Road Bike, Cycling Forums banner

Chain Autopsy (long)...

1K views 7 replies 6 participants last post by  Kerry Irons 
#1 · (Edited)
Here’s a few of my recent findings on chain and cog wear. If you’ve ever read Jobst Brandt’s (author of The Bicycle Wheel) writings on chain wear measurement, you know he’s a firm believer that measuring the change in the chain pitch is the only valid way to determine if a chain is worn. This can be done with a 12” (or longer) scale or by comparison of the total length with a new chain. The many brands of chain wear measuring tools that bear against a roller at each end create the false impression that the chain pitch is longer than it is. From Brandt:

“The Rholoff chain measuring tool makes some invalid assumptions because it includes the roller clearance in the measured link pin wear. Roller clearance is not uniform among chain manufacturers and can grow faster or slower than the chain elongation, depending on method of chain "care". The Rholoff gauge measures between two rollers a several links apart. Because rollers always have clearance, they can be displaced longitudinally on the chain. Roller clearance is not important to chain performance while the pitch of the links, pin to pin is.”

I now know this for a fact, having just disassembled some links of a Campy chain with 6000 miles of use. This chain showed only the smallest amount of elongation, after all these miles. The only way I could get an accurate estimate of the increase in pitch was to compare the tightly stretched chain to a new one. I found an elongation of only 1/16” over four feet, or about ¼ of the allowable elongation. This figures out to be as little as .0003 (three tenthousandth’s) of an inch wear on a pin, assuming that the pin and bushing wear equally. Micrometer measurements of several used pins proved this to be accurate.

The wear on the rollers was a different matter. The OD of a typical roller showed a reduction of about .006 inch, or about 20 times the wear on the pins. If you’ve ever disassembled a chain link, you’ll see that the roller rides on a bushing or sleeve, formed into the inner side plates. This is a 2-piece bushing with a gap in the center. A severely worn roller will have a diameter in the center that’s about the same as new (since it never touches the bushing), but an area on both sides, where it contacts the bushing halves, that is much larger. It’s quite easy to see the depth of the wear with the naked eye, since they are so deep. Getting an accurate measurement of the ID is a bit tough, but it was much larger than the wear on the OD, with an increase in diameter of about .010 inch, which is about 30 times the wear on a pin.

The question that comes to mind is whether using this chain which has a perfectly acceptable pitch, but severely worn rollers, had an adverse affect on cog life. Brandt claims that is does not. I measured the roller displacement on the 6000 mile chain and found that it increased by a huge amount, from an original .200 inch to .235. That suggests to me that roller wear has little affect on cog life. What I can tell you is that I managed to baely wear out a 19 tooth steel cog on one cassette in 6000 miles and both the 19 and 21 tooth Ti cogs on a Record Ti/steel cassette with only about 4000 miles of use. Both of these cassettes were used with only one chain. While this might seem like a short “cassette” life, you have to put this life into perspective. My cog wear results from a great many miles coupled to a 28T chainring that creates a much higher chain tension than if the cogs were used exclusively with 39 or 53 tooth chainrings. I also use these cogs with the larger chainrings, of course, but far less often. Another rider on different terrain might spread the wear out more evenly over 3-4 cogs and get twice the mileage before a single cog wore out. One thing I’m going to try is changing back to a 30T little ring, to move the wear up one cog, to the 21 and 23. Of course, I’ll lose one low gear, but it doesn’t see a lot of use.

Another thing to keep in mind is that either cassette and chain could have been used much longer, since the chain pitch was well within tolerance. I put a lower mileage (4000) chain on the cassette with 6000 miles on it and it worked fine. Only a couple of weeks earlier, I installed an almost new chain (300 mile of use) on this cassette, just to check for wear. No chain skip occurred at all, indicating that the 19T cog was just barely worn beyond the limit of working with a new chain.

The problem that Brandt points out with using worn cogs for too long is the risk of actually breaking off teeth, since the load becomes concentrated on the last tooth engaged (I’ll take his word for that). That was enough to convince me to bite the bullet and replace both of my cassettes. It also convinced me to never buy another Ti cog (no more pricey Record Ti/steel cassettes).
 
See less See more
#2 ·
Just a thought, correct me if I'm wrong:

I think that if the ID and OD of the rollers are worn in a similar fashion at all of the links that the loading of the cassette teeth will also be even. This may be acceptable, though the decrease in OD of the roller would distribute the load to the tooth differently. Chain stretch on the other hand tends to concentrate loading on the last engaged tooth.

Why is there a greater tension in the 28? Is your cadence dropping or do you just feel stronger in the little ring? :D

Maybe you're wearing out cassettes because you've got mad, crazy watts, yo. :D :D :D
 
#3 ·
chain tension...

For a given gear ratio, chain tension is proportional to the number of teeth on the chainring. My 28/19 for example is about the same as 39/26 or 39/27, but the tension is about 39% greater. That certainly can't enhance cog life.
 
#7 ·
i would think decreased roller diameter would wear the cassett due to load distribution on the teeth. in fact i had some interesting observations that may be related.

i use Ti cassetts by cycle dynamics. when i first installed the Ti cassett, i installed a new sram pc 69 i think....their high end 9 speed chain. after a short time i observed deformation of the Ti teeth. they were slightly smashed and mushroomed. this condition didn't get worse and didn't effect performance at all. some time later i switched to a campy set-up and installed all new cassett cogs with the new campy chain. in this configuration the deformation is way less if at all. i never tried measureing anything to figure this out.

as a side note, IMO campy chains are very nice. i love the way they look and in my application they work better
 
#8 ·
Tooth loading

C-40 said:
The question that comes to mind is whether using this chain which has a perfectly acceptable pitch, but severely worn rollers, had an adverse affect on cog life. Brandt claims that is does not.

The problem that Brandt points out with using worn cogs for too long is the risk of actually breaking off teeth, since the load becomes concentrated on the last tooth engaged (I’ll take his word for that). It also convinced me to never buy another Ti cog (no more pricey Record Ti/steel cassettes).
It seems (logical?) that the only way you could have differential loading of the teeth on a cassette cog or a chain ring is when the chain was elongated. As you wear a tooth, the width (viewed from the side) may get smaller, but the pitch cannot change. This is also true if the bushings of a chain are wearing - the pitch remains the same and so the loading per tooth remains the same. However, when a chain elongates, you now have a situation where each subsequent tooth is loaded less and less, or at least where the roller is contacting each subsequent tooth at a different point, "higher up" on the tooth profile (farther out on the radius of the chain ring or cog). I have seen cases of extreme chain wear when you could actually see daylight between the chain and chain ring as the chain was riding higher and higher on the teeth as the chain wrapped aroung the chain ring.

My own experience with Ti cog wear was that a 19t Ti cog (Record 9, "6+3" cassette) wore about 5X faster than the adjacent 18t steel cog. This was based on a general visual observation that both cogs looked comparably worn, but that the 18 had about 5X the miles on it compared to the 19. Not too scientific, but in the ballpark, I think.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top