jim p

I bet that if you put these chain parts in salty water for a month that they will be magically welded back together. Then you can soak them in the mineral oil for just a week and they will be good as new.

If you believe this then find you a 10% hill and coast up it without pedaling.

cyccommute 

Not necessarily. Because the damage in the picture is nearly identical, I suspect that there is a stress riser that is introduced into the system during manufacture. It could be the stamping of the numbers on the plates or the swaging of the pin into the plate but it's a stress riser, nevertheless. Stress risers can cause damage through 'stress'...i.e. use...of the product but a stress riser would also open the door to chemical damage through corrosive action as well. The stress riser opens the door to chemicals attacking the metal at a certain site, which opens the crack a little wider, which opens up more area for chemical attack and so on.

This would be something like chloride stress cracking of steel. I'm not saying that chloride is necessarily the culprit here but it would be a similar mechanism.

On second thought, I could even see chloride as the culprit in this case. joejack951 is from Delaware and he said the chain had been soaking for 8 months. That puts the time frame in February just when salt application to the roads is highest. If he had soaked the chain in 'real' mineral spirits, I doubt we'd be having this discussion. But since he used the 'green' spirits that may contain water, he probably introduced chlorides (in the form of sodium or magnesium chloride) into the emulsified mineral spirits. The salt attacks the stress part and, as they say, the rest is history.

joejack951

While that's an interesting theory, I had two chains soaking and both cracked in very similar ways. The other chain was not put in the "green" mineral spirits during the winter. I'm pretty sure it was during the summer based on my typical mileage. Yes, that does make it over a year in the GOMS (green odorless mineral spirits).

LarryMelman

You know, I read this accusation here and there. But I just don't see the logic behind it.

There are a couple types of checkers. I have the simple kind which gives you a go/no-go indication at 0.75% and 1%. Unlike a ruler, it only needs to be accurate in those two locations. I trust whoever made the tool (in my case, Park) that it is. It's a clear pass/fail indicator, I don't have to squint at my ruler and fuss over "is that really a 1/16th or not". The tool puts tension on the chain, I don't have to "hold with one hand and measure with the other" like I'd have to with a ruler. And if my gauge says the chain is worn, I'll certainly get some second opinions - I'll let the LBS use their chain checker, etc.

So, maybe I'm missing something. Just why is a chain gauge inferior?

cyccommute 

A few questions:

1. Did you reuse the solvent from previous soaks? Soaks that were done during the winter time?

2. Was the bike ridden during early spring, after the snow but before the rain had time to wash the salt off the roads?

3. Do you live near the coast or ride near the coast?

Look at this graph from the link I provided



Notice that the temperature range. Yes, that's a high temperature and the cracking is a slow process even at those temperatures. However look at the concentration of chloride and oxygen. A part per million is a milligram per liter. That's roughly a single grain of salt in a quart of water. Even 100 grains of salt is still small.

There are lots of chloride sources throughout the world and picking up a little bit of salt would be very simple. A couple of drops of sweat, handling the chain after a ride, dust on the road, just being near an ocean, etc could all lead to getting salt on the chain.

Going into full on geek mode now: If the solvent you used does contain water, which is likely, that water is only there in about 60%, so a single grain of salt in 1 L of solvent would not be a 1ppm solution but would instead be a 1.7 ppm solution. If you reduce that volume to a cup (250 mL), a single grain of salt could boost the salt level to 7ppm.

But you don't have just plain water in your jar. You have a nonpolar solvent forced into a polar medium with chemical black magic...it's really solid and simple science but many people would consider it magic Nothing in that mixture really wants to be there which is why the mixture is cloudy. Even your chain doesn't want to be there. The metal is perfectly happy to attract a layer of water to itself to keep the oil off of it*. And any salt in the solution will be just a happy to sit down on the metal. The salt in the mixture may even migrate towards the metal surface, further increasing the localized ppm concentration of the salt. Even the oxygen may migrate towards that metal surface rather than get involved with the nonpolar components of the mixture.**

Now you have all the ingredients you need for chloride cracking: increased salt level at the metal boundary, water at the metal boundary, a mixture forcing water away from it, bare metal (remember you are removing the oils from the surface) and, most importantly, time. You don't have temperature but that just makes everything happen faster. A long time period is almost as important as temperature.

The end result is a cracked part.

If you had used mineral spirits without the water in it, none of the above would have happened. There's no water in the mineral spirits...we call it anhydrous which is a 50 cent word meaning 'no water'...so no chloride interaction with the metal. Any chloride on the chain would end up as a lump in the corner of the jar, hugging it's sodium to it's little electronic chest and sobbing quietly to itself***

I love getting my geek on!

*Please excuse the anthropomorphizing but it helps to explain some complex interactions

**To further anthropomorphize, oxygen likes water better than it likes the petroleum distillate. Water has lots of electrons in the right configuration to share with the oxygen making everyone happy.

***Now that's anthropomorphizing!

BikeWise1

Chain gauges measure the distance between the rollers, not the pins. Different chains have rollers of different thicknesses, therefore accuracy is not to be taken for granted.

I use gauges, and rulers, but I measure a dozen or more chains a day, and know what gauge is accurate with what chain. If you own a KMC chain, I can say authoritatively that the Park gauge will declare your chain is nearly 3/4 worn when brand new....

A nice, precision metal ruler is by far the most accurate route.

joejack951

Thanks for the chemistry lesson! I did read it all but snipped it to save space in this reply.

I can't be positive that I did not re-use solvent after cleaning a chain used in the winter. These chains were on my commuter bike so they undoutably were exposed to road salt at some point but were just as likely rinsed off by the rain before getting cleaned. I am certain that I used fresh GOMS for the long term soaking though. That doesn't mean that some salt didn't make it into the mix though.

cyccommute 

To geek out just a little more...and give you an example of the power of chloride corrosion...I have a stainless steel reactors at work. It's made of 316 stainless which is more resistant to corrosion than the stainless steel cladding you find on appliances (304 stainless). It's a high chromium (16% to 18%), high nickel (10% to 14%) steel alloy. The chromium and nickel boosts the metal's ability to resist corrosion to incredible levels. I was running the reactors at 140C, with 0.5% sulfuric acid in an organic/water solvent mixture at pressure of 5 atmospheres. It's a damnably harsh environment.

At one point we started to play around with the organic/water ratios as part of an experimental design. We usually run the reactors at around 50% water/50% organic (through manipulation of some solvent properties, it's a single phase mixture even though it doesn't like being one). We increased the organic to 80% (still single phase) and found the water had separated out into a layer of bright green crystals. The crystals had also coated the biomass were trying to process. We hadn't seen green crystals before and, quite frankly, for the work that I usually do, if it ain't brown there's something seriously wrong.

Well, green crystals are usually associated with certain chromium salts and/or nickel salts. Because of the high chromium and nickel levels in our reactors, green meant that something was chewing up our reactors. I remembered that I had tried to do a really cool demonstration from Scientific American years before, in which you use a solution that is sensitive to chloride. Our facility uses deionized water which has a few ppm of sodium chloride in it. The chloride was corroding the inside of the reactor and I hadn't noticed it until I reduced the water enough to precipitate out the chromium/nickel salts.

We cured the problem...and saved the project and got a patent on the process...by using megaohm water which is as pure as you can make water. We also changed to Hastalloy for the reactor which is much more resistant to chloride. Much more expensive but it stands up better.

dchsueh

As BikeWise1 wrote, chain gauges measure between rollers.

If you grab a roller with your fingertips gently, you can wiggle it w.r.t. the pin. Now imagine a brand new chain with zero wear - the chain gauge pushes the two rollers it measures against away from each other - the new chain is already measuring longer by 2x the wiggle amount per roller.

I expect that different brands of chains vary in the wiggle amount.

I decided I would use a ruler for my own bikes.

JTGraphics

I will resurrect this post in a year for you.
Put it on my calendar notifications
Result will be I have a super clean link and chain LOL

joejack951

More pics! Below is the second quick link which had the pin fall out due to the cracked plate as well as various photos of the cracked chain links. My favorite is the second to last.

deermouse

Maybe instead of changing your solvent, you should change your chain brand, or maybe both. I'm still not convinced that this would happen to a new chain, but I wouldn't soak any chain is a caustic (cleaning) solution like simple green for an extended period. Caustic solutions can corrode metal just like acid. Simple green has a pH of 9.5 which, while not extremely caustic) puts it between see water and ammonia. I wouldn't soak my chain in either sea water or ammonia for an extended period.

Still I wouldn't expect the chain to crack like that unless it was already on the verge of failure. I'd consider a different brand of chain.

On a side note, Sram recommends using new master links whenever removing and reinstalling the chain.

joejack951

The chain was soaked in "green" odorless mineral spirits, not Simple Green. I wouldn't intentionally soak steel parts long term in a water based solution. This was a mistake, that I'll never make again, and which hopefully no one else does either. I'm on my third X9 chain now and they have all performed flawlessly, the first two for 4000 miles and this one for almost 5000 now. I'm not at all concerned about the chains themselves provided I don't do something stupid to them again.

We think differently. If 4000 miles happened to be the breaking point (pun intended) for a KMC chain, how has my third one made it past that point?

Only for 10 speed "Powerlocks". All of their other quick links are ok to remove as needed. I've used both 8 and 9 speed versions of SRAM chains. For the record, I've never had a problem with SRAM chains either (not that I've ever soaked any of them for months in GOMS), or Shimano chains, or Wipperman chains. I'm not concerned about chains. I am slightly more concerned about road salt though.

JTGraphics

O hell their goes my test I don't use that green sprits. O well LOL anyway good to know I guess.

thirdgenbird

while we agree that park gauges are worthless, i had the opposite experience. i measured several campagnolo 10spd chains with various wear and they all measured as if they were perfect.

davidad

What brand green mineral spirits? I haven't seen the green. I used to use unscented spirits to clean my chains until I got an ultrasonic cleaner. Now I use a 50-50 mix of Simple Green and water. Then rinse and dry.
My home brew chain lube is comprised of 4 parts mineral spirite to 1 part chain saw bar oil.
I am at 10,000 miles on an 8 speed chain and the same on a 7 speed.

Rowan

This also assumes that the machining accuracy of a new (stamped) gauge is spot-on, as confirmed by measurement with a pair of vernier calipers. Then there is a wiggle in the rollers.

I can understand the convenience of using a chain checker in a shop environment, and as the guys say, a mechanic would get to know which chains did what.

Indeed, I have replaced a chain on a bike, and week later after fine-weather riding, I've had an associate proudly proclaim with his chain checker that it was entirely worn out and must be replaced. I lost faith in the tools then (not that I ever used them), and more so when I went home in a cold sweat and found my one-foot ruler showed absolutely no deviation in the rivet measurement.

Rowan

Looking at the new set of pix, and in particular the first one, I'd say long-term stress. Those links look damaged on the leading edges, and those in themselves would have acted as stress risers and should have emulated the damage detailed elsewhere on the links, if it was a chemical issue.

cyccommute, would ibreaking the links at the cracks to reveal the structure of the metal on either side of the cracks be a clue as to whether this is chemical or mechanical?

joejack951

Klean Strip.

Sixty Fiver

In the multiple shops environments I spend my time in, we use a ruler and have gotten rid of the chain checkers due their unreliability.

It is not hard to read 1/16 on a ruler.

joejack951

First, it was more than just the quick links which cracked. I posted several pictures of chain links that also cracked. For those claiming this happened during use, all of those cracks would have had to happen simultaneously unless you believe that a cracked plate is strong enough to crack a fully intact plate.

Second, I didn't take any pictures of the reverse side of the linkk but what you are seeing on the leading (and trailing) edges is distortion caused by the coining done to the edges of the link on the opposite side. For whatever reason, there is a short length on the leading and trailing edge that is not coined so the perimeter is not uniformly deformed.

jim hughes

How - chemically speaking - is water able to form cracks in steel parts? Is it just enhancing invisible cracks that were already present?

davidad

Looks like rust on some of the links. The water may be the problem.

p2templin

Why are you expecting chain perfection to 3999 miles, with failure in the subsequent 10560 feet?

I think the summary of the story is:

Chains wear faster than cassettes.
You believe it to be a 3:1 ratio between the two (4k/chain, 12k/cassette)
Therefore you rotate three chains so the chain wear is typically in line with the cassette wear.
This theory has merit: three $50 chains rotated so you can get more life out of a $100 cassette.

However, the chains aren't lasting as long as you hoped. You might have to introduce a new chain into the rotation, or just use the third chain until it's done before replacing everything anew.

Haven't you gotten your money's worth? If two chains are toast, you either shell out $100 for two chains or you might have to shell out $250 for three chains and a cassette. Isn't it about time to shell out the $250 anyway?
However,

cyccommute 

Read posts #55 and #58. Water provides the medium, other stuff does the heavy lifting