Bicycle Mechanics - Failure of unstretched chain?

So this morning's commute was fun - as I start pedaling after an intersection, the chain breaks. Yay! Chain breakage is one of those things I'd like to avoid as I enjoy my teeth where they are. Thankfully this time it broke as soon as I put my foot down, so no harm done.

I'd been checking for stretch quite often, but the thing was (and is still) under 1/16" of stretch over 12 links. But still it failed. The failure was at the connector pin on a 7-speed Shimano Hyperglide chain. The lip on the pin wore down and the plate pulled away. The chain had never been broken since installation.

Is there anything one can check for other than stretch? For what it's worth, the chain in question was original on a bike I've owned for 16 years, but has only seen about 3000 miles. So certainly it was rather old. But my impression was that years don't matter if the chain checks out OK. I've had a replacement chain at home waiting, but it seemed wasteful to toss a functioning chain.

So is there a sane limit beyond which one should toss a chain for sanity's sake, even if the tape measure says it's OK? Looking back on it, I was an idiot not to just throw on a new chain, but I was raised not to waste anything. In any event, I'm still intrigued why it didn't show virtually any wear.

"The lip on the pin wore down and the plate pulled away"

This is more or less the definition of chain "stretch" (see pictures in attached -- http://www.sheldonbrown.com/chains.html) so I am at a loss to explain why the measurment wouldn't show wear.

A seven speed chain is much thicker in terms of plates and pins than a 9 or 10 speed or even an 8. The chain can get pretty worn around the pins before it "stretchs" that much.

Chain elongation, or stretch is a very small amount of wear on the pins and bushing formed into the inner plates, multiplied by 100+ links. It has NOTHING to do with chain failure. Most chains come apart when an outer plate pulls off a pin. There are many reasons for that to happen, including improper installation.

Use a Master Link, however, re-use is sometimes not recomended.
The older Shimano systems did not recomend their use but the new 10s chains are supposed to come with one.
KMC is another chain supplier you may want to consider.

"It has NOTHING to do with chain failure."

Of course it has to do with failure. Its not the only cause or probably even the most common one; but chain elongation can certainly contribute. It is a symptom of pins and bushings being worn away (i.e. pins get smaller and bushing openings get bigger) just as the OP described his pin being worn away; as that happens, pins and plates are more likely to work themselves free from eachother; particularly if , as in this case, the chain is old so is stiffer generally than a newer chain. True, lots of thigs can contribute to chain failing, one of those things is chain wear. So in this case I would surmise that wear of the pin combined with other factors stemming from the chain's age combined to cause the failure.

So this morning's commute was fun - as I start pedaling after an intersection, the chain breaks. Yay! Chain breakage is one of those things I'd like to avoid as I enjoy my teeth where they are. Thankfully this time it broke as soon as I put my foot down, so no harm done.

I'd been checking for stretch quite often, but the thing was (and is still) under 1/16" of stretch over 12 links. But still it failed. The failure was at the connector pin on a 7-speed Shimano Hyperglide chain. The lip on the pin wore down and the plate pulled away. The chain had never been broken since installation.

I would vote on either a defective, too soft, connector pin or the connector pin being originally misinstalled. I would presume when you say connector pin you are referring to a Shimano replacement pin as used to reassemble a shortened Shimano chain.

I would vote on either a defective, too soft, connector pin or the connector pin being originally misinstalled. I would presume when you say connector pin you are referring to a Shimano replacement pin as used to reassemble a shortened Shimano chain.

In this case, the chain is as originally assembled new by the bike shop (in 1993!), but it was indeed the pin used to connect the shortened chain. It definitely seemed suspicious to me that it happened to break on the connector pin, since it's what, one pin in about 50?.

Any idea on how often chain failures happen on the connector pin? Is it a usual weak spot if installed correctly, or just if incorrectly?

Incidentally, how does one know if the connector pin is installed just right? Last time I installed one on another bike, I just pushed it through until it felt even on both sides and the link moved freely.

Thanks everyone for the input. I'm just generally curious about it now. I put a lot more miles on my bikes now than I did as a kid, so there's no chance of a chain having such low miles at 16 years again, but it's still interesting.

In this case, the chain is as originally assembled new by the bike shop (in 1993!), but it was indeed the pin used to connect the shortened chain. It definitely seemed suspicious to me that it happened to break on the connector pin, since it's what, one pin in about 50?.

Any idea on how often chain failures happen on the connector pin? Is it a usual weak spot if installed correctly, or just if incorrectly?

Incidentally, how does one know if the connector pin is installed just right? Last time I installed one on another bike, I just pushed it through until it felt even on both sides and the link moved freely.

Thanks everyone for the input. I'm just generally curious about it now. I put a lot more miles on my bikes now than I did as a kid, so there's no chance of a chain having such low miles at 16 years again, but it's still interesting.

Dude.

16 years.

Shimanos "special rivet" for attaching the chain is ridicuously stupid and ********. Every other chain manufacturer has gone to a quicklink type of connection because it is 30 billion times more foolproof than pushing a "special rivet" through.

The reason why your chain failed is two reasons.

1) Chain wear
2) Improper/faulty installation

Even if you have a shimano chain, most of the time you can fit an sram powerlink on it. If your current chain doesn't suffer from extreme wear (this metric depends on who you ask, sheldon brown defines it as 1/16 for chain only replacement and 1/8 for chain + cogs) then replace the faulty link with a sram masterlink. Should probably run you about $5 at your favourite LBS.

#2 usually happens quite quickly, espeically in high torque situations but I can imagine a situation where pin installation was borderline leading to premature failure sometimes (or much later in your case) down the road. Screw shimano and their stupid chain connections, sram produces a superior and much more foolproof chain at equivlalent cost. (e.g PC48 7/8 speed chain = < $5 wholesale, x2 retail).

I've had the same thing almost happen to me, I noticed the change in chain tension and stopped to look before it broke off though.
Same thing about a plate pulling away from a pin.
Thinking it through, I concluded that the problem was caused a few days earlier when I dropped the chain off a sprocket and had it jam into my bottom bracket. At that time I didnt notice but the chain must have been subjected to sideways bending that started working the pin out.

Dude.

16 years.


I know. It's beautiful, isn't it? Completely nuts. At some level I was so impressed by the stupidity of it I almost wanted to see how long it would go.

In fairness, I did post a thread a few months ago asking if there was any other sign that a chain needed to be replaced other than 1/16" of stretch. General consensus was, that was probably good enough. Though I did probably fail to mention the "16 year old chain" tidbit, mayhap the advice would have been a tad different if I had!



Shimanos "special rivet" for attaching the chain is ridicuously stupid and ********. Every other chain manufacturer has gone to a quicklink type of connection because it is 30 billion times more foolproof than pushing a "special rivet" through.


I think I'm coming around to that line of reasoning.



The reason why your chain failed is two reasons.

1) Chain wear
2) Improper/faulty installation

#2 usually happens quite quickly, espeically in high torque situations but I can imagine a situation where pin installation was borderline leading to premature failure sometimes (or much later in your case) down the road. Screw shimano and their stupid chain connections, sram produces a superior and much more foolproof chain at equivlalent cost. (e.g PC48 7/8 speed chain = < $5 wholesale, x2 retail).

I might be obsessing unnecessarily on this, but I cleaned the chain and looked at the inside of the plates. It seemed to support exactly what you stated - the wear wasn't even on the inside/outside plate and pin, and my guess is it was installed just a little off. Not enough to cause it to fail during, say, the first decade of use, but a tad later.

The hilarious bit is, I put an order in to Jenson last Sunday for parts to do a drivetrain upgrade to 9-speed. So if this thing would have lasted another week on top of its 16 years, it could have had a proper burial. For what it's worth, I went with an SRAM chain.

Now, don't get me started on the original factory tubes I still have from 1993. They're still intact...but I did replace them last year just to be on the safe side.

How does one determine which link was had the original connector pin after 16 years and 3000 miles of use? And why the f*ck would anyone care?

How does one determine which link was had the original connector pin after 16 years and 3000 miles of use? And why the f*ck would anyone care?
The 16 years is immaterial unless the chain was stored badly and corroded. 3000 miles isn't much service. The original connector pin is easy to find by inspection as it has a different appearance from the other pins.


Shimanos "special rivet" for attaching the chain is ridicuously stupid and ********. Every other chain manufacturer has gone to a quicklink type of connection because it is 30 billion times more foolproof than pushing a "special rivet" through.
Not every other manufacturer. Campy (you may have heard of them) also uses a connector with special pins to join their chains.

Installed properly, the special pins make a reliable connection and I've had many chains last 5000-8000 miles using them and never had a failure.

Step one - throw Shimano Chain in trash
Step two - install SRAM Chain
Step three - ride bike for long time between chain replacement.

"It has NOTHING to do with chain failure."

Of course it has to do with failure. Its not the only cause or probably even the most common one; but chain elongation can certainly contribute. It is a symptom of pins and bushings being worn away (i.e. pins get smaller and bushing openings get bigger) just as the OP described his pin being worn away; as that happens, pins and plates are more likely to work themselves free from eachother; particularly if , as in this case, the chain is old so is stiffer generally than a newer chain. True, lots of thigs can contribute to chain failing, one of those things is chain wear. So in this case I would surmise that wear of the pin combined with other factors stemming from the chain's age combined to cause the failure.

I just have to disagree. The only wear that causes elongation is between the pin the inner plates. It has nothing to do with the outer plates. When a chain is elongated by .5%, that wear is a very small .0025 inch, divided between the pin and the inner plate bushing. It is certainly not enough to weaken any of the parts.

Older chains are of the protruding pin type that even less prone to coming apart, providing that the person who joined the chained inserted the joining pin properly. In the days of that chain, special joining pins were not used and not needed. Pins were typically pushed out of one outer plate and left partially inserted in the other. Upon rejoing, the pin was merely pushed back in. Protruding pin chains rely mostly on the press fit of the pin and the protrusion to keep them together, unlike modern flush-pin chains that rely on heaving peening at the ends of the pins.

The 16 years is immaterial unless the chain was stored badly and corroded. 3000 miles isn't much service. The original connector pin is easy to find by inspection as it has a different appearance from the other pins.

The bike spent some years in environments that were moist (basements), though not exposed to moisture. It never got badly rusted. Also, it was the pin that failed, not the plates, which would lead me to believe that corrosion failure wouldn't play a significant role - no? For what it's worth, the rest of the chain seems to be in worn but respectable shape. So based on what I'm hearing, I'm inclined to think that the failure was bad installation + 3000 miles, not 16 years + 3000 miles.

I'm now quite interested in how one properly installs the connector pin, since I have two bikes that currently utilize them (one of which I recently installed myself). It looks like the original pin on this bike was a bit off from examining the failure, in which the outer ridge on the pin wore away due to excessive contact with the plate. How does one know if the pin is lined up just right?

Thanks, folks!

The bike spent some years in environments that were moist (basements), though not exposed to moisture. It never got badly rusted. Also, it was the pin that failed, not the plates, which would lead me to believe that corrosion failure wouldn't play a significant role - no? For what it's worth, the rest of the chain seems to be in worn but respectable shape. So based on what I'm hearing, I'm inclined to think that the failure was bad installation + 3000 miles, not 16 years + 3000 miles.

I'm now quite interested in how one properly installs the connector pin, since I have two bikes that currently utilize them (one of which I recently installed myself). It looks like the original pin on this bike was a bit off from examining the failure, in which the outer ridge on the pin wore away due to excessive contact with the plate. How does one know if the pin is lined up just right?

Thanks, folks!

When I buy a chain I do not care who made it - but I always use Sram or KMC quicklinks to reattach - they have always worked fine with Shimano or whatever chain I get, as long as they are designed for the right number of speeds.

I just have to disagree. The only wear that causes elongation is between the pin the inner plates. It has nothing to do with the outer plates. When a chain is elongated by .5%, that wear is a very small .0025 inch, divided between the pin and the inner plate bushing. It is certainly not enough to weaken any of the parts.


Interesting. So it sounds like the 1/16" guideline is a fair estimate of when you should replace the chain to avoid damage to components (stretch from inner plate wear on pin), but not catastrophic failure (from detachment of pin from outer plate).

So how does one prevent or diagnose oncoming failure?

Step one - throw Shimano Chain in trash
Step two - install SRAM Chain
Step three - ride bike for long time between chain replacement.

LOL- this is the exact advice my father gave me when I first started biking regularly. Haven't used a shimano chain since.(And haven't had chain failure since)

Interesting. So it sounds like the 1/16" guideline is a fair estimate of when you should replace the chain to avoid damage to components (stretch from inner plate wear on pin), but not catastrophic failure (from detachment of pin from outer plate).

So how does one prevent or diagnose oncoming failure?


See above post. To prevent failure, I use SRAM. I broke every shimano chain I used for three seasons. When I switched to SRAM, no more issue. I now watch for chain "stretch" and replace after two seasons of racing whether it needs it or not. I get more life out of cassettes and have never had a chain failure since.

You are WAY over-thinking this. Decent midrange chains can be had for $30-$40. That is $2.50 per year if your last chain can be used as a guide. Just give up one Latte per year and you're all set.

+1

use powerlinks.

Ill never touch another shimano chain. those pins are straight up stoo-pid when there are quick/powerlinks

I broke every shimano chain I used for three seasons.
You really need to work on your installation technique. I've ridden over 130,000 miles and about 120,000 of it has been on Shimano 8-speed and 9-spreed chains using Shimano's supplied installation pin. I change chains at 5000-6000 mile intervals and once ran one for 11,00 miles. Also, I ride in a lot of hills so I'm not easy on chains either.

In all that time and distance, I NEVER broke a Shimano chain.

[QUOTE=operator;8385936]Dude.

Shimanos "special rivet" for attaching the chain is ridicuously stupid and ********. Every other chain manufacturer has gone to a quicklink type of connection because it is 30 billion times more foolproof than pushing a "special rivet" through.

Sorry but I disagree.
My wife and I ride Shimano 9-speed and Campagnolo 10-speed systems and have never used a quick link. We always get 5,000 to 7,000 miles out of a chain. When Campy introduced the first 10-speed chain it came with a quick link that proved to be problematic so they discontinued it within a few months.

[QUOTE=operator;8385936]Dude.

Shimanos "special rivet" for attaching the chain is ridicuously stupid and ********. Every other chain manufacturer has gone to a quicklink type of connection because it is 30 billion times more foolproof than pushing a "special rivet" through.

Sorry but I disagree.
My wife and I ride Shimano 9-speed and Campagnolo 10-speed systems and have never used a quick link. We always get 5,000 to 7,000 miles out of a chain. When Campy introduced the first 10-speed chain it came with a quick link that proved to be problematic so they discontinued it within a few months.

I'm not sure how you can refute that chain installation using quicklink is more foolproof than pressing rivets, but ok.

[QUOTE=Al1943;8391595]

I'm not sure how you can refute that chain installation using quicklink is more foolproof than pressing rivets, but ok.

Actually I was disagreeing with your statement that every manufacturer other than Shimano uses a quick link. Campy does not. I'm not saying that a person should not use a quick link, but I am saying that the two largest producers of components do not.

You really need to work on your installation technique. I've ridden over 130,000 miles and about 120,000 of it has been on Shimano 8-speed and 9-spreed chains using Shimano's supplied installation pin. I change chains at 5000-6000 mile intervals and once ran one for 11,00 miles. Also, I ride in a lot of hills so I'm not easy on chains either.

In all that time and distance, I NEVER broke a Shimano chain.

The chains did not break at the installation pins. Perhaps it was just a coincidence or I went through a bad batch of material but they were all nine speed Shimano chains. Granted that I am tough on this bike as it is used for MTB racing but the SRAM chains I have used since live in the same environment and have been far superior in longevity. I know many people who swear by Shimano chains but my experience with them has not been the same. I just buy SRAM now for all the bikes.