Fixed gear / drivetrain wear distribution

[silva]

I've read that people swap cheaper parts in order to make more expensive parts wear slower, ex "rotating" 3 chains on a single chainring so that the single chainring lasts longer.
The rear cog I use has a "lifetime guarantee" on condition that it is flipped every X miles (of course this guarantee is a practical non-sense - it all wears, and what's a "lifetime"?
So this is about the idea to flip or "rotate" (on their mount) chainrings / cogs (so all sprockets).
Due to the pedaling force nothing being a linear force but a fluctuating one due to 2 separate strokes every turn, certain teeths must wear faster than others.
Same for chain links, especially in case of a dividable / integer ratio alike 48/16=3, or an offcenter chainring or its mount.
So who here follows a "regime" to distribute the wear, ex flipping/rotating a chainring/cog, and which one?

 

[Cycleops]

I suspect nobody does that, I certainly don't.
How long do you think you will live and what is the price of your 'lifetime guarantee ' rear cog?

 

[fossyant]

Nope never bothered. Just replaced the fixed sprocket and chain at the same time. Not exactly expensive.

 

[rogerzilla]

Use a sprocket and ring combo that gives even, or nearly even, wear on the sprocket. 47 x 18 or 49 x 18 are optimal. 48 x 18 is marginal. 48 x 16 is as bad as it gets. A 19T rear sprocket works with most chainring sizes.

 

[fossyant]

I had a 16T and if I didn't swap the chain (move it on a link) it would run like a tractor if I'd put the chain back on. This happened once, when I'd changed the tyres for Durano Plus, and didn't have any flats for about a year, so I'd not 'removed' the wheel. The sprocket had work with the chain, and when I put the chain on again, it was terrible until I advanced the chain by 1 link. After this I deliberately moved the chain every month or so. Only happens with 'evens' sprockets.

 

[sleuthey]

Same for chain links, especially in case of a dividable / integer ratio alike 48/16=3

You have a point, and you are mathematically correct, but if this concerns you why not just shift the chain along a few links every so often?

 

[Ming the Merciless]

Use a chain with a prime number of links and you'll get even wear on your chain rings and rear cog.

 

[silva]

You have a point, and you are mathematically correct, but if this concerns you why not just shift the chain along a few links every so often?

That's what I did with the first and second chain of the bike. And that last is precisely what I'm asking here. I initially had a 48/16 and I shifted the chain 12 links "forward" direction, that 12 based on the thought that it was a quarter of the chainring teeth number. A pedal stroke causes the chainrings teeth directly on and after it to wear more, two pedal strokes then means 2 most worn and 2 least worn teeth(series), so shifting a quarter should bring the least worn teeth towards the most wearing position. The fluctuating load also causes a wear pattern within the chain, with max and min difference depending on the fractional part length of the gear ratio. My initial 48/16 setup thus was such a worst possible case. My current 47/16 should spread that out way more, limiting any shifting effort to solely the chainring.
But it needs an exact shifting, any deviation causes a number worn teeth / a few chain links to still be positioned on a more load / more worn. So there might be a system possible, that defines precisely which regime to follow for which gear case.

 

[fossyant]

Don't shift a chain an even number of links forward on an evens sprocket/chain wheel setup - just move it forward 1 link every so often. The issue is worse with evens sprockets - chain rings are much bigger so never an issue.

 

[Deleted member 1258]

Just ride the bike, change the chain when it's knackered, change the sprocket when it's knackered, you're making something simple unnecessarily complicated.

 

[sleuthey]

I completely see where the OP is coming from but like they say on the "we buy any car" add, I value my time and I'd prefer to just replace the chain and sprocket when they are worn. I know this might slightly shorten the time between changing but I'm willing to pay for convenience and I do not buy ones made from solid gold.

 

[silva]

Upon all worn, I changed all to new, and a 47 instead of 48t chainring.
That solved (proved afterwards) the problem of an increasing chain tension variation.
The basic variation, about equally sized, remains unsolved.
But since it doesn't grow, it doesn't become problematic (choice wreck bearings against risk chain falling off) it's just annoyant, extra work at retension.

 

[silva]

As a last try to get rid of that remaining part of the chain tension variation, and since I had to replace the crankset anyway due to both arms broken, I decided to let the square taper axle and cranks replace by the same Octalink system I have on my previous bike that never had such a chain tension variation.
Bad luck though - the system wasn't available with my current chainring mount bcd 144 mm so I had to chose 130 mm, rendering my stock spare rings useless and requiring me to order new, which I did couple days ago.
So over a couple weeks I will see if this last try succeeds or fails its goal.

Aside, in the past I found on two second hand occasions stainless steel chains marked "08BSS" which I decided to buy. I measured, and it looked like these would fit a bicycle. My interest in these was for winter / road salt occasions. But as I later discovered when trying to mount one - they didn't fit the chainring. Today I found their specifications and the roller has a slightly bigger diameter so these purchases (same prices as new common bicycle chains) were a fail, although not costly ones.
Today I checked specifications of motorcycle chains. 415 and 420 appear to be totally bicycle compatible (singlespeed only ofc). I checked clearance for a 420 and it should fit the bike. 415 is 3/16" wide and 420 is 1/4" wide. I checked around and an IRIS RX420 looks like a good gain/cost candidate.
My motivation is to minimalise wear at its first cause: the chain. If that wears slower then the rest will too. It's the chain that eats sprocket teeth when increased pitch due to wear, not vice versa.
This chain will have quite some play on a 1/8" sprocket, but I don't think this will give away anything substantial of the achieved benefit. Maybe this play could be seen as some lateral movement space for the chain, to compensate (in a degree) for the stiffness of such heavy chain.

 

[fossala]

I decided to let the square taper axle and cranks replace by the same Octalink ...the system wasn't available with my current chainring mount bcd 144 mm

Dura ace 7710 is octalink and 144bcd. I'd guess whatever you picked up is a road crank and not a track one. The upside of 7710 is you can get a fixed gear stages powermeter for it.

 

[silva]

Dura ace 7710 is octalink and 144bcd. I'd guess whatever you picked up is a road crank and not a track one. The upside of 7710 is you can get a fixed gear stages powermeter for it.

I asked the dealer about hollowtech and the answer was negative, given reason was a fixed axle lenght causing no chainline adjustment possibilities. My chainline is 57 mm. True / not true?