Originally Posted by
Bill Kapaun
At what temperature? Just because it's a digital caliper doesn't mean you have an accurate reading, considering the inherent inaccuracy of a digital display and thermal expansion of a steel chain re: temperature.
My checker reads ."about" .25% worse than reality on the "short" side with a "snug" fit. I haven't had a worn enough chain to check the "long" side.
So, as long as the short side says "good" to "snug", I know I don't need to check further.
Good point, I measured before and after rides, under different temperatures. That's why I gave the worse case range in my original post. If the chain is cold I measured as little as 5.001 ~ %0.02.
%0.25 is a lot but I am also experiencing something in the range of %0.15 so I wouldn't be surprised.
Originally Posted by
gauvins
I have a reasonably good caliper, certified at 0.02mm accuracy.
The problem wrt measuring a bike chain is that the way you position your caliper will have a definite impact on the reading. I took 30+ readings over a couple of weeks, on a slightly used but clean chain. The measurement MAPE (mean average percent error) turned out to be 0.28%. (compare it to the critical value of 0.5% at which a chain might/should be retired and you may conclude that it is too high or somewhat ok). My personal conclusion is that using a caliper isn't particularly useful.
I find it much easier to nail a new chain to the wall use it as a reference for the chains I use in rotation -- I hang the used chain on top of the reference (i.e. hung to the same nail) and retire it it gets close to one link longer. A chain is much cheaper than an accurate caliper, more accurate (probably < 0.05%), and is fast and foolproof.
I completely agree that positioning plays a big role in taking measurements that's why I am also trying to see if the discrepancy was due to my inability to measure it right.
That is also the reason why I am not trying to measure roller to roller distance from inside the chain since those knife edges make it almost impossible to get a consistent reading (they don't fit well and make it hard to alight the caliper with the chain line ). However from outside the chain, just like using a ruler, I believe I get fairly consistent measurements. I try to get at least 5 measurements from different sections.
Since these chain checkers are also measuring a similar, around 5"/5 full links, distance I think it is fair to compare the digital caliper readings with a chain checker. I believed these drop in chain checkers were machined precisely (otherwise it is just a piece of metal) so it should be by design that they measure less than what they should. Have you tried to measure the chain checker itself with your caliper?
As you have mentioned, hanging the chains side by side is the better way of measuring average chain wear (unfortunately I don't have the space for it atm). One problem with this approach is that there will be variance section by section and while the average may be well within %0.5 there may be individual sections that exceed it. With chain checkers we take the worst section(say 5 full link) wear rather than the average.