[QUOTE 3627974, member: 9609"]I regularly measured my last chain during its life, as I knew I was changing the cassette anyway I kept it going until it failed.
here's a little chart showing wear against distance (mm/12 links/miles)
Wear seems to be quite exponential, not much change in the first 2000 miles, but it was never going to make it to 6000 miles
0.5% wear would be 306mm
@Dogtrousers Interestingly your chain started to slip at 310mm a little before mine did but same ball park.
some pictures of the cassette and chain ring in
this thread
[/QUOTE]
I can't tell you how much I like this graph. It tells the whole story of chain life mathematically. Like you say, chain wear is not linear. The inflexion point is at 307.8, which represents 0.5% elongation. Chains and sprockets are engineered to not damage each other up to this point. Thereafter, wear of both components is accelerated. This is the point where the incoming tooth on the chainring starts to hit the incoming roller on the chain and forces its way past it. At the rear, the chain now starts to engage only one roller in the job of tensioning the sprocket and thus concentrates wear on only one tooth.
Your start point is interesting too. The immediate elongation a ride or two after fitting the new chain shows how the chain quickly elongates during the first ride. This is not because of wear but because of link correcting. Imagine a wooden latter made with morticed and tennoned rungs but no glue. t's rungs aren't perfectly perpendicular just after manufacture but as soon as its has seen a bit of use, the rungs settle perfectly in the mortices and the ladder straightens out. The same happens with a chain.
I suspect there is a bit of measuring error in the first five measurments. I would have expected the length to suddenly increase (a bit) and then settle. In other words, the 3rd, 4th and 5th data points are too small. I just can't explain them.
PM me your postal address. I have a little gift for you in this line.