DaveSSS 

Getting only 4500 miles or two chain's worth of riding with a cassette is not very good life. If new-chain skip occurred when a third chain was installed, it usually means that the first two were used too long. The exception might be the case when a small sprocket wore out. Apparently, some flat land riders use some of the smallest sprockets a lot and wear them out because they have so few teeth to share the wear.

Troul 

The factors that I use to determine as to when my chain should be considered for being replaced comes down to personal experiences that incorporates a slew of things like:
operational use when all else is in good working order [cogs/cables-housings] ; does it hang up during shifting, have trouble maintaining linear pulling force (ie: pedaling during powering up some climbing) is it loud or quiet when rotating
reliability risks; will I be stranded in a worse moment
price; if I find a deal to stock up on chains, it comforts me to know I will not have readiness impacted greatly when the weather is allowing for a nice ride

The matrix aiding in the above is often related to where I have been riding. Meaning, my riding is 99% on the road. I have not had a chain not become dirty & gritty after a few times of being used no matter how prudent I am to cleaning it. Dirt finds a way in & on the chain. I've accepted that my riding areas are just going to make the chain nasty since road construction is a contain thing that kicks up contamination. I will never expect a chain manufacturer to develop an affordable chain that can stay clean & quiet for its life expectancy without me heavily maintaining the chains condition.

So I replace the chain when I determine it could impact my ride in a worse outcome.

rm -rf

Ha, your math is correct. Thanks!

San Rensho 

its a steel 9 spd cassette and shimano 105

jgwilliams

Some interesting feedback here. The test pretty much confirms my opinion of the SRAM 1110 chain; my bike came with that and the matching cassette (which is a shame considering the rest of the groupset was Force). It lasted 2000 miles which I thought was appalling. Having previously used Record 9-speed chains on my previous bike I never got less than 8,000 miles. However, judging by the comments here 2,000 miles doesn't seem that bad. I've gone for a SRAM 1170 chain this time - we'll see how that does. I've also now got a Shimano 105 cassette, but that was purely because I didn't like the way the spacing on the SRAM cassette, with the jumps getting progressively bigger as the gear got lower.

mstateglfr 

As much as I love bikes as a whole and broken into the components, I just cant get into chains.
I geek on geometry theory, frame materials and design, as well as the mechanics of components...but chains are just a total bore.

I made it only part way thru the article before I started scrolling for something interesting. I made it halfway thru this thread before scrolling for something interesting. Its a chain- clean it from time to time and replace it when its worn.

A chain is the pinky toe of the bike. Chains are boring, plain, and forgettable...but vital to proper operation of the bike.

Sy Reene

but but but... chains have more moving parts than any other component on the bike.

Reynolds 531 

I'm guessing you've never experienced "New Gold Chain Day".

You should try it.

It could change your life.

63rickert

Back in the day of 5 and 6 speed freewheels it was common to ride chains to 1% elongation. That was because 1% was where conspicuous problems would begin. Jobst Brandt is dead wrong on this one. There's a reason a lot of us who can't avoid running into him disagree on everything and only pay attention to his engineering explanations when it seems his reality has some contact with our reality. Back in the day of Campy Nuovo Record being the universal chainring for anyone who did high miles, wearing out 'rings was a big deal. It was not just that they were expensive, you could not always buy them at all. Out of stock and none expected. Mail order meant snail mail. Phone calls to LBS around the country were priced by the minute. So we replaced chains at 0.5% elongation. And still wore out 'rings and freewheels.

When I tell others I can run a Campy chain 6000 miles and only get 0.25% elongation most simply will not believe me. Like you I find that the drivetrain is showing wear if I let a new chain go that far. Others on thread are talking 0.5% and 0.75% with 11 speed setups. That sounds completely impossible to me. It is different for different people. So many say that chains are only good for 2000 miles they cannot all be completely wrong. So many say their 11 speed indexing works fine at 0.5% elongation they cannot be completely wrong.

Some outliers. My wife weighs a mere 105# and has been lighter than that most of her life. Her daily rider - she does ride almost every day and always has - has a freewheel installed in 1983. It shows no wear that anyone could care about. This includes every winter riding all through the road salt season in Chicago. When she gets new chainrings - they keep getting smaller as she ages - I put on a new chain but it is not really necessary. Over on C&V there are posters who talk about using the same Campy chainrings since late 60s and claim these are bikes that have always been ridden. They go on to say Campy rings just do not wear out. Just vintage NR rings I am sure I've worn out 30 pairs of them. It is different for different riders.

Few people can or will accurately measure to the tolerances discussed here. Refer to the Velonews article and see how Connex measures. That will not be replicated at home. I know how I measure and believe my results. That does not mean I am perfectly accurate. It means if I keep doing the same thing I will keep getting same results. Even I do not believe the number, only the results. Any number cited by anyone take with a grain of salt.

The Connex test is with an electric motor putting out a constant load. Humans don't pedal that way. Smooth pedal style makes everything on the bike last longer. Rough pedal action wears out everything and chains are first. Big rough guys trash everything very quick and big racers the worst.

DaveSSS 

These days, I pay a lot more attention to the increase in individual pairs of rollers, since I know that a full length measurement of a Campy chain will never show much elongation. I also don't trash chains prematurely in an attempt to lengthen the life of a cassette that will eventually have a couple of cogs skip with a new chain. Instead, I pick however many chains I think constitutes a respectable life for the cassette and use each of those 4-5 chains longer than most would. I'll never use a 12 speed chain for 6,000 miles, but if I can get at least 3,000 from a 12 speed chain and 15,000 from a cassette, I'll be happy. These days, I only ride 3500-4000 miles a year, and split the mileage between two bikes, so those two cassettes might be the last I'll ever need. At best, they'll both be replaced and I'll just get started using them before my replacement knees wear out.

I've never had to replace a chain ring on any of my bikes, dating back to the 80's, even the ones that had 53/39/28 rings with the 28 used a lot while climbing the Colorado mountains. In the 2003-2010 years, I was riding at least 5,000 miles a year and splitting the mileage unequally, between a winter bike and my main bike. I did have two top level bikes for a couple of years.

aplcr0331

But which one saves moar WATTZZ!?

jadocs

I've known forever and keep telling people about these chains, in addition to the benefits of their quick link, but nobody listens LOL.

jadocs

Hot paraffin wax and PTFE powder, is what I am using on my Connex chains...best of both worlds!

DrIsotope

Get me an 11-speed Connex chain for less than $25 a pop and we can talk. Until then, Connex chains are just too expensive. Nearly 4x the cost per mile over what I use now.

base2 

I just can't for the life of me figure out how some here just blow through chains so regularly. 800 miles? 1000 miles? Is anyone serious? People blow through a chain every 6 weeks to 2 months? My Gawd! They must stand & stomp every gear at 50rpm to abuse a component like that.

I run exclusively Shimano 11s Dura-ace/XTR chains & get 3500 miles until I can force the early wear indicator between a couple of rollers...and I usually need to check 3 or 4 places to find a spot it'll go. Out of an abundance of caution, I change it then. I have no idea how many miles until the go-no-go indicates an actual need to change is imminent.

I just don't understand.
Do these same people who get only 800 miles out of a chain also blow through 2 or 3 or 4 bottom brackets a year as well? How many bottom brackets & freehub bodies does Mr. 800 Mile Chain go through? I'm asking as a guy that has 7000miles on a bBright with nary so much as a creak & still pulls splines so hard on cassettes they sometimes jam on Shimano aluminum freehub bodies in less than 3000 miles. (The have to be forcibly removed...& it sucks)

On a related note: Freehub bodies get expensive. Get yourself an Edco Monoblock.

Serious, I don't get it. Is Mr. 800 Mile Chain powering toasters?

DaveSSS 

Whenever the new Connex 12 speed chain becomes available, I'll try it, as a comparison with my Campy Chorus 12 and SRAM AXS 12 chains. The last time I tried a connex chain, it was either 9 or 10 speed and I was not impressed, but the connex link is great. I suspect that their 12 speed chain will be a bit on the wide side and sort of universal, for use with both MTB and road 12 speed, but of course it will not have the proper roller diameter for use with SRAM AXS drivetrains, that should use a #41 chain, not the most common #40 .

rydabent

What I would like to see is the same test done with a chain "lubed" with wax, and one lubed with a high quality oil such as Mobil 1.

Drew Eckhardt 

I use Campagnolo chains. Never clean them beyond wiping off excess lubricant, use Prolink Gold whenever they stop running silently, swap them around 4500 miles because front shifting is slowing down from the flexibility although elongation is just 1/32". I have 25,000 miles on my current cassette, but expect to change it when my current chain wears out.

Oldguyonoldbike

Out of habit I use Campagnolo chains and have never had any complaints. I've used several different levels, Record, Chorus and Veloce. I tend to change them after about 4000 miles and have never had one fail. The drivetrain on the bike I ride most has about 8500 miles on it and is still in very good shape. There may (or may not) be better chains out there, but I will probably always stick with what has always worked for me.

Tomboniboi

Hi Sy, It's a tongue-in-cheek comment.

jadocs

I can see the logic in that.

However if other chains in the test fail before Connex under those conditions (smooth pedaling), would they not fail even sooner under rough pedaling?

63rickert

Well, yes. But no way to know if all would wear or fail proportionally to the lab result. Constructing a lab test that would replicate real world results is not likely. This is why anecdotal reports, as on this forum, are still interesting. And why manufacturers supply race teams and monitor how the equipment survives.

base2 

I was reading in a .pdf regarding Rolhoff hub efficiency, of all things, that in a pedal motion of an average power of 200 watts the load cycles between low value X & high value Y. There is little if any change in efficiency at the low value, and the efficiency (or lack thereof) is magnified at the high load.

As the inefficiency of the drivetrain (hub) is caused by increased friction, that can be presumed to also cause additional wear of components & shortened service life.

So you are correct, essentially. Little to no wear (friction) happens at low loads. Greater wear (friction) happens at higher loads. The trick is finding a means of normalizing the results to make a meaningful comparison.

In their test methodology:
To test a cyclical load of 200 watts average by a typical rider, the analogue methodology would be to test a higher constant load consistant with the peak load applied by the rider. 200 watts being the average between 0 & 400. 300 watts being the average between the average power & the peak power where friction losses & thus efficiency measurements have consistant meaning.

Around 300 watts constant is what they used in their field test to normalize & compare efficiency numbers across the various drivetrains & configurations tested.

They were testing IGH hub efficiency vs derailleur system efficiency, but it's not hard to see how the same math/logic applies to the chain only in isolation of other factors.

base2 

...If I remembered right.
It may be peak torque that a human puts out is equivalent to constant (average) torque of a motor on a testing rig...average power being the value midway between 0 & peak.

In anycase, it's the high peak load that causes the greatest wear.

Nevermind, I'll see if I can go dig up the .pdf now, cause I think I confused myself.

edit: Source of my mis-remembered rambllings. It's the pdf entitled: Efficiency Measurements, Never ending story