Cycling UK Forum

As has already been discussed elsewhere at some length, there is a school of thought that says that including some roller wear in chain wear measurement (as per most chain checkers on the market) is not a good thing, in that it gives different measurements with chains of different types, even though they might be similarly worn/unworn.

So is there a simple alternative? I'm happy to use a 12" ruler but I know some folk are not happy with this method and would prefer some kind of unequivocal measurement. Below I have written up a method which might fit the bill. All you need is a set of 6" digital Vernier calipers (typical cost £12-20 for a basic version) which many cyclists will have already or find very useful for other purposes.

Step 1. Zero the caliper. Whilst taking a between rollers measurement (BRM), set the caliper to zero in the 'inches' range.



Step 2. Take the measurement. Slide the caliper out to an indicated ~4.9" and then take a measurement between the LH roller as before and the appropriate RH roller.



Step 3. Do the calculation. An unworn chain should measure almost exactly 5.000" on this test, (but +/- 0.005" isn't unusual even on new chain)

Any excess over that represents pin wear in the chain. Thus a 1% worn chain will measure 5.050", and a 0.5% worn chain will measure 5.025", and so on.

The chain below is (with the caliper re-zeroed for that chains BRM value) +0.043" so is almost 0.9% worn.



Checks and balances: There are additional checks that you can do if you want;

1) Check uniformity of roller wear. With the calipers zeroed as step 1, you can check other roller gaps. With 1/8" chain using most verniers you can check any chain gap, but with derailleur chain you can only check between outer side plate links (as per the photos) unless you modify your verniers by grinding the internal jaws to make them slimmer than normal. Most worn chains I have checked thus show remarkably uniform roller spacing, but if a (singlespeed) chain is used on even tooth count rings/sprockets, every other link may wear differently.

2) Check uniformity of chain wear. You can repeat steps 1-3 as many times as you can be bothered to do along the chain length. If the chain you are measuring shows uniform BRM then you can take measurements without having to re-zero the caliper each time.

3) check roller wear. In the image below you can see the difference in BRM as measured using a short length of unused chain of the same type (KMC 'inox').



You can see that the roller wear is ~0.024" different between new and used chains. Had this been included in the main wear measurement, this arguably would have constituted a >50% error. Even new chains show a variation of +/- 0.25mm (~0.010") in this measurement, due to small variations in roller OD and bushing clearances.

So there you have it; I think this method may be a both simple and effective one for keeping track of chain wear. Having said that, I don't think it adds much to the results you can get by simply using a ruler, but not everyone is happy to do that. The vernier method arguably has a flaw (which it shares with all commercial chain checkers, pretty much) in that it relies on the rollers being free to move in the normal way that occurs when the chain is in use. For various reasons this might not always happen.

NB I have taken these measurements on short lengths of loose chain, but there it is probably best if the measurements are taken whilst the chain is still on the bike. For example if the chain is very dirty, whilst the chain is still in its usual position, (just used) the rollers will be able to assume their usual positions without difficulty, but once the chain is removed or otherwise disturbed, dirt in the rollers may affect the measurement. For this reason, some folk advocate cleaning the chain before you measure it. If you are going to remove it and clean it anyway, why not.... but if you are not, it means that you will regularly waste your time cleaning a chain that is actually headed for the bin....

cheers

The zeroing is easier said than done with my calipers as the 2 pointers barely fit within a link ( well they dont) and I am not sure what part of the curved surfaces they are measuring? ( unless I havent understood what the zeroing is all about. I tried this idea some time ago an rejected it in my case ) would be better if there was a zeroing like the measureing involving more than one link so only one caliper per link?

When I do it I as suggested I get my new chain is 50.025 one time ( 8 speed Sram) I find it hard to get repeating consistant results. Good idea if you have some calipers which are thin.

The Shimano TL-CN42 tool is quite effective. Splitting the chain, hanging on a nail and measuring again a steel ruler is a lot cheaper.

The Campag Method is the same as Brucey's but done in metric.

REPLACEMENT: Use a high precision caliper gauge to measure, in different points of the chain, the length as indicated in fig. 5. If even one of the measurement is longer than 132.60 mm the chain must be replaced.

This is a very timely thread as only this morning I measured the wear on two of my bikes using the 12" ruler method.
(My 10-speed chain appeared to have 1.5% of wear - how did that creep up on me?)

I revisited both bikes and tried Brucey's method. With a new chain on the worktop, the jaws of my Aldi calipers were just thin enough to squeeze in one gap and the results indicated zero wear. However on the bikes it was more difficult until I resorted to zeroing with the jaws in gaps one inch apart (and measuring over four inches).

This gave consistent and repeatable results and confirmed this morning's results.

Mick F wrote:The Campag Method is the same as Brucey's but done in metric.

REPLACEMENT: Use a high precision caliper gauge to measure, in different points of the chain, the length as indicated in fig. 5. If even one of the measurement is longer than 132.60 mm the chain must be replaced.

Chain measurement.png

No, it isn't the same; they include roller wear in their measurement, because they don't zero the calipers initially.

cheers

Yep. Ok.
The fact that they say what they say, it is obviously about Campag chains.
I don't use that method, and although I've had a go with it, It's too fiddly and "variable" in its results.

Best way I've found over the years is to use a long steel rule. Mine is a metre long/39 inches.
I have a small G clamp that I clamp the chain on top of the steel rule using our big pine kitchen table, and lay the chain along the rule.
One eighth of an inch over the 39inches is too long IMHO and the chain is on its way out. If it gets beyond that, it gets retired.

However ................ now I've discovered the delights of Shimano, I don't give a toss because the cassettes and the individual sprockets are dirt cheap. Keep the chain clean and happy, and within reason, it can be as long as it wants.

Mick F wrote:One eighth of an inch over the 39inches is too long IMHO and the chain is on its way out. If it gets beyond that, it gets retired.

That's only about 0.3% wear if I'm correct. I used to let mine go to 1% (I use 1/8" in 12 inches), but now I try to retire when they get to 1/16" in 12 inches as I found the 1% would give a little jumping with a new chain for a few miles.

I have a new chain hanging on a nail in the shed. I hang my used chain after cleaning from the same nail. This gives a good visual indication of chain stretch and indicates when I need to change it.
You could even hang a worn chain up and use this as a reference to compare with a used chain to indicate end of life.

That is what I do, except I have replaced the two chains with a couple of lines drawn with a pencil.

To mark the ends of the chain, not the whole length, just in case somebody thought I had gone for the whole artistic representation.

rjb wrote:I have a new chain hanging on a nail in the shed. I hang my used chain after cleaning from the same nail. This gives a good visual indication of chain stretch and indicates when I need to change it.
You could even hang a worn chain up and use this as a reference to compare with a used chain to indicate end of life.

OK, well a used chain will always measure longer than a new one. What criteria do you use to tell if your chain is stretched too long? Chains are also different lengths depending on the bike application.

MikeDee wrote:

rjb wrote:I have a new chain hanging on a nail in the shed. I hang my used chain after cleaning from the same nail. This gives a good visual indication of chain stretch and indicates when I need to change it.
You could even hang a worn chain up and use this as a reference to compare with a used chain to indicate end of life.

OK, well a used chain will always measure longer than a new one. What criteria do you use to tell if your chain is stretched too long? Chains are also different lengths depending on the bike application.

Half a link over 100 links (or 108 or 112 whichever you use) is 0.5%. This is easy to assess and about right as a change point in my experience. Absolute accuracy is not important as the best degree of "stretch" at which to renew is only a subjective estimate anyway.

Brucey wrote: I'm happy to use a 12" ruler but I know some folk are not happy with this method and would prefer some kind of unequivocal measurement. Below I have written up a method which might fit the bill. All you need is a set of 6" digital Vernier calipers (typical cost £12-20 for a basic version) which many cyclists will have already or find very useful for other purposes.

Step 1. Zero the caliper. Whilst taking a between rollers measurement (BRM), set the caliper to zero in the 'inches' range.

Image234.jpg

Step 2. Take the measurement. Slide the caliper out to an indicated ~4.9" and then take a measurement between the LH roller as before and the appropriate RH roller.

Image235.jpg

Step 3. Do the calculation. An unworn chain should measure almost exactly 5.000" on this test, (but +/- 0.005" isn't unusual even on new chain)

Any excess over that represents pin wear in the chain. Thus a 1% worn chain will measure 5.050", and a 0.5% worn chain will measure 5.025", and so on.

The chain below is (with the caliper re-zeroed for that chains BRM value) +0.043" so is almost 0.9% worn.

Image237.jpg

So there you have it; I think this method may be a both simple and effective one for keeping track of chain wear. Having said that, I don't think it adds much to the results you can get by simply using a ruler, but not everyone is happy to do that. The vernier method arguably has a flaw (which it shares with all commercial chain checkers, pretty much) in that it relies on the rollers being free to move in the normal way that occurs when the chain is in use. For various reasons this might not always happen.

NB I have taken these measurements on short lengths of loose chain, but there it is probably best if the measurements are taken whilst the chain is still on the bike. For example if the chain is very dirty, whilst the chain is still in its usual position, (just used) the rollers will be able to assume their usual positions without difficulty, but once the chain is removed or otherwise disturbed, dirt in the rollers may affect the measurement. For this reason, some folk advocate cleaning the chain before you measure it. If you are going to remove it and clean it anyway, why not.... but if you are not, it means that you will regularly waste your time cleaning a chain that is actually headed for the bin....

cheers

I'm liking this method. Many thanks

Just checked 3x 8 speed KMC chains in situ on 3 bikes with a steel rule and was gobsmacked at how nigh on worn out they appeared to be having only covered between 747 and 1435 miles

Then used this method to find the most worn was only 5.021 in. My eyes and or steel rule method were deceiving me

an intersting article that summarises it all

https://www.bikeradar.com/road/gear/article/bicycle-chain-wear-explained-46015/

mercalia wrote:an intersting article that summarises it all

https://www.bikeradar.com/road/gear/article/bicycle-chain-wear-explained-46015/

that is one of the better articles on the subject but they fail to mention whether the KMC digital gauge measures pin wear only or whether sloppy rollers can affect the measurement. There are only a few chain checkers that measure pin wear only.

If you use the same brand of chain and in the same conditions all the time then you may get sufficiently consistent results using one of the inferior chain wear checkers, but it is a far better idea to use a better chain wear measuring tool to start with. However a problem with all such tools (if they are made simply) is that they have some kind of a wedge action into the chain link; this is to give accuracy but it has the unintended effect of making it possible to exert a large force using the tool, which can give an erratic reading or even damage the tool.

The method I described upthread measures pin wear only (at low force, in situ) and easily/quickly gives you an unequivocal measurement. It may be the only method that does this, in fact.

cheers