficult to assess wear accurately if you are looking at amounts less than 0.25%. I did it with an accurate ruler, a vernier and a chain wear guage and got quite a range of results on the same chain. It is difficult to see and accurately align measurement marks at sub- 1mm distances. I'm guessing most folk couldn't be bothered so we still need a decent chain wear guage that measures accurately in the 0.1% to 0.5% range.

[raptor-22]

The bigger the disparity between pin spacing and gear tooth spacing the higher the chain rides in the teeth and the faster the teeth will wear. Hence the quite common advice to start a new set of gears with several new chains and swap out the chains regularly so the gears experience smaller incremental steps in chain growth between changes and hopefully wear more gradually as a result. This all assumes proper lubing and cleaning as well. We know how dirty a mtb chain gets so riding conditions will also affect wear even with a good cleaning regime. If we rode in diamond dust I think abrasive wear on the gears would totally overshadow chain stretch issues.

 

Even though the rollers wear, this does not affect the pin spacing (which the writer sees as the dominant influence on gear tooth wear in chains that are still functional). Yes, in a badly worn chain the rollers don't roll and then they eat the teeth even worse but we're talking about chains that get changed before this.

 

The most common advice I see written about chain wear is to change before 0.5% for 9 and 10spd drivetrains but in an era where nice wide range cassettes are costing upwards of R1500 perhaps we need to change even more often. The problem is measuring this wear. Even when you hang and measure a chain it is difficult to assess wear accurately if you are looking at amounts less than 0.25%. I did it with an accurate ruler, a vernier and a chain wear guage and got quite a range of results on the same chain. It is difficult to see and accurately align measurement marks at sub- 1mm distances. I'm guessing most folk couldn't be bothered so we still need a decent chain wear guage that measures accurately in the 0.1% to 0.5% range.

 

 

Indeed, I'd agree with at.

[Uni]

Is there a way to check your chain without removing it? Are there any tell tale signs one can look out for,or a general rule of thumb..after x kms replace.

[johannrissik]

Is there a way to check your chain without removing it? Are there any tell tale signs one can look out for,or a general rule of thumb..after x kms replace.

Uni,

When in doubt, ask Sheldon (well, metaphysically anyway).

 

http://sheldonbrown.com/chainclean.html

 

Once you've cleaned it and re-assembled it, you can use a chain-checker and be ready for the great "chain-stretch debate"

 

[Uni]

Uni,

When in doubt, ask Sheldon (well, metaphysically anyway).

 

http://sheldonbrown.com/chainclean.html

 

Once you've cleaned it and re-assembled it, you can use a chain-checker and be ready for the great "chain-stretch debate"

I should start with a what not to do list

 

Nice one Johan... You know I would have gone and done that .. And then get to the end of the page and gone doh!

 

[droo]

Maybe I'm oversimplifying this, but if there's already a chain checker on the market that measures pin to pin wear, why not just use that one?

 

I just measure sideplate to sideplate with an inch ruler after cleaning the chain and replace at slightly less than 1/16" elongation over 12", which works out to 0.52%.

[Eddy Gordo]

http://roadcyclinguk.com/how-to/maintenance/maintenance-chain-checking-and-replacement-part-one.html#reWDkxegFq2QRC2h.97

 

 

http://cdn.coresites.factorymedia.com/rcuk/wp-content/uploads/2013/02/NewChainOldChain-620x413.jpg

 

The inner one being worn

[MrJacques]

Maybe a silly question, but does the type of chain (e.g. 9 / 10 / 11 speed) make a difference to how you measure chain wear? Do you need to change the measurements used?

 

Are the various chain wear checkers 'speed' specific?

[Jacoo]

Kinda. You use the same chain checker though. Ideally you want to replace 11 and 12 speed chains at 0.5, and 10 speed and lower at 0.75. Single speed at 1.0. I run 10 speed and replace the chain not too long after 0.5 just to make sure my cassette lasts as long as possible since its an expensive cassette.