bobn

There have been bicycle recalls due to faulty chains breaking. Is it possible that these chain brands pictured in this thread could be the same chains being used on the recalled bikes?
Check Google for "recalled bike chains"

joejack951

I'm not. deermouse apparently is though. I have zero concerns with KMC chains due to this incident.

I think the summary of the story is:

Let's start over because you seem to be misunderstanding the concept of rotating chains.

My cassette is an Ultegra 6500, cost anywhere from $70-85 depending on sales. Not uber-expensive, not throwaway either in my opinion. KMC X9 chains can be had for $12-15 (https://www.speedgoat.com/Catalog.asp...C104&Prod=7712).

I settled on 4000 miles as my initial chain rotation period based on measuring chain wear during that period. At 4000 miles I was finally able to measure some wear (1/32"). I found the same wear profile in the next two chains. At that point I had 12,000 miles on the cassette and no signs of slipping or poor shifting. Assuming that my original chain was not destroyed by my misunderstanding of the contents of GOMS, I could have reinstalled it, used it another few thousand miles, then done the same with the other two chains. At that point, I likely would have discarded the chains and cassette so as not to induce too much wear on my chainrings. I figured ~24,000 miles for an outlay of at most $130 is pretty good. Unfortunately, I'll only get to ~16,000 miles now because I only have one chain left and new chains skip on my cassette. Unless I break in another chain on a different cassette, when my current chain is toast so is my cassette.

What are you suggesting I do differently, other than perhaps not be so frugal?

joejack951

In addition, those steel links are under constant stress from the peened/pressed-in pins. A little corrosion between the pin and link adding some undesired bulk between the press fit surfaces could be enough to split the link, or simply an overall weakening of the material leading to the ever-present stress causing a crack. This picture (posted before) should put to rest any claims that these cracks are related to the chain's use rather than it's storage. There's no way this crack is the result is cycling-related stress:

joejack951

Nope, of the bikes recalled where no chain brand is mentioned, none of the bikes are 9 speed equipped. The only brand chain I've seen mentioned is a 10 speed SRAM. I didn't dig much further than Google page 3 though.

Remember, this chain did not fail in use. It only failed when stored for a long time in an obviously (in hindsight) incompatible solution. This is my fault, not the manufacturers.

Sixty Fiver

You need to take two of these chains and inspect each one extremely closely to prove that neither has any signs of damage... soak one for 8 months and ride the other one and see what happens.

Also, check and see if the chain you are using has been subject to recall.

Still not convinced this is a chemical issue or that chemicals are the sole reason for the failure.

deermouse

I don't expect anything, but I do think that 4,000 miles is very reasonable mileage to get from a chain. It's pretty clear that 3 chains is not a large enough group to be statisticlly segnificant. I just find the whole situation to be very strange and think there may be more going on than you are seeing. I'm reluctant to jump to conclusions.

joejack951

Here's a summary of what's been done:

Chain 1, KMC X9: used 4000 miles (Sept. '09 - April '10). soaked for 16 months. Result: cracked quick link and chain links
Chain 2, KMC X9: used 4000 miles (May '10 - Dec. '10) . soaked 8 months. Result: cracked quick link and chain links
Chain 3, KMC X9: used 4800 miles (Jan '11 - present). no long term GOMS soaking. Result: no cracks

Nope, Google doesn't turn up anything for a recall on the X9 chains.

I've considered mounting one of the cracked chains and seeing how much (or how little) pedal pressure it takes to split the links fully. Would that convince you that the chain failed during storage and not during use?

The bit about the Statue of Liberty corrosion problem in the link below is similar to what I think could have caused the chain links to fracture like they did.

https://labspace.open.ac.uk/mod/resou....php?id=391137

joejack951

You already told me to change chain brands based on this incident. To me, that's jumping to conclusions. If any of these failures had happened during riding, I might agree with you. But they didn't and I'm as certain as one can be that none of this damage existed before soaking the chains. The corrosion on the links, especially the second quick link I posted, shows that something corrosive was happening during the soak. The quick link was not corroded like that before I soaked it. As far as I'm concerned, all signs are pointing towards the long term soak in GOMS as the source of these cracks:

1. no chain failure during use (3 KMC X9 chains)
2. no visible signs of cracking after 4000 miles of use (3 KMC X9 chains)
3. no other existing reports of KMC X9 chains failing during use
4. chemical composition of GOMS known to cause corrosion to chains
5. visible corrosion on chains that cracked during soaking

deermouse

I suggested that you consider changing chains. I also said that 4,000 miles was better than most people get from a chain and provided a link to support that. Clearly your chains have a lot of damage, more than I would say a chain could have and still be at all usable. I'm just not convinced that a caustic solvent soak did all that. If you want to prove the point I think you would need to reproduce the "experiment" using chain samples with varying degrees of wear. Maybe a new piece of chain, and ones with 1,000, 2,000, 3,000 and 4,000 miles of wear. If after soaking for 8 months all samples show cracks, I think you can safely say that the solvent caused the cracks. On the other hand if only the high mileage samples show cracks I would have to conclude that either the cracks were already there or the samples were close to failure before soaking and the solution attacked the weak areas.

I'm not trying to be argumentative and I agree that long term soaking in a caustic solvent is more damaging than you would want to subject your chain to. I just don't believe the solvent did all that damage. If the chains were near failure when you put them in the solvent, having the solvent destroy them could have saved you from a broken chain on the road. The fact that you got an extra 1,000 miles on a chain that wasn't soaked could be luck or a quality control problem in the manufacturing process or because the chain wasn't exposed to the solvent (or maybe road salt) or maybe something else. Show me some more evidence and I'll believe it.

p2templin

Cry more.
Buy more (chains and cassettes).
Post less.
Expect chains to last less, or at least get over it quicker when they don't.
Expect cassettes to last less, or at least get over it quicker when they don't.

Life goes on. Buy a new cassette and a new chain, and ride.

jim p

I would not mount the chain to see if it would break but I might put some weight on it to see how much weight that it would hold.

When I first saw this post, I thought that there was no way that a chemical could cause a crack like that in metal without at least damaging the metal in other places. Now I am not so sure that it is not a chemical crack. I remember of some industrial piping that was developing cracks and over lay welding was used to strengthen the pipes. I always assumed that these cracks were caused by vibration and thermal expansion but maybe it was the chemicals in the fluid that was causing the cracks.

If you find the root cause of the cracking to be the chemicals, that will be some good information. Makes me wonder what other chemicals or cleaners may be doing to other products.

joejack951

I didn't plan on riding the bike. I figured I could snap it quite easily (the quick link pins are loose) simply by holding the bike steady and pushing down on a pedal with one foot. If I could find some weights with which to apply a load that would save my bike from possibly incurring any damage and give an actual load value that broke the chain. I'll see what I can do.

I haven't experiment with metals as much as others but I've seen chemical attacks on plastics from solvents like acetone. The solvent went after any high stress areas in the plastic (small radiused corners in an injection molded part) and created cracks at those junctions. I believe something similar has happened here. I wasn't entirely sure at first either but I certainly knew that I did not start with a damaged chain, at least nowhere near the damage that exists now. Remember, the chain only had 1/32" stretch when it was removed.

I will be cutting and installing another X9 chain on a friend's bike very soon. I just realized that I can experiment with the extra links (I should have a few) which will be brand new. I don't have quick links to try but I have a pretty good chance of finding cracked links in 8-12 months based on my experience.

You've also hit the nail on the head as to why I posted this in the first place. It was intended as a warning to anyone else who might not have considered the difference between "green" OMS and regular OMS (or any other similar solvent combination) and ASSumed like I did that there wasn't anything "bad" (like water) in GOMS. It's a mistake I'll only make once and hopefully this post stops someone else from making the same.

joejack951

Dude, where have I cried that my chains/cassette didn't last long enough? The only thing I've complained (and only once I believe) about in this thread is how I ruined two perfectly good chains and thus lost ~8000 miles of use on my cassette. As mentioned above, the intention of this post was not to complain about my chain or cassette wear. It was to point out the damage caused by what I thought was a safe solvent in which to soak chains. Had I not noticed the damage and attempted to ride those chains, it could have been painful. I don't want someone making the same mistake.

joejack951

If anyone else has found this thread merely annoying, like p2templin apparently has, let me know and I'll let it die. I'm ASSuming again, based on the activity and comments, that most have found it interesting at least and informative at best.

joejack951

You said this:

I might have parsed your sentence incorrectly and perhaps you meant "maybe you should change your chain brand." Either way, you were quick to blame the chain which is my point.

That some people get far less mileage than I had on these chains is irrelevant. Both chains only had 1/32" of stretch. I would guess that fewer than 1% of cyclists replace their chains at that point.

Yes, you'd have to be nuts to try and ride a chain where the quick link and several chain links showed obvious cracks. You'd be nuts to use a quick link with loose pins.

You seem to be in a growing minority.

Well, you will get your wish on the new chain at least. I will be installing an X9 chain on a friend's bike and plan to put the excess links in some GOMS to soak for a while.

The current chain has definitely seen road salt. I've also used other bikes around road salt (and ocean air) and never had cracking problems before (SRAM and Shimano chains). That includes a bike that has been dedicated to only being used on snowy/icy days when the road salt is at its worst. I am willing to believe that chloride potentially left on the chain could have caused the cracking due to soaking in it for so long. The new section of chain in GOMS will eliminate that variable so it will be interesting to see the results. If anyone wants to donate a new KMC quick link to the cause I'd be interested in soaking it too.

catmandew52

I have been following this thread from the beginning.
When it was first posted, I thought, stress/fatigue cracks, pain and simple. Abused chains and drive train wear had to be prime cause.
Chemical reactions are waaaay outside my area of knowledge.
I did take several metallurgy classes quite a long while ago and I faintly recall going over some chemical reactions of chloride precipitates and metals.
The only other time I have seen a crack appear after cleaning, when there definately were none before, was from an ultrasonic cleaning tank. The stresses for the crack were already there, the hi-frequency just induced more stresses at that spot causing the crack.
Stamping and automated assembly machines have got to put a lot of stesses in those skinny little plates.
Chlorides could, I quess, attack those micro stress cracks that are already there causing epic fail.
I have always used Gunk parts cleaner (it's chlorinated) to clean chains and such, but I never leave any parts in the tank longer than overnight. I do, however, soak some parts in oil before before storing them away for use much later.

cyccommute 

So your suggestion is to ignore the problem and just move along? joejack asked a question about something that happened to him. He inquired. I applaud him for that and for having an open enough mind to see an answer. In asking why something happened, he has learned something that may be valuable to him in the future. We are all enriched by his question and by the responses. A well deserved , joejack.

To paraphrase the Most Interesting Man in the World, 'Be curious, my friend.'

cyccommute 

While chlorinated solvents have other problem...<Cough> health...they won't cause corrosion. The chlorine in those molecules is very tightly bound to the carbon chain. The issue with chloride corrosion is when you have free ionized chlorine in the solution. The free chlorine binds to the iron making iron chloride. But because the iron in iron chloride can exist in different oxidation states and the chloride can be replaced by oxygen to form iron oxide (rust), there's always a supply of free chloride around to do more damage. It's what we chemists call a 'catalytic reaction' because the reagents are constantly being regenerated to push the reaction forward.

Now don't get me wrong. joejack's cracking issue is not completely chemical. If you were to put a piece of steel that had not had anything done to it except being formed into his solvent mixture, nothing would have happened. The stamping and the numbers and the addition of the pins all give rise to places that let chemical cracking occur. The chloride just takes advantage of the opportunity.

catmandew52

Chemical bonds always put my head in a fuzz bag even when my brain still worked correctly.
Stress/fatigue cracks and failures and resolution of said failures (from a stricly mechanical point of view) is one of the things I did over the last 40 odd years.
A stress/fatigue failure on an old or very old part, in continuous service, is not unexpected, at least not to the initiated.
Component age is determined by the life cycle designed into that particular item. To increase the life cycle (if that's what the customer wants), a change in materials, a change in manufacturing process or a redesign of the part in question is required.
I have seen parts still in process crack/fail because a needed step was skipped or not deemed necessary the first time around.
"How it's Made" has a segment on bicycle chains,https://www.youtube.com/watch?v=h8j5-dC6_x8, and I am wondering if the chains joejack has were improperly heat treated, or the quenching medium was contaminated? Coluld the cholorinated solvent weaken the molecular bonds at the micro cracks/stresses left over from the manufacturing process, other than turning it into ferric oxide(rust, which I am very familiar with living in Michigan), I don't know, but obviously something is going on.

deermouse

I have nothing against KMC chain. My point was that there are variables other than soaking. But since you plan on testing a new piece of KMC chain it would be interesting to also test some other brands and see how they compare after being subjected to the same soak.

Science and the truth isn't a popularity contest. Did you start this thread because you wanted to warn others or because you were interested in what caused the chain to crack? Earlier in this thread I stated that I had an Ultegra chain that broke after only 1,100 miles. What would you think if I stated that Ultegra chains were only good for 1,100 miles, or maybe only 10 speed Ultegra chains since mine was a 10 speed chain.

It's not irrelevant! I agree that few cyclists (myself included) would replace their chains with 1/32" of stretch. But I would question why a chain with 4,000 miles had only 1/32" of stretch. About the same stretch that my 1,100 mile Ultegra chain had when it broke (here in Colorado we don't have much of a salt problem). At that rate you should be getting 8,000 miles on a chain before it is worn out; not many cyclists get that kind of mileage. There may have been reasons that your chain didn't show more stretch. Maybe there was still dirt or corrosion inside the rollers that filled some to the space caused by wear. Maybe there is some other explanation like KMC chain is extremely wear resistant or you're using a very good lube.

I don't have a new KMC quick link, but I will donate a new section of Sram 1031 chain if you would like to soak that to compare with your section of KMC X9 chain. Maybe you'll find the Sram chain performs better in corrosive environments. I don't know what the result will be, but after testing you will know. Maybe others will donate sections of other brands too.

The important thing is to determine better ways to maintain your bike and know why they are better.

deermouse

Cool video, but my avatar is afraid of your avatar.

catmandew52

Here is some more info to ponder about joejacks cracked chain links.
https://en.wikipedia.org/wiki/Stress_corrosion_cracking
and since it looks like those chains are plated, https://mechanicalplating.com/hydrogen.htm and https://metallurgyfordummies.com/hydrogen-embrittlement/

garage sale GT

could it be the solvent was actually alcohol instead of mineral spirits?

Perhaps the plating on the chain protected it except for where it was worn through around the pin, which is also where the radial stress would be the highest.

I think stress makes metal more susceptible to chemical attack, and the riveted pin is constantly putting outward pressure on the link. As a crack develops, it serves to concentrate stress at its vertex.

DaveSSS 

9 speed chains won't elongate as fast as a 10 speed chain, if ridden in the same conditions, due to their larger surface area. Different chain brands wear differently due to the choice of material. Campy chains, for example may elongate as little as 1/4 the amount of others. I tested a Campy 10 chain against a Shimano 7800 chain and a cheap KMC DX10SC. The Shimano chain elongated the fastest and the KMC was close behind. Both would reach .5% or 1/16" per foot in 3500-4000 miles. The Campy chain elongated at about 1/4 the rate, but that does NOT mean that the chain would last 4 times longer. In addition to elongation, there is much greater wear on the rollers and side plates of all chains. The Campy chain's rollers still wore less than the others, but not a lot less. After 6,000 miles of use, one of my Campy 10 chains had about .15% elongation, but the roller spacing had increased from .200 inch to .240 inch and the side clearance was about .013 inch or twice the amount of a new chain. That chain was totally shot and the previously new cassette had enough wear on one cog that it skipped with a new chain. That shows that there is more to avoiding cog wear than just measuring chain elongation.

Campy suggests using calipers to measure between the rollers over a length of about 5 inches and toss the chain when the length increases by about .020 inch. If you do that, the reading is mostly the wear on the rollers at each end of the measured length, plus a small amount of elongation. IMO, that chain is about half worn, particularly if you alternate the use of several chains with one cassette. When the roller spacing increases by about .020 inch, then it's time to change to the next chain in the rotation, but not time to toss it.

crashmo

Ummmm, I'm going to go compulsively inspect my chain. Right now!