Adventures in OCD: Today's Chain Waxing

[alex_cycles]

No surprise the 9 speed is lasting longer than the 11. There's more contact area between rollers/bushing and pin the lower speed the chain is, more contact area, less wear.
Tbf you're doing well, I'd struggle to get 2000 miles out of a 9 or 10 speed chain but then I did like to be out the saddle on inclines, it doubtless puts an awful lot more force through the chain.

Since I started waxing I have yet to wear out a chain, so it's hard to say quite how well I'm doing. (Probably because I'm rotating three chains, but also I didn't do a lot of outdoor in 2023 because the weather wasn't great.)

I'm sure a 100kg rider will put more force through the chain than a 50kg rider, so that must have a bearing on things too as regards wear.

I've often said, been there, done that (seemingly ridiculous focus on extending chain life or mileage)...but it was interesting.
Eventually came to the conclusion my variety of methods made little real difference... but it was interesting finding out.

I'm "all-in" on chain waxing now. Still on my original batch of wax. Three chains for each of my three main bikes (road, gravel and indoor).
No going back for me. But I do try to avoid riding on wet roads (because I don't enjoy it).

 

[Ajax Bay]

No surprise the 9 speed is lasting longer than the 11. There's more contact area between rollers/bushing and pin the lower speed the chain is, more contact area, less wear.
Tbf you're doing well, I'd struggle to get 2000 miles out of a 9 or 10 speed chain

Agree to second experience. If 'getting out' means +0.5%. With @mickle oiling my experience is 2500km ish, but I have procured an ultrasonic cleaner and aim to begin 'dark side' waxing this year.
But is there "more contact area between rollers/bushing and pin the lower speed the chain is"?
It's the outer dimension of the chain that decreases with drivetrain speed and what you've said is true if comparing a 6-8sp chain with a 9sp or higher, but that's not the context (you compare 9sp and 11sp).
Multi speed chains from 9 to 12 speeds have inner width of 11/128″ (2.18 mm). The pins (and rollers) are exactly the same between the inner plates. So there is no more contact area and I would not assume a 9sp chain lasts any longer. Paging an expert. @Yellow Saddle
Do people think/experience that higher speed chains wear more quickly, ceteris paribus? If so, why?

 

[Ajax Bay]

I can’t get my Enigma white powder to melt in normal wax. I’ll just have to use it for brushing on afterwards.

From another place, suggesting post-waxing (and break in) powdering may not add value (reduce losses):
Quote from ZeroFrictionCyling (2020):
"OPTIMISED RACE CHAIN PREP – up till 20/11/20 this was done with moltenspeedwax (msw) and race powder. After controlled chain break in, ultrasonic cleaning x 2, ultrasonic wax, wax break in run (on rig) – then the race powdering application would occur. Recent data release with UFO drip v2 has shown Silca Hot Melt is around 1w faster than race powdered msw chain, as such FULL RACE optimised chains now undergo all the same steps but will be waxed with Silca Hot Melt and not race powdered after wax break in run)."

 

[wafter]

Agree to second experience. If 'getting out' means +0.5%. With @mickle oiling my experience is 2500km ish, but I have procured an ultrasonic cleaner and aim to begin 'dark side' waxing this year.
But is there "more contact area between rollers/bushing and pin the lower speed the chain is"?
It's the outer dimension of the chain that decreases with drivetrain speed and what you've said is true if comparing a 6-8sp chain with a 9sp or higher, but that's not the context (you compare 9sp and 11sp).
Multi speed chains from 9 to 12 speeds have inner width of 11/128″ (2.18 mm). The pins (and rollers) are exactly the same between the inner plates. So there is no more contact area and I would not assume a 9sp chain lasts any longer. Paging an expert. @Yellow Saddle
Do people think/experience that higher speed chains wear more quickly, ceteris paribus? If so, why?

Good work!

I think the main area of wear responsible for chain elongation is between the pin and the inner links that articulate around it. You're correct that the internal width between inner links and roller width are the same on 9-12 speed chains; however to reduce the outer chain width the pin length and plate thickness are reduced - hence there's less bearing area between the inner plates and pin.

From ealier in this thread, below is a picture 11sp (left) and 9sp chains - with the plates on the latter looking decidedly fatter.

As such it's a natural assumption that the fatter chains will last longer; however it's been suggested that this might not be the case as chains that cater for more speeds are likely made to better tolerances.

I run everything from 5-12 speed on various bikes, however differences in application plus the longevity afforded by the waxing process make direct comparisons difficult.. however for now it seems that the 11sp is proving most resistant to wear.

From another place, suggesting post-waxing (and break in) powdering may not add value (reduce losses):
Quote from ZeroFrictionCyling (2020):
"OPTIMISED RACE CHAIN PREP – up till 20/11/20 this was done with mspeedwax and race powder (after controlled chain break in, many ultrasonic cleaning rounds, ultrasonic wax with fresh wax, controlled wax break in run – then the race powdering application would occur). Recent data release with UFO drip v2 has shown Silca Hot Melt is around 1w faster than race powdered msw chain, as such FULL RACE optimised chains now undergo all the same steps but will be waxed with Silca Hot Melt and not race powdered after wax break in run)."

Tbh I think that you'll get 99% of the benefit with pure wax (maybe softend a bit with some liquid paraffin). While I run moly powder I'm not sure I would again as it adds significant cost, while other additives such as PTFE are crap from an environmental perspective.

 

[Ajax Bay]

I think the main area of wear responsible for chain elongation is between the pin and the inner links that articulate around it. You're correct that the internal width between inner links and roller width are the same on 9-12 speed chains; however to reduce the outer chain width the pin length and plate thickness are reduced - hence there's less bearing area between the inner plates and pin.

You may wish to rethink that analysis. Thank you for making me think about this!
The inner plates do "articulate around the pin" but the relative movement under tension (so on the final tooth leaving the sprocket) is small# and likewise wear (can see trace evidence in image (from Sheldon Brown)).
The area of wear in a chain, several magnitudes higher, resulting in its measurable elongation, is each pin (see image below), worn by the rollers, augmented by less wear of the rollers*.
On an 11sp chain the overall pin length is shorter (because the outer chain width has to be less to manage the reduced sprocket spacing) but the section of the pin (9sp - 11sp) experiencing stress causing friction and thereby wear is the same: < 2.18 mm so the rollers will be 2mm).

Edit - added (from https://bike.bikegremlin.com/3555/bicycle-drive-chain-dimension-standards/ )

*Which also wear a bit (internal radius) but that wear is spread around whereas the relevant wear on/of the pin is only in the sextant that always takes the maximum stress (and with chain articulation, friction thus wear).
# The smaller the sprocket the greater articulation and thus wear (with associated reduction in chain efficiency btw).

 

[wafter]

You may wish to rethink that analysis. Thank you for making me think about this!
The inner plates do "articulate around the pin" but the relative movement under tension (so on the final tooth leaving the sprocket) is small# and likewise wear (can see trace evidence in image (from Sheldon Brown)).
The area of wear in a chain, several magnitudes higher, resulting in its measurable elongation, is each pin (see image below), worn by the rollers, augmented by less wear of the rollers*.
On an 11sp chain the overall pin length is shorter (because the outer chain width has to be less to manage the reduced sprocket spacing) but the section of the pin (9sp - 11sp) experiencing stress causing friction and thereby wear is the same: < 2.18 mm so the rollers will be 2mm).
View attachment 719000

*Which also wear a bit (internal radius) but that wear is spread around whereas the relevant wear on/of the pin is only in the sextant that always takes the maximum stress (and with chain articulation, friction thus wear).
# The smaller the sprocket the greater articulation and thus wear (with associated reduction in chain efficiency btw).

lol - the plot thickens!

From that image your point about the wear distribution seems reasonable. However, I'd counter that the problematic element of chain wear is the change in effective pitch between the rollers; which (IIRC) allows the leading roller on the sprocket to ride up the tooth, wearing this area and getting progressively worse to the point where it eventually slips and can no longer transmit any force to the sprocket.

Assuming wear on each roller / the corresponding face of each pin is even across every one in the chain, and that all rollers are all loaded in the same direction during use (which they are) then the effective pitch between them doesn't change - so wear in this area effectively cancels itself out.

Also, wear between the rollers and pins does nothing to contribute towards the overall elongation with the chain's length that we all associate with chain wear.

Hence, the pins could be worn half-way through their diameter but the effective pitch between the rollers would remain constant at the original half inch, while the chain length would also be unaffected.

The only way for the chain to get longer is if the inner and outer plates move relative to each other / the centreline of the pin. This can only happen through wear between the moving parts (inner plate and pin) or stretch in the plates; the latter being unlikely since in normal use the forces involved should be well below the yield stress of the material.

Pity I don't have an old knackered chain kicking about to chop up for science..

 

[Yellow Saddle]

Agree to second experience. If 'getting out' means +0.5%. With @mickle oiling my experience is 2500km ish, but I have procured an ultrasonic cleaner and aim to begin 'dark side' waxing this year.
But is there "more contact area between rollers/bushing and pin the lower speed the chain is"?
It's the outer dimension of the chain that decreases with drivetrain speed and what you've said is true if comparing a 6-8sp chain with a 9sp or higher, but that's not the context (you compare 9sp and 11sp).
Multi speed chains from 9 to 12 speeds have inner width of 11/128″ (2.18 mm). The pins (and rollers) are exactly the same between the inner plates. So there is no more contact area and I would not assume a 9sp chain lasts any longer. Paging an expert. @Yellow Saddle
Do people think/experience that higher speed chains wear more quickly, ceteris paribus? If so, why?

Most amateur "findings" of what chain wears better are based on anecdotes and bias. It is extremely difficult to make such a judgement on your personal experience. This is because of several reasons. I'll name a few:

1) You don't put in enough miles to get enough samples in the experiment.
2) Any "experiments" are just vague recollections.
3) Your conditions are not controlled.
4) Your method is not standardised and repeatable.
5) You pick up bias from ignorant bicycle writers working for underfunded online magazines with no scientific background.
6) All sorts of other biases - just list the various fallacies and try and find one that's NOT applicable.


Chains above 9-speed have thinner outer plates-plates and flat, countersunk rivets. This is clearly visible in the photo above. Countersunk rivets have problems of their own in that the countersink tends to crack open outer plates because of the compressed wedged pin forced into a thin, countersunk hole. All things being equal, all these chains have the same wear rate.

However, all things are not equal. Chains have improved a bit thanks to better steel and better heat treatment. This is also reflected in the cost of these chains at a factor much higher than their slightly improved longevity.

 

[Yellow Saddle]

lol - the plot thickens!

From that image your point about the wear distribution seems reasonable. However, I'd counter that the problematic element of chain wear is the change in effective pitch between the rollers; which (IIRC) allows the leading roller on the sprocket to ride up the tooth, wearing this area and getting progressively worse to the point where it eventually slips and can no longer transmit any force to the sprocket.

Assuming wear on each roller / the corresponding face of each pin is even across every one in the chain, and that all rollers are all loaded in the same direction during use (which they are) then the effective pitch between them doesn't change - so wear in this area effectively cancels itself out.

Also, wear between the rollers and pins does nothing to contribute towards the overall elongation with the chain's length that we all associate with chain wear.

Hence, the pins could be worn half-way through their diameter but the effective pitch between the rollers would remain constant at the original half inch, while the chain length would also be unaffected.

The only way for the chain to get longer is if the inner and outer plates move relative to each other / the centreline of the pin. This can only happen through wear between the moving parts (inner plate and pin) or stretch in the plates; the latter being unlikely since in normal use the forces involved should be well below the yield stress of the material.

Pity I don't have an old knackered chain kicking about to chop up for science..

"From that image your point about the wear distribution seems reasonable. However, I'd counter that the problematic element of chain wear is the change in effective pitch between the rollers; which (IIRC) allows the leading roller on the sprocket to ride up the tooth, wearing this area and getting progressively worse to the point where it eventually slips and can no longer transmit any force to the sprocket."

That's not how it works. A worn chain should be examined from two scenarios: a derailer bicycle and a single-speed. On singlespeed without a tensioner, the chain doesn't enter the sprocket under slack provided by a spring-loaded derailer. Therefore, any discrepancy in pitch between sprocket and is forced away. The chain will, even if elongated, get forced onto the sprocket. In this scenario there is no slip, even at high wear conditions. The cogs (teeth) will literally wear to a thin strip and then break off.

On a derailer bike this doesn't happen. The chain enters the sprocket under slack and should it be mismatched with the sprocket, it will not engage and fall into the sprocket but ride over the top and on the power stroke, skate over the top of the sprocket. In other words, it will start to skate long before wearing the teeth to slivers and therefore cannot wear the teeth beyond a certain point.

"Assuming wear on each roller / the corresponding face of each pin is even across every one in the chain, and that all rollers are all loaded in the same direction during use (which they are) then the effective pitch between them doesn't change - so wear in this area effectively cancels itself out."

Rollers don't come into contact with the pins. Contact is between pins (rivets) and inner plates. The latter is swaged into a bushing of sorts. The wear takes place between pin and inner plate only. Note that the outer plates do not wear because the rivets are fixed to them. You do mention that elsewhere but there seems to be a contradiction in your text.

Plates do not stretch through tensile forces, as you mentioned.

 

[Yellow Saddle]

If I may: Chain mechanics is expertly discussed in this cheap book:

Click here .

PS - Wax is crap.

 

[Ajax Bay]

The only way for the chain to get longer is if the inner and outer plates move relative to each other / the centreline of the pin. This can only happen through wear between the moving parts (inner plate and pin) or stretch in the plates;

I have succumbed and cut a link off a new chain (had 57 and I only need 56). All dimensions to the accuracy my calipers allow/afford but may also be relative, which is relevant.
A pin's diameter is 3.61mm.
The holes in an inner plate are circa 3.69mm (outer's are <3.5mm btw which is why they have to be pushed in with force) so there's already 'wriggle room' there, but the pitch is maintained (or not with wear) by the pin diameter (along the chain's axis).
In use, chains articulate (at most, 11t sprocket) 33 degrees.
Each half link (12.5mm long) only has to wear 0.07mm before the chain is elongated 0.5%. That wear is a combo of pin wear plus side plate hole wear: means more slack at the inner plate hole.
I assume the pin is made of softer steel than the sideplates (so wears more).

Rollers don't come into contact with the pins. Contact is between pins (rivets) and inner plates. The latter is swaged into a bushing of sorts. The wear takes place between pin and inner plate only.

It's clear (from measurement) that (as said) rollers don't come in contact with the pins (I always assumed they did) but roll on the swaged (sort of) bushings. Please tell me/us, then, what causes that f***ing great groove in the Sheldon image above.
Why aren't there two grooves in each worn pin? What causes that central groove?

 

[wafter]

"From that image your point about the wear distribution seems reasonable. However, I'd counter that the problematic element of chain wear is the change in effective pitch between the rollers; which (IIRC) allows the leading roller on the sprocket to ride up the tooth, wearing this area and getting progressively worse to the point where it eventually slips and can no longer transmit any force to the sprocket."

That's not how it works. A worn chain should be examined from two scenarios: a derailer bicycle and a single-speed. On singlespeed without a tensioner, the chain doesn't enter the sprocket under slack provided by a spring-loaded derailer. Therefore, any discrepancy in pitch between sprocket and is forced away. The chain will, even if elongated, get forced onto the sprocket. In this scenario there is no slip, even at high wear conditions. The cogs (teeth) will literally wear to a thin strip and then break off.

On a derailer bike this doesn't happen. The chain enters the sprocket under slack and should it be mismatched with the sprocket, it will not engage and fall into the sprocket but ride over the top and on the power stroke, skate over the top of the sprocket. In other words, it will start to skate long before wearing the teeth to slivers and therefore cannot wear the teeth beyond a certain point.

OK, that's going to take some thinking about to respond to in any meaningful sense.

"Assuming wear on each roller / the corresponding face of each pin is even across every one in the chain, and that all rollers are all loaded in the same direction during use (which they are) then the effective pitch between them doesn't change - so wear in this area effectively cancels itself out."

Rollers don't come into contact with the pins. Contact is between pins (rivets) and inner plates. The latter is swaged into a bushing of sorts. The wear takes place between pin and inner plate only. Note that the outer plates do not wear because the rivets are fixed to them. You do mention that elsewhere but there seems to be a contradiction in your text.

Plates do not stretch through tensile forces, as you mentioned.

Good point; of course you're right re. roller and pin contact. Not sure where I've contradicted myself but yes, I agree with the above.

I have succumbed and cut a link off a new chain (had 57 and I only need 56). All dimensions to the accuracy my calipers allow/afford but may also be relative, which is relevant.
A pin's diameter is 3.61mm.
The holes in an inner plate are circa 3.69mm (outer's are <3.5mm btw which is why they have to be pushed in with force) so there's already 'wriggle room' there, but the pitch is maintained (or not with wear) by the pin diameter (along the chain's axis).
In use, chains articulate (at most, 11t sprocket) 33 degrees.
Each half link (12.5mm long) only has to wear 0.07mm before the chain is elongated 0.5%. That wear is a combo of pin wear plus side plate hole wear: means more slack at the inner plate hole.
I assume the pin is made of softer steel than the sideplates (so wears more).

It's clear (from measurement) that (as said) rollers don't come in contact with the pins (I always assumed they did) but roll on the swaged (sort of) bushings. Please tell me/us, then, what causes that f***ing great groove in the Sheldon image above.
Why aren't there two grooves in each worn pin? What causes that central groove?
View attachment 719007

Good work and that all makes sense. The groove Sheldon's image above is the result of wear from contact with the two inner plates; which as we've been reminded don't contact the pin at the outsides of the roller that sits between; rather effectively form a bush that contacts the central portion of the pin's length as bounded by the inside faces of the two inner plates

An interesting aside of this correction being that the wear area of the higher-speed-count chains is proportionally greater than suggested by our incorrect assumption earlier about roller / pin interface and plate thickness...

 

[alex_cycles]

All this talk about chain wear I am reminded that the "change the chain" threshold is 0.5% for 11-speed plus and 0.75% for 10-speed and lower.
Does this contribute towards the "impression" that the 7,8,9 speed chains last longer?

Also gonna need a bit more than "waxing is crap" before I ditch my crock-pot
It seems to ride well and the chain wear is good and it doesn't get oil everywhere.

 

[Sea of vapours]

Also gonna need a bit more than "waxing is crap" before I ditch my crock-pot
It seems to ride well and the chain wear is good and it doesn't get oil everywhere.

Indeed so. Me too.

As I've doubtless said above, the cleanliness alone makes immersive waxing a no-brainer choice for me, but having a chain that's only half way to 0.5% wear after 10,000km is a nice bonus too :-)

 

[Ajax Bay]

All this talk about chain wear I am reminded that the "change the chain" threshold is 0.5% for 11-speed plus and 0.75% for 10-speed and lower.

Where do you get those %ages from? In particular where do you get the different thresholds for <=10 and =>11sp?
I assume it's this Park Tools article: https://www.parktool.com/en-int/blog/repair-help/when-to-replace-a-chain-on-a-bicycle
"Every drivetrain manufacturer has their own specifications, so consult with the manufacturer for the recommended wear level for your system. Below are some general guidelines. Your chain should be replaced by the time it reaches the wear value specified.
"Single-speed 1.0%
10-sp or less 0.75%
11-sp or more 0.5%"
However Park Tools offer no rationale for the different %. Any idea?
Earlier in that article they say:
"Method 2: Measure with Ruler: 24 rivets and your last rivet should be at the 12″ mark of your ruler. If it is off by more than 1/16″ [0.0625" so 0.5%] your chain is stretched to the point of replacement."
KMC FAQs: "When the elongation reaches 1%, it is the time to change your chain." (Comment: do that and the new chain will slip on the by now worn most used sprockets ime)
SRAM: "When should I replace my SRAM Eagle [12sp] chain?
"Replace the chain when it reaches 0.8% wear using a Pedro’s Chain Checker Plus, Pedro’s Chain Checker Plus II, or a Park Tool CC-4 tool to measure chain wear." (Comment: Maybe SRAM probably thinks that Americans can't handle 0.05 type number so they've gone for a single digit (8).)
BikeRadar: "anything past 12 1/16in (0.5%) would be the time to replace a chain. Anything past 12 1/8in (1.0%) has been worn to death and so a new cassette is likely needed."
Bike Gremlin (Relja) "When the chain gets elongated 0.5% of its length, it is time for a new chain. If the chain is changed [then], the life of chainrings and sprockets will be longer. Rear sprockets will last from 2 to 4 chain changes this way, before they too get worn and need a change."

Back on the 'where and how a chain wears' I found another image on Bike Gremlin (Relja) which show the wear (two grooves) on the pin by the two sideplates(+bushings). https://bike.bikegremlin.com/3306/bicycle-chain-wear-elongation/#4

 

[alex_cycles]

Where do you get those %ages from? In particular where do you get the different thresholds for <=10 and =>11sp?
I assume it's this Park Tools article: https://www.parktool.com/en-int/blog/repair-help/when-to-replace-a-chain-on-a-bicycle
"Every drivetrain manufacturer has their own specifications, so consult with the manufacturer for the recommended wear level for your system. Below are some general guidelines. Your chain should be replaced by the time it reaches the wear value specified.
"Single-speed 1.0%
10-sp or less 0.75%
11-sp or more 0.5%"
However Park Tools offer no rationale for the different %. Any idea?
Earlier in that article they say:
"Method 2: Measure with Ruler: 24 rivets and your last rivet should be at the 12″ mark of your ruler. If it is off by more than 1/16″ [0.0625" so 0.5%] your chain is stretched to the point of replacement."
KMC FAQs: "When the elongation reaches 1%, it is the time to change your chain." (Comment: do that and the new chain will slip on the by now worn most used sprockets ime)
SRAM: "When should I replace my SRAM Eagle [12sp] chain?
"Replace the chain when it reaches 0.8% wear using a Pedro’s Chain Checker Plus, Pedro’s Chain Checker Plus II, or a Park Tool CC-4 tool to measure chain wear." (Comment: Maybe SRAM probably thinks that Americans can't handle 0.05 type number so they've gone for a single digit (8).)
BikeRadar: "anything past 12 1/16in (0.5%) would be the time to replace a chain. Anything past 12 1/8in (1.0%) has been worn to death and so a new cassette is likely needed."
Bike Gremlin (Relja) "When the chain gets elongated 0.5% of its length, it is time for a new chain. If the chain is changed [then], the life of chainrings and sprockets will be longer. Rear sprockets will last from 2 to 4 chain changes this way, before they too get worn and need a change."

Back on the 'where and how a chain wears' I found another image on Bike Gremlin (Relja) which show the wear (two grooves) on the pin by the two sideplates(+bushings). https://bike.bikegremlin.com/3306/bicycle-chain-wear-elongation/#4

View attachment 719054

Park Tool instruction book that came with the gauge. https://www.parktool.com/en-int/product/chain-checker-cc-2
I see they print it on the tool now too...

I don't use the gauge any more though since I own a nice pair of Mitutoyo Calipers and found them a much better way to measure.