Ultegra & Dura Ace chains barely elongated after ~3500 miles (5600 km)

Back by popular demand, the general all-things Road forum!

Moderator: robbosmans

talltales
Posts: 185
Joined: Mon Mar 11, 2019 10:43 pm

by talltales

TobinHatesYou wrote:
Wed Jun 10, 2020 10:53 pm
The chain might be even less elongated that indicated. The KMC digital chain checker measures roller inner to inner instead of inner to outer, so it doesn't isolate out roller wear. Also check multiple sections of the chain just in case.
id suggest roller wear should be included. From my experience only taking true elongation into account may leave you with very worn rollers on a chain that is still within spec. Imo both should be considered, that is a chain that is not overly stretched and does not have excessive roller wear.

The KMC tool seem to have a reasonable approach to this, measuring over 8 links, ~100 mm (minus the roller diameter), allowing 0.8 mm comes out to ~0.8% total wear from the sum of true elongation and roller wear. Ive been doing the same thing with a digital vernier caliper before knowing about the KMC tool, allowing for up to 1mm, compared to a new chain, before discarting the chain. At that point the chan IS notably worn, but may not exceed 0.5% true elongation.

by Weenie


Visit starbike.com Online Retailer for HighEnd cycling components
Great Prices ✓    Broad Selection ✓    Worldwide Delivery ✓

www.starbike.com



Krzysio
in the industry
Posts: 41
Joined: Thu Jul 11, 2019 8:45 am

by Krzysio

I can see discussion stopped long ago, but there is so many discrepancies between chain checkers, factors influencing chain wear, people trying to find out how to deal with this problems, and people losing whole drivetrains by not understanding highest importance of the chain, that I'll dare to drop here several posts for those looking for advise.

Krzysio
in the industry
Posts: 41
Joined: Thu Jul 11, 2019 8:45 am

by Krzysio

Roller wear therefore only impacts how far the rollers shift relative to the pins, but has no impact on chain pitch.
[/quote]

You're right – rollers have no influence on the chain pitch, but worn rollers have thinner walls, and this way they make pins' line around chainring shorter, letting pins to get closer to the bottoms of the teeth valleys. But my experience shows - I'll explain later how – this influence is much smaller than pins' wear. And my guess is, ( so far it's only guess ) influence is smaller, because pins' wear happens by losing material on most critical point to cause elongation, while rollers get rolled, without losing material, so this never lost material still keeps pins away of bottoms of the valleys more than if material was lost.

Krzysio
in the industry
Posts: 41
Joined: Thu Jul 11, 2019 8:45 am

by Krzysio

TobinHatesYou wrote:
Thu Jun 11, 2020 1:05 am
bm0p700f wrote:
Thu Jun 11, 2020 12:55 am

If you rode in 53 12t all the time your would get fewer miles from your chains.

And? Hardly anyone who actually knows what they are doing rides at the cadences you do.
I know mate. Stay with me, as I'm going to tell you funny story adequate to this topic.

Krzysio
in the industry
Posts: 41
Joined: Thu Jul 11, 2019 8:45 am

by Krzysio

pdlpsher1 wrote:
Thu Jun 11, 2020 5:26 am
That makes no sense. If the rollers become smaller/thinner then the distances between roller contact points become longer. Thus the chain would want to 'ride' higher up on the cassette tooth. Roller wear has the same detrimental effect as elongation.

Sent from my KFMAWI using Tapatalk
No mate. Distances between rollers stay exactly the same as they were, as long as pins are not worn, because pins carry them, and rollers need to obey, and lay exactly where pins will put them. But I'm not saying, you're 100% wrong about importance of rollers' wear. Stay with me, and I'll say more.

Krzysio
in the industry
Posts: 41
Joined: Thu Jul 11, 2019 8:45 am

by Krzysio

alanyu wrote:
Thu Jun 11, 2020 9:26 am
TobinHatesYou wrote:
Thu Jun 11, 2020 9:12 am
alanyu wrote:
Thu Jun 11, 2020 7:49 am
+1 roller wear should also counts. A chain with little elongation but significiant roller wear has more strain on the cassette and chainring, shortening the period of switching the cassette and chainring.

This isn't true... to put more strain on a cassette/chainring you need to reduce the points of engagement. An elongated chain puts more wear on those components because the chain/pin pitch doesn't match the gear pitch. This means that when one roller is engaged with one tooth, the next roller is only partially engaged or not at all, and it gets worse and worse farther away. When chain pitch is conserved, as with consistent roller wear, the gears still mesh with the chain correctly.
It is not only the spacing counts. The new or little weared rollers match well with the gear shape while the weared ones not. The less contacting area, though all the rollers seem to engage, puts more strain under the same power.
Contact area is even larger as rollers get rolled to larger diameter, while not losing material. However there is still enough space to take significantly tired guys between undamaged teeth, like on this photo.
DSC06880.JPG

Krzysio
in the industry
Posts: 41
Joined: Thu Jul 11, 2019 8:45 am

by Krzysio

alanyu wrote:
Thu Jun 11, 2020 11:35 pm
TobinHatesYou wrote:
Thu Jun 11, 2020 11:19 pm
alanyu wrote:
Thu Jun 11, 2020 10:52 pm
TobinHatesYou wrote:
Thu Jun 11, 2020 10:20 pm



That page just says roller wear occurs. That's a given. The point you continually ignore is that the wear does. not. matter. in the context of functionality, only structural integrity.
So you want a scientific paper on the function effect? I can easily find a lot of papers on this topic.

https://www.researchgate.net/publicatio ... on_fitness
https://www.sciencedirect.com/topics/en ... ller-chain

It seems like you just searched for "roller" and "wear" and came up with random papers that broadly discuss wear on "roller chains." Drivetrain issues occur with elongation. Roller wear does not result in elongation. Give up.
It seems like you read the title and think ok it is like this I give up.

Several sentences in the paper:

The effect of wear on a roller chain is to increase the spacing of the links, causing the chain to grow longer. Note that this is not from any actual stretching of any metal, as too many engineers and mechanics intuitively believe but is due to the effect of wear at the pivoting parts. It could be said that the roller chain loosens with wear. After a long period of running time the pitch of the chain increases uniformly which results in an increase in the length of the chain.

A lubricant-starved chain drive shows a brownish or rusty coloration around the joints and in the roller-bushing areas when the link is disassembled and the pin inspected. The normal highly polished surface of the pin will have deteriorated to a roughened, grooved, or galled surface that can eventually destroy the hardened surfaces of the chain parts and increase wear until the drive is completely destroyed. This also will be true if a lubricant that does not meet the chain's technical specifications is used.

First paragraph teaches you what the wear of the pivoting parts, rollers, will do. The second paragraph teaches you, lube, actually protects the wear of the joint and roller aera, not elongation of the plate.
There is plenty papers. Some of them are wrong, some of them easy to misunderstand, but still, though you're not 100% wrong about importance of rollers' wear, you're wrong about what it's actual influence is.

Krzysio
in the industry
Posts: 41
Joined: Thu Jul 11, 2019 8:45 am

by Krzysio

TobinHatesYou wrote:
Fri Jun 12, 2020 1:28 am
The clauses you linked are saying articulation increases with chain wear because the rollers are being displaced outward causing the plates to sit at a closer radius to the gears and also that spacing increases if pushed in opposing directions. All true. Once again, no one is denying that roller wear exists, and no one is disputing that it can’t accelerate overall chain wear with the increased articulation. What it doesn’t do is count toward measures elongation of a chain or cause noticeable adverse gear engagement.

If you want a practical demonstration, remove all the rollers from a chain, install it on a bike and ride. It will work.
Although I generally agree with what you say on this discussion, I have to say you went to far with this riding on bare pins. Imagine what happens then at 3 o'clock on your chainring – pins will be dropped right to the bottoms of chainring valleys, by distance of reduced to zero rollers' walls. That'll cause reduction of ring diameter wrapped by pins. So, linear lenght of chain's section wrapping the chainring stays the same, but say – 30 links of chain will wrap shorter diameter of your chainring, while pins are not being pushed away of teeth valleys anymore, as they normally are by rollers. Final result of shortened pins' wrap line will be the same, as with elongated chain – 30-th pin climbing onto the tip of the tooth. Or even earlier one. I've obviously never tested it, so I'm open to be corrected. Bu if I think changed geometry...

Krzysio
in the industry
Posts: 41
Joined: Thu Jul 11, 2019 8:45 am

by Krzysio

I never buy Shimano chains, but I think it's also a riding style that makes a bigger difference in how the chain wears. With today's gear ratios, most cyclists ride with a cadence of around 90rpm. But I know a few, especially older, even experienced cyclists who could not switch to a higher cadence. One of them, who covers 20,000 km a year, rides with a cadence of 53rpm, even though he only rides on flat terrain, steady pace, all the time sitting on the saddle 52/14! He has roughly the same power as me, but he declares that his chains last 60-70% more than mine, even though he washes the chain only once or twice a year, and usually only wipes it and lubricates it with medium-quality lubricant.
I decided to take a closer look at this conundrum, suspecting that his chains would be worn beyond measure, and I took his worn chains: Ultegra 10-speed and Dura Ace 11-speed.
Here's what I discovered first: the chains were almost as short as the new one! The pins were intact, but the rollers had dished tops, and were rolled. I think that this turn of events is partly due to the fact that when riding at such an extremely low cadence, the chain travels about 40% less distance than the distance of the biker's riding at high cadence ( gear ratio difference makes them both to cover same distance with different chain speed ). This slow chain, in turn, determines that the pins at the same time make 40% less movements on the abrasive attacking them. And the rollers crease much more, where the constant pressure needed to turn a hard gear is much greater, than when riding in a soft gear.
The owner of these chains considered them to be worn, because that's what the chain wear indicator showed. However, the indicator does not show whether pins or rollers are worn. Strictly speaking, in this case it showed only the wear of the first and last roller of the measured section, while on a chain with worn pins, the sum of the wear of all pins of the measured section is shown, and then the indicator gives a useful value. At first glance, it seems that the described chain should be suitable for disposal much sooner, than it was, since only 2 worn rollers made gauge already showing 100% wear of the chain. What if you add up the wear of all rollers from between! But this is not the case, because the wear of the rollers - as long as the pins are not worn - does not change the length of the chain, because the pins determine the position of the individual rollers anyway, and more importantly, the position of all chain elements. So, it is wear on the pins, not the rollers, that causes the chain to elongate. And when the rollers are significantly worn, they still fit properly into the valleys between the teeth ( see photo ), until the pins have wear that increases the play of the links. In turn, it is the length of the chain section entwining the chainring (i.e. about 28 links on a large chainring) that determines the threat to the rest of the drivetrain elements from the elongated chain. This is so, because in the new chain the spacing of the links/pins is correlated with the spacing of the teeth, and when the chain gets longer - for example 6mm on 30 links - the chain wraps around the ring, the first roller lies at the bottom of the teeth valley, and the 30th roller is 6 mm too far, i.e. at the tip of the tooth . And destroys it.
When I examined these chains, my city bike's chain was about to be replaced. So I changed it to this Dura Ace with partly damaged rollers ( but still short ) and it still works technically OK, and feel as good as new one rather than mine just worn out. Almost no slack can be pulled on the chainring.
So, in this case, a chain worn by the reading of instrument is still safe to use - for a long time so far - although it probably shouldn't be put on a $10,000 bike at this stage. Especially it's safe to be used by high cadence, not a strongest guy on the planet, to start wearing pins/stop smashing rollers. And that “long time” on my commuter is not particulary long during Polish slush, road salt period to be prcise. In such circumstances pins start disappearing quickly even with Silca Synergetic lube.
Attachments
DSC06880.JPG

alanyu
Posts: 1503
Joined: Thu Jun 06, 2019 1:10 pm

by alanyu

Krzysio wrote:
Thu Mar 30, 2023 3:00 pm
alanyu wrote:
Thu Jun 11, 2020 9:26 am
TobinHatesYou wrote:
Thu Jun 11, 2020 9:12 am
alanyu wrote:
Thu Jun 11, 2020 7:49 am
+1 roller wear should also counts. A chain with little elongation but significiant roller wear has more strain on the cassette and chainring, shortening the period of switching the cassette and chainring.

This isn't true... to put more strain on a cassette/chainring you need to reduce the points of engagement. An elongated chain puts more wear on those components because the chain/pin pitch doesn't match the gear pitch. This means that when one roller is engaged with one tooth, the next roller is only partially engaged or not at all, and it gets worse and worse farther away. When chain pitch is conserved, as with consistent roller wear, the gears still mesh with the chain correctly.
It is not only the spacing counts. The new or little weared rollers match well with the gear shape while the weared ones not. The less contacting area, though all the rollers seem to engage, puts more strain under the same power.
Contact area is even larger as rollers get rolled to larger diameter, while not losing material. However there is still enough space to take significantly tired guys between undamaged teeth, like on this photo.
DSC06880.JPG
LOL. Are you joking? When pedalling with worn rollers, the driving force will make the rollers move to the thinner part of the teeth as I drew one page before, opposite to what your finger does.

Krzysio
in the industry
Posts: 41
Joined: Thu Jul 11, 2019 8:45 am

by Krzysio

alanyu wrote:
Thu Mar 30, 2023 3:33 pm
Krzysio wrote:
Thu Mar 30, 2023 3:00 pm
alanyu wrote:
Thu Jun 11, 2020 9:26 am
TobinHatesYou wrote:
Thu Jun 11, 2020 9:12 am



This isn't true... to put more strain on a cassette/chainring you need to reduce the points of engagement. An elongated chain puts more wear on those components because the chain/pin pitch doesn't match the gear pitch. This means that when one roller is engaged with one tooth, the next roller is only partially engaged or not at all, and it gets worse and worse farther away. When chain pitch is conserved, as with consistent roller wear, the gears still mesh with the chain correctly.
It is not only the spacing counts. The new or little weared rollers match well with the gear shape while the weared ones not. The less contacting area, though all the rollers seem to engage, puts more strain under the same power.
Contact area is even larger as rollers get rolled to larger diameter, while not losing material. However there is still enough space to take significantly tired guys between undamaged teeth, like on this photo.
DSC06880.JPG
LOL. Are you joking? When pedalling with worn rollers, the driving force will make the rollers move to the thinner part of the teeth as I drew one page before, opposite to what your finger does.
I'm not joking mate. I'm not telling you where they are exactly being pushed. The photo is only to show there is enough space for even worn roller, as this one is, even that worn roller has actually larger diameter, producing larger contact area. I read you probably saying there is smaller contact area close to the tips of the teeth. However I have to say I'm actually riding now this chain, and because pins are in perfect condition, there is hard to pull the chain off the chainring more than any new chain. Because pins are not worn, overall lenght of the chain not changed, pins hold the rollers close to the bottoms. If any force wants to move them to any other area of the teeth, it has to do to all of them wrapping chainring, having not elongated chain. Hard to imagine it at 3 o'clock on the chainring, so all rollers have to stay at the same positions. I have camera good enough to record what happens on real ride, but may have not enough determination. Feeling on ride is much closer to new chain, than just replaced one.

alanyu
Posts: 1503
Joined: Thu Jun 06, 2019 1:10 pm

by alanyu

Krzysio wrote:
Thu Mar 30, 2023 4:07 pm
alanyu wrote:
Thu Mar 30, 2023 3:33 pm
Krzysio wrote:
Thu Mar 30, 2023 3:00 pm
alanyu wrote:
Thu Jun 11, 2020 9:26 am


It is not only the spacing counts. The new or little weared rollers match well with the gear shape while the weared ones not. The less contacting area, though all the rollers seem to engage, puts more strain under the same power.
Contact area is even larger as rollers get rolled to larger diameter, while not losing material. However there is still enough space to take significantly tired guys between undamaged teeth, like on this photo.
DSC06880.JPG
LOL. Are you joking? When pedalling with worn rollers, the driving force will make the rollers move to the thinner part of the teeth as I drew one page before, opposite to what your finger does.
I'm not joking mate. I'm not telling you where they are exactly being pushed. The photo is only to show there is enough space for even worn roller, as this one is, even that worn roller has actually larger diameter, producing larger contact area. I read you probably saying there is smaller contact area close to the tips of the teeth. However I have to say I'm actually riding now this chain, and because pins are in perfect condition, there is hard to pull the chain off the chainring more than any new chain. Because pins are not worn, overall lenght of the chain not changed, pins hold the rollers close to the bottoms. If any force wants to move them to any other area of the teeth, it has to do to all of them wrapping chainring, having not elongated chain. Hard to imagine it at 3 o'clock on the chainring, so all rollers have to stay at the same positions. I have camera good enough to record what happens on real ride, but may have not enough determination. Feeling on ride is much closer to new chain, than just replaced one.
When talking about worn rollers, you mean worn outside but perfect inside? No, during my use I haven't met that situation. Rollers are always worn both sides, together with pins, though the rate is different.

How can a worn roller has a larger diameter than a new one? Even the worn roller stay perfectly as the same position as a new roller, the contact area is smaller, unless the ring/cassette has been already deformed to match the worn roller.

One case is that, some commuters don't maintain their bike, and the drive system are extremely worn out, including chains, rings, pulleys and cassettes. The worn chain doesn't skip under gentle pedalling, but the new chain on the old system will skip immediately, because the teeth are strongly deformed to match the worn chain.

Krzysio
in the industry
Posts: 41
Joined: Thu Jul 11, 2019 8:45 am

by Krzysio

alanyu wrote:
Thu Mar 30, 2023 4:31 pm
Krzysio wrote:
Thu Mar 30, 2023 4:07 pm
alanyu wrote:
Thu Mar 30, 2023 3:33 pm
Krzysio wrote:
Thu Mar 30, 2023 3:00 pm


Contact area is even larger as rollers get rolled to larger diameter, while not losing material. However there is still enough space to take significantly tired guys between undamaged teeth, like on this photo.
DSC06880.JPG
LOL. Are you joking? When pedalling with worn rollers, the driving force will make the rollers move to the thinner part of the teeth as I drew one page before, opposite to what your finger does.
I'm not joking mate. I'm not telling you where they are exactly being pushed. The photo is only to show there is enough space for even worn roller, as this one is, even that worn roller has actually larger diameter, producing larger contact area. I read you probably saying there is smaller contact area close to the tips of the teeth. However I have to say I'm actually riding now this chain, and because pins are in perfect condition, there is hard to pull the chain off the chainring more than any new chain. Because pins are not worn, overall lenght of the chain not changed, pins hold the rollers close to the bottoms. If any force wants to move them to any other area of the teeth, it has to do to all of them wrapping chainring, having not elongated chain. Hard to imagine it at 3 o'clock on the chainring, so all rollers have to stay at the same positions. I have camera good enough to record what happens on real ride, but may have not enough determination. Feeling on ride is much closer to new chain, than just replaced one.
When talking about worn rollers, you mean worn outside but perfect inside? No, during my use I haven't met that situation. Rollers are always worn both sides, together with pins, though the rate is different.

How can a worn roller has a larger diameter than a new one? Even the worn roller stay perfectly as the same position as a new roller, the contact area is smaller, unless the ring/cassette has been already deformed to match the worn roller.

One case is that, some commuters don't maintain their bike, and the drive system are extremely worn out, including chains, rings, pulleys and cassettes. The worn chain doesn't skip under gentle pedalling, but the new chain on the old system will skip immediately, because the teeth are strongly deformed to match the worn chain.
I don't mean worn one side only. I'm not talking about situation you met, or even Me. Personally I never damaged any roller to be precise. And there is plenty situations happened to others but not you, whether you say "no", or not. Unless you are God there is no must to be allowed by you. Take a bit of plasticine, roll it and you'll se how worn roller can have larger diameter, than new one. Worn rollers will basicaly not lose material. They get worn significantly by being rolled with presure. Pins are abraded, and this, and only this elongates chain. Unless someone will report links elongation. What I never met, but I may just don't know. Next you say rollers are always worn together with pins. Agree, but to my surprise this chain used by guy to ride 11000!!! km has no significant wear to the pins. See the photo: There is 3 chains stretched to the full lenght, right ends in line. Top one is just worn. Bottom one is brand new, washed with coresponding outer link about 5mm to the right, so, this much shorter. And 11000km one in the middle, with corresponding link showing almost exactly the same lenght, as the new one. So, to be understood this time: of coure, precisely - pins wear as well, but in this case negligibly. Personally I'll say - no wear to the pins. That's why I took this chain for examination. Because I haven't met this situation. If you O.K. with that, please read next ( or maybe previous )? post for whole story of this guy.
Attachments
DSC06844.JPG

Krzysio
in the industry
Posts: 41
Joined: Thu Jul 11, 2019 8:45 am

by Krzysio

11000 km chain owner story - previous post. No offence alanyu. Just to share what surprising things we've found.

Krzysio
in the industry
Posts: 41
Joined: Thu Jul 11, 2019 8:45 am

by Krzysio

I've seen completely disintegrating rollers only once, 35 years ago. It was on all steel 2x5 clunker of guy who had no idea about chains, and even started replacing chain sections with random, different width ones. His freewheel sprockets were gone, but thick, steel chainrings survived very well.
Very experienced, and deserving huge respect people like Dave Rome from Cyclingtips, or Adam Keirin – ZFC, prefer certain models of chain checkers. Dave even said something like: “Chain checkers are best way. And you can disagree. And you're wrong!”. I'd pick up the gauntlet. In my opinion the best chain checkers may be the best way for experienced mechanics, who can clearly see what happens to cause all discrepancies, and they have feeling of small differences between how much force to apply, or not to apply, etc. But average folks need to learn as well, instead of throwing tons of chains every week onto counters of mechanics, just to constantly check their wear.
And IMHO, the best way for average guy is to clean the chain ( dirt prevents chain off the full stretch ), put it stretched on the bench, measure long section of the chain, look at the easy to notice, ( instead of uncertain, microscopic ) difference of the chain lenght, look at the rollers, move them with screwdriver to see how much they move on the inner links, and don't worry about their wear to much, until they are really disshaped comparing to the new ones. You'll see, when they're next to disintegrate, what is very rare to happen before chain is too long. ( With above story of the guy, who on 11000km elongated rollers without readable pins' wear I'm not saying impossible ) .Again – rather avoid installing still short chain but with evidently loosy rollers onto expensive bike. This way you know in enlarged scale what happens to your invisible pins as well, as to your relatively easy to assess rollers. If you're not technically skilled at all, you periodicaly take your bike to trusted mechanic anyway. This is how you can check you chain easier way, say as if with help of magnifying glass, and without hesitating about how much your checker is not telling you. Being “metric” I take 30 links: new chain 381mm, worn 383, 384 – chainring teeth start changing shape, 386mm – teeth chewed up within few hours, 396!!! - record one I've ever seen - no teeth at all, rider continued on smallest of the triple chainset. I needed to count links to make sure it's true.
If 2 mm from new to worn is still to small difference for you, go for 90 links to have 5mm space. Although nowadays most people ride with missing link, 30 links is doable with chain on the bike as well. On the photo: left pin edge to left pin edge.
Attachments
DSC06870.JPG
DSC06871.JPG

by Weenie


Visit starbike.com Online Retailer for HighEnd cycling components
Great Prices ✓    Broad Selection ✓    Worldwide Delivery ✓

www.starbike.com



Post Reply