Rowan

Sheldon does say that this is a very badly worn chain in his illustration, and I guess it is feasible that the roller has worn away the shoulders on the sideplates and started in on the rivet as well.

The idea of reversing a worn chain is simply answered: The pitch of the chain is what counts, and it doesn't matter which way you stretch it out (left to right, or right to left) and which way you measure it (left to right, or right to left), the distance of 12 links is still going to exceed the desired parameters (1/16 or 1/8th of an inch). Compressing isn't an option and remember, too, it is the pseudo-bushings that have worn most, and it can be expected that there will be wear on all sides, as the chain rotates through 360 degrees in its circular path (shown most simple, of course, on a single-speed or fixed gear).

TallRider

I think this is worth re-mentioning:

So mox, keep us posted on what you find. If rollers are in good shape on an older and worn chain, then I end up with a lot more trust in the Park chain-measuring tool.

simplify

Maybe there's no reason that seems intuitive, but the rollers DO rotate against the pseudo-bushing! I just put my bike on the stand, and watched those rollers relative to the *inside* plates, and as they come up, over and off the cassette. The rollers stay stationary against the cog tooth and the pseudo-bushing turns inside them! I stand by my hypothesis that the rollers do experience friction against the pseudo-bushing, and that the inner surface of the roller causes wear on the outer surface of the pseudo-bushing. Again, this contributes to slop in the rollers, but does not contribute to chain lengthening.

TallRider

Good point. Makes sense. So now we're all waiting to here moxfyre's experimental analysis...
I wish I still had some of my old, absolutely trashed, SRAM chains around, that I'd used on my commuter bike.

moxfyre

Okay, so I took apart 3 old chains, and I was disappointed to find that only one had enough wear to be visible to the naked eye: it was just shy of 12-3/32". I had trouble getting good up-close shots with my camera's macro mode, so apologies for the crappy images. (On the other hand, ain't my washing machine just the sexiest thign you've ever seen )

All the photos are in https://bike.toleressea.net/oldchain. Here are a few of the better ones:
https://myxo.css.msu.edu/danimal/bike...n/PA060003.JPG
https://myxo.css.msu.edu/danimal/bike...n/PA060009.JPG
https://myxo.css.msu.edu/danimal/bike...n/PA060010.JPG

My thoughts:

• The worn pins are CLEARLY worn in two separate sections centered around 2/5 of the way from the ends of the pin. So, looks like pseudo-bushings are the culprits.
• The roller has a circular raised ridge around its inside. I would estimate it to be uniformly 0.010"-0.020" high. This ridge is possibly an artifact of the forging process? In any case, this ridge has no visible wear on it, which suggests that the interior of the roller does not get worn down much.
• I could not get very good photos of the pseudo-bushings, sorry, but from a visual inspection the OUTSIDE surfaces appear to have little wear on them. Even squinting at them very hard, I couldn't get much idea of the condition of the inner surfaces (which would contact the pins).

So... looks like the pins have borne the brunt of the wear on this particular chain (a 6/7 speed chain, has a stylized "UG" and the word "TAIWAN" stamped on every link). No significant wear on the outside of the pseudo-bushings, or on the interior of the rollers. Any thoughts? Any suggestions on how to take better photos? My camera has a manual mode but I'm pretty rusty on my photography skills.

Rowan

The wear on the pin is from the shoulders or pseudo bushings rotating on them as the chain moves around the cogs, chainring and pulleys. It's natural that the most wear occurs there because that is where most of the drive forces when pedalling are applied. In addition, because most of the movement under pressure occurs in that region, the addition of fine dirt particles accelerates the wear compared with, say, the rollers. This is not to say the rollers or sideplate shoulders don't wear -- they do, but at a different rate. At least, that would be my assessment, but it's open to discussion.

simplify

NO! I think that ridge is actually *evidence* of wear! The pseudo-bushings come together right at the middle of the roller (and pin), but they still have a tiny gap--that's why you can see a similar ridge between the two heavily-worn areas on the pin. That's where the "seam", or gap, in the pseudo-bushing is. So the roller is also wearing away inside, and the ridge is where it contacts the gap between the two halves of the pseudo-bushings. The ridge is what's left, when the interior surface of the roller is worn away on either side of it.

Good work, BTW, and yes, very sexy washing machine!

Carusoswi

I think we need chains made of some other substance - or, perhaps chains with rollers made of some sacrificial material that will indicate wear before damaging the cogs. I thought that I had done something wrong when, after two chains and 4500 miles, I had to replace my cassette along with the chain. Based upon what I read here, that sounds almost normal.

I think mfr's should come up with some better system - different materials (don't have a clue as to what sorts of changes/materials would be available).

Caruso

TallRider

Well, the key is whether the worn roller (with the ridge) is noticably thinner than a roller from a little-worn chain (and whether the little-worn roller has the same ridge). You can't get this in photos with your camera's macro mode, but you can compare the rollers and stack them on top of each other and get a pretty good sense of this.

I must admit that I, too, felt the draw of your washing machine.

HardyWeinberg

so do we think a chain wear gauge is worth it?

simplify

Well, based on what we know so far, I think the most likely scenario is that the chain wear gauge would register more wear that is actually there, because it would mistake the roller slop for lengthened pitch. If you use a ruler, you're really measuring pitch--and only pitch. If you use a gauge, you're measuring both pitch and slop, and it seems that slop in the rollers is not much of a factor.

That said, it's not a bad thing that gauges show perhaps more wear than is actually there, because you'll change your chain a bit sooner.

HardyWeinberg

Beautiful, thank you!

Rowan

Or you could say to hell with chaincheckers and rulers, and ride the whole lot into the ground and put on a new (upgraded) groupset after... ohhhh... say.... 15,000km

Sawtooth

Thanks for the great discussion guys (I am a ruler guy, myself). The above is exactly what I ended up perscribing for a Bianchi I was working on for a friend over the weekend. The chain was well over 1/8 inches stretched and by that point, the new chain I put on slipped with every single stroke. My advice was to wear his system into the ground and replace entirely when the occasional slipping became too much.

Again, great thread!!!

Rowan

I am unsure where you are coming from here. Wear depends on lubrication -- if there is a layer of oil between the moving parts, wear will be minimised. If the parts are allowed to rub against each other, the wear will be accelerated, and usually the softer metal will wear away most rapidly.

Maybe your eye sees "dependent" but your mind sees "independent"?

operator

Uh. That's exactly what he said.

d_D

Or buy a few chains and rotate them. This brings the wear rate of the chain closer to the wear rate of the sprockets so measuring isn't needed. Plus you have lots of time to clean and lube the dirty chains before they are neeeded again.

HardyWeinberg

I've been piling the lube onto the chain of my (commuter) bike lately, just to keep it spinning through the rust it's picking up from all the rain (20" in Nov), so it's naturally drawing that much more road grit into a nasty grinding paste.

I tried one of these chain checkers, it repeatably claims that I've got 0.75% stretch. A 12" metal ruler, however, shows very clearly pin-to-pin 12" of length, suggesting *no* chain stretch (the chain has maybe 1k miles, but the last few hundred have been very hard miles). I didn't take the chain off to measure it, just turned the crank forward to put some tension on it.

My technique is probably subpar w/ one or both measures, but it is interesting to me that I am at least getting consistent results with both techniques. Regardless of stretch, I'm going to just chuck the chain once the monsoon season passes (maybe Feb-ish), but I think for right now my mind is somewhat eased of fears my cassette will be ground to powder if I don't swap out the chain THIS WEEK.

humble_biker

Nice thread! Hey did you know that the Concorde jet grew an extra 6" when at full speed?