Ruler and Chain Checker are Diverging
 

I measured the chain wear on one of my bikes today. On the Pedro Chain Checker Plus II, the #3 hook is easily going into links, which indicates a wear rate greater than .75%. However, when I measured the chain wear with a ruler, it only stretched by 1/32 of an inch. Based on these measurements, should I replace the chain immediately? Or, can I get a little bit more life out of my chain?

Ruler trumps chain checker.
Use the checker for a quick check. You now know that it reads .25% high on the .75% scale.
Not uncommon at all for them to be "generous".

You might want to use a 3' section of tape measure. It'll give you 3X the resolution vs checking a 12" section. I use the 1 & 37" marks to avoid dealing with the hook.
That would give a little better idea of the chain checkers true error.
https://cimg6.ibsrv.net/gimg/bikefor...6a4d20b69f.jpg

Chain checker trumps a ruler. You're attempting to avoid wearing out your sprockets, and sprockets run on the rollers the checker measures, not the pins a ruler measures.

And this is what that looks like when there is no pin stretch (ruler) but roller wear:
https://cimg3.ibsrv.net/gimg/bikefor...0430f21a3a.png

The rule for chain replacement is 1/16" elongation, so your ruler is saying 50% and the chain checker is saying 75% (and might have read 25% on a brand new chain). And don't get confused that there is a magic "worn vs. not worn" point. Wear is linear, not digital. Things don't instantaneously get bad when chain wear crosses a threshold.

While Kontact is theoretically correct his example assumes a significantly different wear rate on rollers vs. pins, which is not the case. There are lots of examples of chain checkers being wrong, but rulers never lie.

My example shows how roller wear alone can damage the sprockets. But roller wear AND pin wear don't cancel each other out - together they are going to produce the geometrical effect depicted in the graphic - that the pitch of the chain is effectively different when curved around a sprocket.

What examples do you have of chain checkers being wrong? How did you test that they were wrong?

And the whole 25% thing is a canard. It doesn't matter what the chain measures until that measure exceeds whatever tolerance the manufacturer recommends as "No go". We don't measure the chain to monitor its wear - that wear might not be linear. We measure the chain to take note of whether it is still in-spec or not. When it is not, you replace it to keep the rest of the drivetrain in shape.

Chain wear absolutely 100% is not linear.

The easiest way to measure for me is hanging the entire cleaned chain against a new one. Measure in mm with one of those digital vernier caliper thingies. Makes the math easier.

My eyeballs are not so good at 1/32, they like 5 or 6 mm better

I was not suggesting that roller and pin wear cancel each other out. I'm simply stating that chains don't exclusively wear on one or the other, which would be the argument that a ruler is not a good way to measure chain wear. If the chain is wearing, the rollers are wearing AND the pins are wearing and the ruler will show the latter, which is a good marker for roller wear as well. For years Campagnolo used simple elongation as the measurement for wear. While roller wear might be slightly more precise than just length, most chain checkers offer no real advantage over a ruler and several mislead the user.

Over the decades lots of people have reported their chain checker showing 75% wear on a brand new chain. That sounds wrong to me.

My "linear" comment was meant to be a contrast with the "go/no-go" mindset induced by chain checkers. I don't know that anyone has ever published research showing that the rate of chain wear somehow changes as the mileage increases, but perhaps you can share what you know.

There is no need to be able to see 1/32" since it is 1/16" that is considered the replacement point. That is easy to see in a 12" length of chain, even with a cheap ruler that has only 1/8" gradations. There's no need to remove, clean, and hand a chain to see if it is worn out. Just measure it on the bike.

I have seen a lot of people complain, as you did, that chain checkers show 25% "wear" right off the bat, but not 75%.

Some chain checkers may be out of spec. When I ran service departments I rounded up all the chain checkers and allowed only one model to be used, then checked all examples as behaving the same on several examples of worn and new chains. There is little chance that four Park CC 3.2 are all wrong and by exactly the same amount.

A home mechanic would be wise to check their tool against new chains as well.

Elongation alone may work fine - if the rollers hold up as well as the part of the chain that's being measured. If the rollers wear faster than the pins and links, then you're measuring the wrong thing.


The goal here is to replace your chain before it puts unnecessary wear on other components. I have yet to hear anyone complain that their chain checker caused them to replace their properly lubricated chain every 500 miles.

Any instrument can lie. Or, more properly, can be read incorrectly. A ruler measurement of a chain is fraught with issues. The end tab on a tape measures move which throws off the measurement if you are using that end. Measuring between, say, 1” and 13” would give more accurate results but precision can still be quite bad because of the way that individuals measure things. The start and stop point are different from person to person or even from measurement to measurement. If a rule is used, it should be measured several times and the average taken. And, as has been pointed out above, reading the rule can be an issue.

There’s also the problem of how the chain is measured with a rule. If the chain slack or tight? If it is slack, that’s going to throw off the measurement. Tight will give you a better measurement, something that chain checkers do automatically. The advantages of a chain checker are that it tightens the chain and it doesn’t depend on reading anything.

Some people suggest using a 12” rule and then estimating that last 1/16”. Estimating anything is by definition not accurate. Additionally 1/16” is only 0.5% wear. Most chain checkers will give 0.75% as the no go measurement for 10 speed and lower chains and 0.5% wear for 11 to 12 speed chains. 0.75% on a chain would be 3/32”, not 1/16”. That’s even harder to measure and especially to “estimate”

Chains do wear exclusively on one side. The leading edge of the chain is always going to be under pressure as the chain goes over the top of the chainring. As the chain goes under the cog on the back, the same edge is going to be under pressure. The pin doesn’t spin in the chain so the same side of the pin is under pressure and is going to wear.

It is wrong and can be attributed to user error. Again, any instrument can lie.

Also make sure the chain is nice and clean before using a chain checker. I am sure a straight edge measuring tool is a fine tool but the chain checker is generally my go to so long as you have a clean chain. However anything can be inaccurate like cyccommute said it just depends on the user and the quality of instruments.

My normal routine is to use a chain gauge (coincidentally, the CC-3.2 @Kontact mentioned) to find out if it's time to break out the ruler. However . . .

Using a ruler isn't automatic. I'll go with a longer rule, such as a measuring tape or the Park SBC-1, so I can center the beginning (1" on the tape or the triangle on the Park) on a rivet. Measure to the center of the '12" rivet', then go back and make sure the beginning mark didn't shift. Repeat at least one more time to make sure you've got the right measure. Or you can measure to one end or the other of the rivet, if you're not sure you can eyeball the middle of a rivet accurately.

And then go measure another one or two 12" sections of chain, to see if the measurements match or are longer than you want the chain to be to stay on the bike. If there's significant divergence, go with the longer measurement IME, and replace the chain.

IIRC the Pedros tool is supposed to take roller slop out of the measurement, FWIW.

My guess is that your "ruler" moved while you looked from one end to the other. That's one free opinion that's worth everything you paid for it! :)

wear rate increase rapidly once the surface treatment wears off

aside from measurement sensitivity, crude inside will invalidate the measurement

https://www.velonews.com/gear/measur...ar-accurately/

I find it interesting that people who staunchly rely on dedicated brandname bicycle tools like Park or Pedro’s won’t trust the chain checking tools from those companies. I mean they produce accurate tools for every other aspect of bicycle maintenance…including torque wrenches for delicate CF fixtures. But for some reason their chain checkers can’t be trusted??? I even know of one case where a person deferred to a Park Tool ruler instead of a chain checking tool. If the chain checker is supposedly inaccurate…who’s to say the ruler sold by the same company isn’t inaccurate?

Dan

I've found the newer Park CC 3.2 with the .5 measurement easy to use and sufficient. I wouldn't discourage anyone from using a ruler but how do you find one to trust? Is a HF China- made assumed to be accurate? Or is it accurate enough, same as a tool?:foo:

If this is aimed at my post, the answer's simple. There've been many laborious threads on why chain checkers aren't reliabile, having to do with precisely what is being measured. The Park chain checker works well for me as a gross check, but roller wear means that after the 0.5% mark, I trust a ruler more than any chain checker.

Post not aimed at anyone in particular.

Dan

I am one of those guys that buys economical freewheels and chains. I also ride allot of dusty trails so I end up cleaning my chains frequently. I have wasted my money on three different chain-wear tools and none of them beat my ruler so I am sticking with that. Further more I have finally learned that replacing bike chains is much better then replacing cogs and rings. Bill's 3' trick is one I am going to use from now on...

I've never used a chain checker clown device. Never will. LOL.
Measuring any distance more than 12" is just plain dumb. Slackness will make a difference.
Of course you look at 10" to 22" or something, using a roll tape measure is best.
My last 1/8" IGH chain has never had any lube in over 3,000 miles. Lubes all suck dirt.

And .5 is enough for those using 11 speed.

Unfortunately, people don’t read the manual. From Park

5mm is easy to measure with my micrometer.

Clean the chain and hang it side by side with a new chain.

5mm is 0.4% and time for a new one.

A nail is cheaper than a Park go/no go gauge

Always trust the humble RULER. Get rid of the fancy dancy chain checkers.

1/32 inch is only half time. Replace when it wears 1/16 inch.

Good write up and illustrations below:

http://atbicycle.blogspot.com/2014/0...nch-ruler.html

New chain:
https://cimg7.ibsrv.net/gimg/bikefor...3311c69581.jpg


Worn 1/16 inch - Replace chain.
https://cimg3.ibsrv.net/gimg/bikefor...45c6ad75bc.jpg


Worn 1/8 inch - Your cassette is probably trashed at this point.
https://cimg4.ibsrv.net/gimg/bikefor...55c7aee9de.jpg

Wrong! I had a Park Tool torque wrench that quickly became erratic and eventually tightened past the set point. I replaced it with a Shimano Pro torque wrench which has been much more reliable.

1/16” is only 0.5%. 3/32” is closer to 0.75%. I know that’s not much but it could be a few hundred miles worth of chain use.


I’m glad you posted the pictures below. It will illustrate my point about the 12” rule nonsense. I don’t usually leave the pictures in a quote but I’ll do so here.

So far so good.

Any time someone uses “approximate”, there can be absolutely no claim made of any kind of accuracy…nor precision for that matter. No instrument can be used to accurately measure anything beyond its range. Going beyond the range is useless data. What is shown is ‘12” plus something’. To make the claim that this is more “accurate” than other methods is to not understand what “accurate” means. All that can be said about this measurement is that the chain has some wear.

Making the estimate larger doesn’t make for a more accurate statement.