Cassette question

For discussions about bikes and equipment.
fastpedaller
Posts: 2283
Joined: 10 Jul 2014, 1:12pm
Location: Norfolk

Re: Cassette question

Postby fastpedaller » 11 Feb 2020, 10:55pm

Going by the weight (and my budget at the time) certainly steel!

Brucey
Posts: 36822
Joined: 4 Jan 2012, 6:25pm

Re: Cassette question

Postby Brucey » 12 Feb 2020, 12:52am

I keep learning new things about chains and sprockets.

A fairly recent realisation is one of how chains fit sprockets. It is fairly obvious that a worn chain will run on a newer sprocket but there will inevitably be some movement of the chain rollers on the teeth whilst under load. I have always thought that this would result in more fatigue damage to the sprockets, as well as more wear.

It is likewise obvious that if the chain pitch is too short for the way the sprocket has been made (or has worn) then it won't mesh with the teeth well and in fact it is this which causes skipping under load.

However what hasn't been clear to me previously is that the chain has to be a fairly slack fit even on a new sprocket, else it too may skip. How slack is this fit? -pretty slack, it turn out. I recently got hold of some older NOS SunTour sprockets (with symmetrical teeth) and I was surprised at how a wrap of new chain around these sprockets was already fairly slack. I tried on a bunch of other new sprockets and although it was less easy to see what was going on (what with funny HG tooth shapes and so forth) the same applied.

This slackness is pretty bad, such that if it were scaled up onto a chainring, you might assume that the chainring or the chain were already worn out!

Thus I suspect that even with a new chain/new sprocket situation, you may sometimes get more severe fatigue loading than you might expect, much as you might expect if you used a worn chain on a new sprocket.

Old sprockets did break occasionally and I'm sure that some were too brittle from the start. But conditions of use have to be a factor too.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

Marcus Aurelius
Posts: 794
Joined: 1 Feb 2018, 10:20am

Re: Cassette question

Postby Marcus Aurelius » 12 Feb 2020, 5:43am

Brucey wrote:I keep learning new things about chains and sprockets.

A fairly recent realisation is one of how chains fit sprockets. It is fairly obvious that a worn chain will run on a newer sprocket but there will inevitably be some movement of the chain rollers on the teeth whilst under load. I have always thought that this would result in more fatigue damage to the sprockets, as well as more wear.

It is likewise obvious that if the chain pitch is too short for the way the sprocket has been made (or has worn) then it won't mesh with the teeth well and in fact it is this which causes skipping under load.

However what hasn't been clear to me previously is that the chain has to be a fairly slack fit even on a new sprocket, else it too may skip. How slack is this fit? -pretty slack, it turn out. I recently got hold of some older NOS SunTour sprockets (with symmetrical teeth) and I was surprised at how a wrap of new chain around these sprockets was already fairly slack. I tried on a bunch of other new sprockets and although it was less easy to see what was going on (what with funny HG tooth shapes and so forth) the same applied.

This slackness is pretty bad, such that if it were scaled up onto a chainring, you might assume that the chainring or the chain were already worn out!

Thus I suspect that even with a new chain/new sprocket situation, you may sometimes get more severe fatigue loading than you might expect, much as you might expect if you used a worn chain on a new sprocket.

Old sprockets did break occasionally and I'm sure that some were too brittle from the start. But conditions of use have to be a factor too.

cheers

Quite right.
As the old adage says “you can’t teach an old cog new links” and vice versa ( to a lesser extent ). Chains and sprockets ‘mesh’ and wear in unison. If you change one half of the equation with out the other, you get problems. As the new ‘equation’ balances, you get increased wear on the new bits.

Brucey
Posts: 36822
Joined: 4 Jan 2012, 6:25pm

Re: Cassette question

Postby Brucey » 12 Feb 2020, 7:41am

my point was in good part, there is some mismatch between even new parts.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

thatsnotmyname
Posts: 189
Joined: 23 Jan 2020, 10:23am

Re: Cassette question

Postby thatsnotmyname » 12 Feb 2020, 8:39am

Marcus Aurelius wrote:Quite right.
As the old adage says “you can’t teach an old cog new links” and vice versa ( to a lesser extent ). Chains and sprockets ‘mesh’ and wear in unison. If you change one half of the equation with out the other, you get problems. As the new ‘equation’ balances, you get increased wear on the new bits.


It’s not an ‘old adage’, it’s a phrase you made up and insist on regurgitating ad nauseam across the internet, using whichever forum/username you happen to be on at the time. It’s been proven wrong so many times I can’t believe you are still trying it on. You are literally the only person who has ever claimed it - which if course makes it very easy to spot wherever else you've tried it. For instance..

https://forum.bikeradar.com/discussion/ ... r-cassette

Marcus Aurelius
Posts: 794
Joined: 1 Feb 2018, 10:20am

Re: Cassette question

Postby Marcus Aurelius » 12 Feb 2020, 10:12am

Brucey wrote:my point was in good part, there is some mismatch between even new parts.

cheers


That’s manufacturer tolerances for you.

Brucey
Posts: 36822
Joined: 4 Jan 2012, 6:25pm

Re: Cassette question

Postby Brucey » 12 Feb 2020, 12:15pm

Marcus Aurelius wrote:
Brucey wrote:my point was in good part, there is some mismatch between even new parts.

cheers


That’s manufacturer tolerances for you.


If it were just that then there would be many more instances of new parts that just don't work with one another. I rather think that there must be a slight mismatch for it to work reliably at all even when the parts are new. That being the case there is no real justification for being precious about thinking there is no place for combinations of parts that are in slightly different wear states; in the simplest terms it either works or it doesn't.


Were this not the case then the whole derailleur scheme would be nigh-on unworkable; the sprockets in a cassette never wear evenly ( with the bulk of the wear being on just e few sprockets), so at any point once the chain is even slightly worn, you are running a chain and sprocket that are imperfectly matched to one another when you are using most of the possible gears.

In a nutshell in a derailleur system, at the sprockets, you either have;

a) the chain pitch slightly shorter than the wear state of the sprockets is best suited to, in which case the chain may skip under load, or
b) the chain pitch slightly longer than the sprockets could use.

In the latter case this arises either by design (e.g. when the parts are new) or through wear. If the mismatch becomes too large all that happens is that the 'bad' wear rate increases, presumably because the loads are more often on only one tooth at a time and there is more likely to be movement under load between the chain rollers and the sprocket teeth.

By contrast in the former case there is little reason to suppose that there is -in the absence of skipping under load- a much higher wear rate than normal, should you use a new chain on slightly worn sprockets.

Hence you should try and avoid using worn chains on good sprockets as a basic principle; it'll work, but it'll probably be wearing everything at an abnormally high rate. However (as many others can testify) if you run a new chain on slightly worn sprockets the worst that can usually happen is that the chain skips under load, which is a surefire sign that you left the chain too long before renewing it.

Slavishly destroying cassette after cassette (not to mention the chainrings) in the mistaken belief that 'they should be kept as a matched set' when in fact the chain is soon worn so that case b) applies is IMHO a flawed approach; they are never 'perfectly matched' in any event and once a chain is slightly worn you have more mismatch than is desirable, and this makes for an ever faster (and ever more harmful) wear rate.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

Marcus Aurelius
Posts: 794
Joined: 1 Feb 2018, 10:20am

Re: Cassette question

Postby Marcus Aurelius » 12 Feb 2020, 1:26pm

Brucey wrote:
Marcus Aurelius wrote:
Brucey wrote:my point was in good part, there is some mismatch between even new parts.

cheers


That’s manufacturer tolerances for you.


If it were just that then there would be many more instances of new parts that just don't work with one another. I rather think that there must be a slight mismatch for it to work reliably at all even when the parts are new. That being the case there is no real justification for being precious about thinking there is no place for combinations of parts that are in slightly different wear states; in the simplest terms it either works or it doesn't.


Were this not the case then the whole derailleur scheme would be nigh-on unworkable; the sprockets in a cassette never wear evenly ( with the bulk of the wear being on just e few sprockets), so at any point once the chain is even slightly worn, you are running a chain and sprocket that are imperfectly matched to one another when you are using most of the possible gears.

In a nutshell in a derailleur system, at the sprockets, you either have;

a) the chain pitch slightly shorter than the wear state of the sprockets is best suited to, in which case the chain may skip under load, or
b) the chain pitch slightly longer than the sprockets could use.

In the latter case this arises either by design (e.g. when the parts are new) or through wear. If the mismatch becomes too large all that happens is that the 'bad' wear rate increases, presumably because the loads are more often on only one tooth at a time and there is more likely to be movement under load between the chain rollers and the sprocket teeth.

By contrast in the former case there is little reason to suppose that there is -in the absence of skipping under load- a much higher wear rate than normal, should you use a new chain on slightly worn sprockets.

Hence you should try and avoid using worn chains on good sprockets as a basic principle; it'll work, but it'll probably be wearing everything at an abnormally high rate. However (as many others can testify) if you run a new chain on slightly worn sprockets the worst that can usually happen is that the chain skips under load, which is a surefire sign that you left the chain too long before renewing it.

Slavishly destroying cassette after cassette (not to mention the chainrings) in the mistaken belief that 'they should be kept as a matched set' when in fact the chain is soon worn so that case b) applies is IMHO a flawed approach; they are never 'perfectly matched' in any event and once a chain is slightly worn you have more mismatch than is desirable, and this makes for an ever faster (and ever more harmful) wear rate.

cheers

This is why I prefer single speed / fixies. So much easier than derailleur gear set ups. ‘Best practice’ is to keep the chains and cassettes matched, the derailleur gear and chain system is fundamentally flawed from an engineering perspective, and there is no ‘perfect match’ keeping them matched ( as best you can) is the best of a bad job, and certainly not a “mistaken belief” no one I know ( who knows what they are actually doing) has ever thought that it’s a “mistaken belief” either.

User avatar
andrew_s
Posts: 5061
Joined: 7 Jan 2007, 9:29pm
Location: Gloucestershire

Re: Cassette question

Postby andrew_s » 12 Feb 2020, 1:36pm

Brucey wrote:Note that with 7s (BITD) the sprockets were 2mm thickness, and up to/including UG most sprockets were made of hard, relatively brittle steel that was relatively slow-wearing but would crack if you shifted clumsily under full power too often. With modern sprockets they are are made of softer steel (which may bend rather than break if you shift under full power) and are as little as 1.6mm thickness. You won't be surprised to hear that I think they wear faster, and it is more important than ever that you don't run with a worn chain

They definitely wear faster.
When I used 6-speed UG cassettes, a new chain would run without skipping at well past 2000 miles.
When I changed to 8-speed HG, the non-skip mileage dropped to around 1000, sometimes less.
(no change of use or maintenance)

The UG teeth did snap off though.
The 6-speed UG bike got relegated to commuting/shopping, which was flat with no need to change gear. After a couple of years in the same gear, the cables rusted up and I no longer had the option of changing. Eventually the sprocket lost enough teeth to be unusable (still with the same chain) and the bike got retired.
Image

Brucey
Posts: 36822
Joined: 4 Jan 2012, 6:25pm

Re: Cassette question

Postby Brucey » 12 Feb 2020, 1:52pm

that is a pretty good illustration of how a worn chain imposes ever more harmful stresses on the sprockets. In fact I suppose that was comparable to a singlespeed in many respects.

FWIW in a singlespeed the chain soon wears so that the chain is loading the teeth in a less than perfect fashion; in practical terms this means that the slack in the chain has to be allowed to become greater as the chain wears, else the chain won't be able to adopt the correct (outward) position on the sprocket, where the (now longer) chain pitch is the best match for the sprocket.

Provided the chain will run happily when it is slack, a chain can be used to well over 1% wear and still run reasonably smoothly on most IGH bikes. If you are using a steel chainring (between about 44 and 50T) then it usually survives this with little ill effect, and IGH sprockets are cheap, so a different strategy can work here. I reckon to get between 3000 and 5000 miles out of a cheap 1/8" chain on my (most used) IGH bike this way, with little maintenance other than occasionally oiling the chain, with little if any actual cleaning.

Horses for courses.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

Marcus Aurelius
Posts: 794
Joined: 1 Feb 2018, 10:20am

Re: Cassette question

Postby Marcus Aurelius » 12 Feb 2020, 2:19pm

Brucey wrote:that is a pretty good illustration of how a worn chain imposes ever more harmful stresses on the sprockets. In fact I suppose that was comparable to a singlespeed in many respects.

FWIW in a singlespeed the chain soon wears so that the chain is loading the teeth in a less than perfect fashion; in practical terms this means that the slack in the chain has to be allowed to become greater as the chain wears, else the chain won't be able to adopt the correct (outward) position on the sprocket, where the (now longer) chain pitch is the best match for the sprocket.

Provided the chain will run happily when it is slack, a chain can be used to well over 1% wear and still run reasonably smoothly on most IGH bikes. If you are using a steel chainring (between about 44 and 50T) then it usually survives this with little ill effect, and IGH sprockets are cheap, so a different strategy can work here. I reckon to get between 3000 and 5000 miles out of a cheap 1/8" chain on my (most used) IGH bike this way, with little maintenance other than occasionally oiling the chain, with little if any actual cleaning.

Horses for courses.

cheers

Yep agreed. You don’t need a fancy chamfered plate climber / “indexed” chain either, so the chains cost a fair bit less too.

User avatar
The utility cyclist
Posts: 2740
Joined: 22 Aug 2016, 12:28pm
Location: The first garden city

Re: Cassette question

Postby The utility cyclist » 12 Feb 2020, 6:20pm

Brucey wrote:
Marcus Aurelius wrote:
Brucey wrote:my point was in good part, there is some mismatch between even new parts.

cheers


That’s manufacturer tolerances for you.


If it were just that then there would be many more instances of new parts that just don't work with one another. I rather think that there must be a slight mismatch for it to work reliably at all even when the parts are new. That being the case there is no real justification for being precious about thinking there is no place for combinations of parts that are in slightly different wear states; in the simplest terms it either works or it doesn't.

Were this not the case then the whole derailleur scheme would be nigh-on unworkable; the sprockets in a cassette never wear evenly ( with the bulk of the wear being on just e few sprockets), so at any point once the chain is even slightly worn, you are running a chain and sprocket that are imperfectly matched to one another when you are using most of the possible gears.

In a nutshell in a derailleur system, at the sprockets, you either have;

a) the chain pitch slightly shorter than the wear state of the sprockets is best suited to, in which case the chain may skip under load, or
b) the chain pitch slightly longer than the sprockets could use.

In the latter case this arises either by design (e.g. when the parts are new) or through wear. If the mismatch becomes too large all that happens is that the 'bad' wear rate increases, presumably because the loads are more often on only one tooth at a time and there is more likely to be movement under load between the chain rollers and the sprocket teeth.

By contrast in the former case there is little reason to suppose that there is -in the absence of skipping under load- a much higher wear rate than normal, should you use a new chain on slightly worn sprockets.

Hence you should try and avoid using worn chains on good sprockets as a basic principle; it'll work, but it'll probably be wearing everything at an abnormally high rate. However (as many others can testify) if you run a new chain on slightly worn sprockets the worst that can usually happen is that the chain skips under load, which is a surefire sign that you left the chain too long before renewing it.

Slavishly destroying cassette after cassette (not to mention the chainrings) in the mistaken belief that 'they should be kept as a matched set' when in fact the chain is soon worn so that case b) applies is IMHO a flawed approach; they are never 'perfectly matched' in any event and once a chain is slightly worn you have more mismatch than is desirable, and this makes for an ever faster (and ever more harmful) wear rate.

cheers

On my commuter bike I deliberately run the inner 26 with the smaller sprockets to try to even the wear, I'm okay to 17mph with 26x12 but as the journey shifts from 8% up to 5% down with other smaller up/down variances in a fairly short space of time I have no choice but to shift to the bigger rings.
I've no idea how much effect it has on the chain/sprockets - obviously cross chaining has some affect, but I thought that trying to even the sprocket wear would be better overall.
I've been doing this now for quite a few years and decided to run the chain to its 'death', the cassette will go with it but the chain has lasted a significantly longer period than I would expect for 10 speed, more so since the bike gets used 4 seasons plus muddy conditions and cleaning regime has been shall we say negligible for 3 years.

User avatar
The utility cyclist
Posts: 2740
Joined: 22 Aug 2016, 12:28pm
Location: The first garden city

Re: Cassette question

Postby The utility cyclist » 12 Feb 2020, 6:26pm

andrew_s wrote:
Brucey wrote:Note that with 7s (BITD) the sprockets were 2mm thickness, and up to/including UG most sprockets were made of hard, relatively brittle steel that was relatively slow-wearing but would crack if you shifted clumsily under full power too often. With modern sprockets they are are made of softer steel (which may bend rather than break if you shift under full power) and are as little as 1.6mm thickness. You won't be surprised to hear that I think they wear faster, and it is more important than ever that you don't run with a worn chain

They definitely wear faster.
When I used 6-speed UG cassettes, a new chain would run without skipping at well past 2000 miles.
When I changed to 8-speed HG, the non-skip mileage dropped to around 1000, sometimes less.
(no change of use or maintenance)

The UG teeth did snap off though.
The 6-speed UG bike got relegated to commuting/shopping, which was flat with no need to change gear. After a couple of years in the same gear, the cables rusted up and I no longer had the option of changing. Eventually the sprocket lost enough teeth to be unusable (still with the same chain) and the bike got retired.

Dunno what chains/cassettes you were y=using in 8 speed but to get skipping in a 1000 miles is ridiculous and I don't think reflective of most people's usage. I can't even say how many miles I've done with a SRAM 1030 chain on the commuter/utility bike (tiagra cassette), but it's done me for at least 5, possibly 6 winters and the last 3 I've barely done anything to it apart from wipe and relube with fully synth motor oil.

Marcus Aurelius
Posts: 794
Joined: 1 Feb 2018, 10:20am

Re: Cassette question

Postby Marcus Aurelius » 12 Feb 2020, 7:22pm

The utility cyclist wrote:
Brucey wrote:
Marcus Aurelius wrote:
That’s manufacturer tolerances for you.


If it were just that then there would be many more instances of new parts that just don't work with one another. I rather think that there must be a slight mismatch for it to work reliably at all even when the parts are new. That being the case there is no real justification for being precious about thinking there is no place for combinations of parts that are in slightly different wear states; in the simplest terms it either works or it doesn't.

Were this not the case then the whole derailleur scheme would be nigh-on unworkable; the sprockets in a cassette never wear evenly ( with the bulk of the wear being on just e few sprockets), so at any point once the chain is even slightly worn, you are running a chain and sprocket that are imperfectly matched to one another when you are using most of the possible gears.

In a nutshell in a derailleur system, at the sprockets, you either have;

a) the chain pitch slightly shorter than the wear state of the sprockets is best suited to, in which case the chain may skip under load, or
b) the chain pitch slightly longer than the sprockets could use.

In the latter case this arises either by design (e.g. when the parts are new) or through wear. If the mismatch becomes too large all that happens is that the 'bad' wear rate increases, presumably because the loads are more often on only one tooth at a time and there is more likely to be movement under load between the chain rollers and the sprocket teeth.

By contrast in the former case there is little reason to suppose that there is -in the absence of skipping under load- a much higher wear rate than normal, should you use a new chain on slightly worn sprockets.

Hence you should try and avoid using worn chains on good sprockets as a basic principle; it'll work, but it'll probably be wearing everything at an abnormally high rate. However (as many others can testify) if you run a new chain on slightly worn sprockets the worst that can usually happen is that the chain skips under load, which is a surefire sign that you left the chain too long before renewing it.

Slavishly destroying cassette after cassette (not to mention the chainrings) in the mistaken belief that 'they should be kept as a matched set' when in fact the chain is soon worn so that case b) applies is IMHO a flawed approach; they are never 'perfectly matched' in any event and once a chain is slightly worn you have more mismatch than is desirable, and this makes for an ever faster (and ever more harmful) wear rate.

cheers

On my commuter bike I deliberately run the inner 26 with the smaller sprockets to try to even the wear, I'm okay to 17mph with 26x12 but as the journey shifts from 8% up to 5% down with other smaller up/down variances in a fairly short space of time I have no choice but to shift to the bigger rings.
I've no idea how much effect it has on the chain/sprockets - obviously cross chaining has some affect, but I thought that trying to even the sprocket wear would be better overall.
I've been doing this now for quite a few years and decided to run the chain to its 'death', the cassette will go with it but the chain has lasted a significantly longer period than I would expect for 10 speed, more so since the bike gets used 4 seasons plus muddy conditions and cleaning regime has been shall we say negligible for 3 years.

A lot of electronic Groupsets auto shift on the chainrings, to maintain a straighter chain line.