Mtracer

Recently my 12-speed Shimano CN-M8100 chain running on a road bike broke with about 4,000 miles on it. One inside plate broke. So, the chain did not completely snap. Just started skipping weird in 11th and 12th gears.

I'm just curious if anyone has any thoughts on this. I'm not a powerful rider, maybe 800 W max sprint effort, and that is rarely. I know how to properly shifts gears, but of course occasional something happens and a shift may be a little clunkier than desired.

Here's more info in an attempt to answer questions before they are asked.

The chain came on new bike and therefore not likely to be counterfeit. When new, I stripped with Silca stripper per instructions and then used Squirt drip wax. No degreasers that would cause the metal to get brittle.

Of course, I understand that 12-speed chains are thinner than 11-speed, which are thinner than 10-speed. But I don't think it is coincidence that it failed at the master link (see pic). Though there's nothing unusual on the inside of the master link, such as burr or similar that would have worn on the broken plate. Plus the crack doesn't follow a curve like I would expect if the master link were wearing on it. No measurable chain "stretch" using a proper measurement tool.

While the chain is dirty, it's just blackened wax. It's not gritty at all. And what looks like a copper color on the inside of the master link is a reflection. That surface is polished very smooth and is still silver in color.

This is my first 12-speed road bike, so I don't have experience to draw on. Is 4,000 miles reasonable for a chain to break? Do these tend to break at master links? Bad luck, or do I need to replace the chain on say a 3,000 mile schedule. Or periodically do an extreme examination in an attempt to see a crack before it becomes a full break.

soyabean

It's manufactured as thin as possible. It's not a 1s bmx chain.

A chain can break anywhere anytime for any or no reason at all.

Crap happens.

Duragrouch

I'm not sure I understand the break. I'm seeing one half of the master link still on the chain, intact, and the other half loose above it, intact. I see part of a broken link on the chain; Where is the other half of that? If that's on the underside of the pin on the loose link, I still don't see how it all went together, and I use master links. I see two completely intact outer plates of the master link.

Beyond that, not surprising to fail at master link, each pin is only press fit (or welded) into one side so will not stay as straight, and bending force on that pin (instead of pure shear) will bend and fatigue the outer plates on the master link, and also adjacent link as shown. Replace link.

EDIT: The thinner side plates of 12s with higher strength material, does not change the lower lateral STIFFNESS of a thinner plate (young's modulus does not change with strength). That is a stiffness related failure, IMO.

Mtracer

The red arrow is pointing right at the break. I think you may not realize that the piece to the left side of the break is the part with some blackened wax on it. It does sort of look like the other inner plate on the far-side of the chain. But it is the other side of the broken plate.

Duragrouch

Got it. See my multiple revisions above.

Mtracer

It certainly seems possible the interface between a master link and standard outer plate to inner plate would be different and might be the source of different stresses.

When you say "replace link" are you suggesting replacing the broken link and continuing to use the chain? That's not something I'm going to do. 4,000 miles is plenty out of a $50 chain. Not worth the risk to me of another failure somewhere else in the chain.

Something else I've wondered is if the plate was stressed when the chain was broken by the bike manufacturer (Trek in this case). I'm sure they get the chain in long spools and break it to length. Versus a retail chain cut to length by Shimano.

Duragrouch

Please see my more recent edit above re: 12s plates.

From breaking off from spool? As my dad would say, "hell I dunno".

Yes, good to replace chain. Just be aware 12s may be less durable in this failure mode, despite the steel being higher grade.

gearbasher

The plate was damaged when the chain was "cut" to length? That link is where a pin would have been pushed out.

grumpus

You beat me to it. I wonder if that was done at the chain factory or by the bike builder.

soyabean

If I was going doing some long distance shiite, then for sure a new waxed chain.

The old chain I would just hoard and use for a repair.

I have friends and family and neighbors that show up with BSO, and that 4000mi chain is still considered an upgrade over their rusted half-seized PYC brand chain.

ScottCommutes

Replacing an entire chain on the side of the road is not a difficult repair. One option, if your type of riding warrants it, would be to carry chain pliers, a spare chain (perhaps already cut to length), and a couple of quick links whenever you get over a couple thousand miles on your installed chain. Alternatively, you could carry a little piece of chain to splice a break.

Personally, I ride 6/7/8 chain. I carried a spare chain all the time for awhile, then quit because I never needed it.

mpetry912 

you got acceptable life out that chain at 4000 miles.

Do you monitor chain wear with the Shimano guage ? a worn chain will wear out the cassette

ONCE in my cycling career I had a chain do this. Still got me home with careful riding.

Stuff does break. I replace chains when they start to show wear. The Shimano chains are exceptionally good.

2500 miles is an acceptable life, at least to me

/markp

Russ Roth

Your peeps have 12sp BSOs? 12sp chains only work reliably on 12 or 11sp bikes and would work like crap on a 7sp bike where it might not even fit over some of the teeth of the FW or chainring. Personally I would never pass off a part that failed to anyone, why it failed really is anyone's guess. I think damage from being removed from the larger spool, based on where this failure is, is a legit reasoning but I wouldn't trust it isn't a batch issue and there's not another failure waiting to happen.

soyabean

My bet is that carrying a spare chain is not necessary if you are religiously checking the chain for 0.50%

Carrying an extra link is sufficient enough, those fall off on their own all the time.

If bike is new, break the chain and put a link in place so you don't have waste weight carrying a chain breaker.

oldbobcat

You said it right there. Anything on a 10-speed chain or thinner after 3000 miles is on borrowed time.

FBinNY 

Just speculating so factor that.

With the understanding that modern chains have plates thinned to the point of having minimal safety margin, we can expect greater instances of breakage owing to variables.

There's always a tolerance in manufacturing precision, alloy formula, heat treating, etc, so there's always a weakest link. With little margin for error, what used to be OK no longer is.

However, I'm not convinced that the failure near the master link is coincidence. 66:1 odds (either side of connector in 122 link chain) argue against that.

So consider that each inner plate carries half the load roughly equally, and can't carry much more.

The dimensional precision of the connector is more complicated than the normal plates, so it's conceivable that one side is a bit off. That would alter the 50/50 load sharing of the adjacent inner plates, taking the harder working one above the fatigue limit, and eventually to failure.
----------
BTW I noticed some nicking of the outer plate in the photo. This is usually caused by shifting under load. Not that it caused the break, but it is hard on chains and not a good thing. If you see much of this type of damage, take a hint and try to time your shifts better.

zandoval 

Could it be constant shifting under load?

Could this a Brifter use technique problem?

FBinNY 

I doubt it. Shifting under load causes nicks and peening of the plates, especially outer ones. It also forces plates outward on pins.

So we see stiff links, and breakage by plates coming off pins. However, I wouldn't associate it with an inner plate failure like we have here. (which, BTW is the 1st I've ever seen).

Mtracer

Well, when I first had odd behavior in 11th and 12th gears, it was about 5 miles into a ~70-mile ride. I'm sure the skipping was caused by a fully broken inner plate. Point being, I put 60+ miles on the chain with that broken plate. While I wasn't sprinting, if the plates were that marginal, I would think as soon as one plate broke the other would follow.

But no question 12-speed chains have less material then 11-speed which have less than 10-speed. So, all other things being equal, they can't hold up as well. Of course, all other things may not be equal.

I also agree that I don't think breaking at the master link is a coincidence. I lean towards stress when the chain was "broken" for sizing when the bike was built. I assume the bike builder is working pretty fast and pulling chain to length with some powered machine that quickly pushes the pin out of the chain. If that builder were perhaps a little too quick to pull the chain free and twisted it a little with some force, maybe they stressed it. Or maybe the machine needs adjusting. No way to know.

Another possibility is stress when I've snapped the master link on and off. I know I did this at least twice. First to take the chain off to strip it when new and another time when I swapped out the crankset. I use the Park Tool Master Link Tool for this. But I don't think opening the master link would stress the inner plates at all as you're squeezing the master link itself. Of course the pliers do apply tension to snap the master link back into place, but I don't think this is anywhere close to the max forces seen riding the bike. But still maybe something about the "snap" itself is a transient stress.

As I mentioned in my original post, the chain has no measurable "stretch". I use a Pedro's Chain Checker.

Yes stuff does break. And in this case I was mostly curious to see if this type of break is common in 12-speed chains. It doesn't sound like it is. But I know I've seen some post online and it does seem that 12-speed chains have more issues.

While I can't complain about 4,000 miles out of a chain, I'd expect wear rather than breaking.

As I mentioned in my original post, I know how to shift gears. But of course, we all have a bad shift once in a while. I estimate ~50,000 shifts over the 4,000 miles on the chain. A few of those are going to be bad.

But I suppose one bad shift in just the right way may be all it takes.

FBinNY 

About the only thing that's certain is that the interval between a single plate failing and the chain snapping is very short.

The chain simply cannot handle loads with a broken plate. So, while it's barely possible you rode a while if you were gentle and there were no hills, I suspect the link broke much closer to when you saw it.

Possibly it was ready, and you finished it off with side load when trying to open the connector.

Andrew R Stewart 

No expert on materials and their manufacturing but I'll toss in my opinion too.

I can't remember the last time I saw a broken inside plate, likely I have but so infrequently they didn't even register in my long term memory (and odd stuff usually does peak my interest, like this post). My guess is that the connecting link had nothing to do with this. So too with any cutting of the chain by whoever did it at this link. For the reasons that the pins are not a press fit within the inner plate and that this crack isn't even at the pin hole anyways. The pin interference fit (with the outer plate) can cause radial cracks at the hole, and there's a history of chains being improperly heat treated during manufacturing, helping to cause these cracks. But without any stress from the pin acting on the inner plates I don't see how the pin can cause this.

What I do suspect is the manufacturing of this chain/inner plate, was "off" and it finally gave out. After what many might consider enough miles to consider the chain having "earned" its keep. I doubt that without a serious and expensive metallurgical assessment we won't really know what caused this. But thankfully the fix/chains are not expensive, even when the cassette needs replacement too.

My suggestion is to replace the chain (and cassette if that's worn beyond problem free use too) and enjoy the ride. Andy

Mtracer

I'm 100% certain the plate was broken early in that ~70-mile ride if not before the ride. The reason is, the symptom was a sort of chain skipping clunking happening in 11th and 12th gear. I thought perhaps the derailleur hanger got bent somehow. I even stopped to make sure the derailleur simply hadn't come loose. It's a Di2 system, so a cabling problem couldn't explain the issue. It wasn't until I got home and on the repair stand I was able to immediately see the chain lift and then fall coming off the 11 th and 12th gears. The reason for this is the larger part of the broken plate had rotated such that it was interfering with the smaller cogs and causing the chain to lift up a small amount. Then of course, once this broken link cleared the cog, the chain fell back down into proper position on the cog. This made for the skipping feeling and clunk sound.

I think it is likely the plate broke a week or so earlier. I developed a slight sound in the rear as if the rear derailleur needed adjustment. Being Di2 this is not expected. But a single indexing step eliminated the noise. So, I believe the chain was broken at this point causing it to track differently on the cassette. It wasn't until the broken plate rotated that the more obvious problem developed.

Fortunately I wasn't doing any max efforts out of the saddle, as perhaps the chain would have snapped and I would have fell.

However, I don't think this says anything about how marginal the chain may be. That 70-mile ride was specifically a zone 2 ride, so it was very much on the low-stress side of what a chain will see. I think I can say with certainty that this single plate, carrying the full chain tension, saw less stress in this time period, than it has at other times when both plates were sharing the stress. Also, my max power is less than half what a powerful rider could do. So, even a single plate wouldn't likely just fail at the power I'm able to produce. Of course, once broken, the remaining plate sees a range of different forces not just that it carries all of the chain forces by itself.

If nothing else, in the future, I'll be more wary of a chain issue if anything seems off. While I don't know that I would have spotted this break easily. It's still worth trying to spot this type of thing in hopes of catching it sooner rather than latter. If nothing else, it could be worth examining the inner plates at the master link position periodically. And I may just be extra cautious and replace the chain every 3,000 miles.

FBinNY 

This is a freak kind of thing. While those of who've been around long enough have seen a few cracked plates, almost exclusively outer, it's not a common thing.

So, other than standard care in maintenance and lubrication, nothing special is called for. Odds are you won't see another cracked in your lifetime, and your concern will be about wear, not breakage.

Though we can't know the timeline you might conduct an experiment with this chain before chucking it. I'd love to know because inner plate failure is so rare that there's not much known about how much load it can manage.

Break a plate, or push a rivet through slightly, so it's not connected at that place. Add some load by either yanking on it, or putting it back on the bike with the broken link on the upper loop, and stepping on a pedal.

We know that a broken outer plate the load is so imbalanced that it immediately pulls apart. However inner plates are close to the center line, so the bending force is much less, which explains how survived that long. (or it might not)

LV2TNDM

Every time the bike industry adds another cog, we hear the same old fears, "Man, these chains are REALLY THIN! I don't trust 'em!" (And I admit, I wonder as well.)

But then you find that 12spd. are regularly used on tandems without issue.* A single rider will never achieve the pedaling force two riders impart on a tandem main chain (and everything else in the drive train). Double the power is one thing, but the clincher is ZERO rear wheel traction loss or front end lift. This is what makes mountain tandems rear hub destroyers. I worry about blowing up my seventh hub. I've attempted Slickrock twice on the tandem and have a zero success rate. First hub failure was, not surprisingly, a DT/Swiss from '97. But the second was splitting a Phil Wood tandem FSC hub in half. And subsequent to that, we've mangled two more sets of PW tandem hub internals. Still on my "to-do" list, but fear I may need to fit a King or Rohloff to avoid my third catastrophic failure on Slickrock. But it pisses me off that I still haven't checked that off my list after 30 years!

So I have yet to encounter epidemic chain failures because they've "made 'em too thin." Despite the hub failures, in 30 years, we've never suffered a main chain failure. Our ONLY chain issue was while touring on the mountain tandem when a rock was kicked up into and breaking the timing chain. Even e-bikes may not be able to achieve tandem torque loads because of wheel traction loss and front wheel lift. But I may be wrong; instantaneous torque electric motors produce might impart sudden high torque loads before the side-effects occur. The ultimate hub killer? An e-mountain tandem. Ay, ye poor freehubs! This is a pretty strong argument for a gearbox.

*Apt analogy. Whenever someone sees a bicycle equipped with S&S couplers, they oftentimes ask, "Wow, are those things strong enough?" Well, they use them on tandems, triples, quads and more. Yup, they're effing strong!

FBinNY 

I don't know why you opted to quote me, except maybe to set up a strawman to argue against. Nobody, and certainly not me, said that modern chains are weak or otherwise not up to the task. Doubly so here, since in the last post before yours I mentioned that breakage is a freak event, and the OP needs or focus his worries on wear rather than breakage.

That said, chains are in fact getting weaker (thought still up to the task). When I started in the bike industry 1400kgf was the accepted standard for chain strength. Later that was 1200kgf, and most recently I see references to 800kgf. That's still much higher than what's needed, but if I may offer an analogy. It's like visiting the grand canyon and walking to some of the lookout vantage points. No problem for the sure footed, but as you get closer to the edge, the risk of falling off if you should slip or trip increases. Doesn't mean you can't or shouldn't venture close, just that you need to be more careful.