Chainring wear on a fixie

[silva]

Discovered something surprising.
The first years I used my current fixie, I had the regime of let things wear, flip now and then, then replace all 3 drivetrain parts.
Begin 2019, I mounted the last new chainring.
Begin 2020, I had to replace the chain (eccenter at its end) and decided to let the worn chainring in place, and added a spare to my equipment in case teeth broke off along the road.
Oktober 2021, I replaced the chain again with a new, and due to the delivery problems caused by the lockdowns, I decided to still leave the worn chainring there.
It's now end may 2022, the chain wore faster than usual (tried a non oil regime, see other topic), and the eccenter sits close to its end.
But the chainrings state is basically unaltered since begin 2020.
Apparently, a new chain doesnt wear an old chainring.
Meaning that when you replace your chain before chainring "continues" wearing, a chainring almost lives forever.

Since some days, I heard a rattle at the chainring. Has been long ago I heard that. Out from experience, I know that this signals that it's time to flip and/or rotate the chainring on its mount, which I did today.
I inspected the chainring while doing it. Remarkably, the teeth were barely worn more than begin 2020.
Meaning that the chainring nearly ceased to wear further.

This indicates that it is the longer becoming chain that inflicts nearly all the wear of the chanring.
Last year I played with the idea of a wider (1/4" motorcycle sprocket, but didn't find a compatible one).
This probably would have proved having been a useless idea, because it wouldn't affect chain wear, and thus its elongation, in any significant way.

And another final element I started to wonder about.
Chainring teeth wear towards a sharkfin shape.
The fin of the teeth is directed backwards of the bike.
Never stood still about that so far.
But now I did..
What happens at the front cog - the chainring, the cranks, the pedals?
Your feet press the pedals. The crank makes the spider turn forwards, the chainring mounted on it rotates forward.
The front edge of the teeth of the chainring thus press against the "back" position N(they can roll...) of the rollers of the chain.
Agree?
So, shouldn't the front edge, being the against a chain link roller pushing edge, of the teeth wear?
Wear = removed material.
So, the "after wear remaining teeth part", the "last material to wear away", should be at the back of the teeth, not the front, and the teeth thus should have a pointed forward instead of backward sharkfin shape, no?
Apparently, it is the opposite, the sharkfin shape points backwards, alike the rollers of the chain push the chainring teeth at their trailing edge, Alike the bike / rear wheel/cog is pushing forward the bike, instead of the cranks, clearly not the case.
Or, I'm wrong somewhere in this thinking.

 

[classic33]

How much do the front of the chain links push against the rear of the teeth on the chainring?

 

[silva]

How much do the front of the chain links push against the rear of the teeth on the chainring?

But at the chainring, the force comes from the chainring, not the chain.
The chain links "follows" the teeth, not vice versa.

The rider applies a force on the cranks, which is transferred along the spider and chainrings teeth FRONT edges to the rollers of the chain links.
Then, at the rear wheel of the bike, vice versa, the chain links rollers transmit the force to REAR edges of rear cogs teeth.

At least, that's how I see it?

So I'd expect the sharkfin shaped result of wear,
- for a chainring / front cog, to point forward (which apparently isn't the case).
- for a rear cog, to point backward.

About the main subject, a bit further thinking.
It even has a benefit to keep a seriously (end position of tensioner reached) worn running.
Because if you fit a new chainring, it will, "has to", wear with the new chain, again.
So again sharkfin shapes, rollers have to "jump" over hooks to disengage, aggravating wear, AND thus the need to flip / rotate the chainring on its mount, again X times during the life cycle of the new chain.
I didn't realize until yesterday, that it has been long ago that I had to flip the chainring.
Well, this must have been the reason. No extra wear on teeth = no extra sharkfin shape = no series of audible ticks. Since that new chain then, now closed in on tensioner end position, the chainring started to wear further again.

Yet another thing: a new chain, has a same default distance between rollers, the pitch.
During wearing, that distance increases.
A worn chainring, gives rollers more space. Removed material increases gaps.
With a new chain on a worn chainring, at first only a single tooth is transmitting the force to a single roller, the other teeth are "out of sync" with the other rollers.
One may expect that single tooth, to wear more due to higher load due to no force distribution over more teeth. But wait, a tooth will also be "saved" from helping other teeth. Instead of X cycles with load Y, 1 cycle with load X * Y. So, no net change of overall wear.
If this wasn't the case, my chainring would have just worn further like before, with the new chain.

I expected the ring to just wore further with the new chain, but as now shown: it didn't.
Hence the conclusion: the chain wears the ring.
If you think about this, single speed chainring producers could design these rings different: always thicker (to postpone teeth break off), and with the short teeth that are result of regular flipping rings on their position.
It would save the flipping hassle that is required to prevent sharkfins and the much aggravated wear these cause.
Drawback of this idea being of course the thicker rings. Velosolo produces them from 4 mm plate.

 

[Roseland triker]

If the teeeef look a bit sharky there done.

 

[silva]

If the teeeef look a bit sharky there done.

Well no, you can flip it, then the other edge of the teeth wears.
The trick is to avoid sharkfin shapes by flipping everytime it becomes noticable.
That way, the teeth become symmetrically shorter.
The problem with the sharkfin is that the rollers have to be forced over the "hooks", that increased force hollows out the sharkfin, aka, makes it even sharper, aggravating the problem in a self reinforcing way.
The key is to stay in the very beginning of that self reinforcing trend.

So it's possible to keep using an old chainring that got worn towards the max chain length.
Wins:
- no chainring replacement cost
- no flipping / rotating work since no additional wear.
Cons:
- at some point, inevitably since there is always *some* wear, a tooth will break off.
Never had that so far, but never kept a chainring running like that. So I don't know if you can continue riding home with a chainring with 1 or more broken off teeth, but likely if 1 breaks, the rest will soon follow. So a spare is needed.

Btw, the rear cog (bolt on chromoly type from Velosolo) is a same story. I have replaced these 3 times so far, and I kept all 3 worn ones, keeping them with me in my luggage. The current is more worn than my spares, still all teeth after a year. It just confirms the chainring story.

If the teeeef looks a bit sharky, then it's time to flip. Then the other side of the teeth will wear towards abit sharky. Repeat. And so on. Gradually, the initially abit wider top will become smaller too, due to the remaining bit extra wear when disengaging.
Probably, once the teeth top will at some point be completely gone, not sure what will happen then, skipping or teeth height dropping?

 

[silva]

Another followup 8 months later, still that same Velosolo 1/8" 47T chainring mounted begin 2019.
It didn't wear further, or so little that it's not noticable. I never expected but it proved so.
The clear conclusion is that the teeth of the ring nearly don't wear at all due to the pedaling force, but due to chain becoming longer due to worn-off material of parts of it, which in turn frets the chainring teeth to accomodate the new length.
The huge benefit of this is that I don't need to flip chainring anymore, once the teeth have worn so much that chain needs replacement.
An obvious idea then is to grind off the teeth of a new chainring directly to that top length, or, if Velosolo would cnc design its chainrings with shorther teeth / wider valleys between them, such a ring would last a mulitude of current life.
One drawback: chains apparently wear faster, no idea why, about 1/3th shorter life. But at a ring price 3-4 times my motorcycle chain, and saved from all the flipping work, I consider it a small con.

 

[silva]

Another follow up and luck saved me a long walk home.
I use a motorbike chain since about 5 years;
I assumed I never had to worry anymore about breaking links, which was true till today.
This morning, upto work, drivetrain made alot clicks and bangs. I had to refrain pushing pedals hard since it was clear something was on the verge of disaster.
At work, check, wooha, my rear cog, had 10 of 16 teeth broken, and my chainring 1.

No big surprise, since the latter is extremely worn, mounted since 2019 and the cog was also near its end. But such disastrous fast give up of so many in such a short time. Over the noon at work i replaced the cog with a mid life worn one. I thought that would make noise mostly go, but when returning from work, not that worse like in the morning but still abnormal noise. And it got worser with the km.

Since my chain was near its end of line (eccenter close to end position) a couple weeks ago I made ready a spare and kept it in my luggage.
Since the noise was still abnormal, I decided to replace the chain already along the road, to avoid the mid life worn one being eaten up or whatever what was happening.
I searched for the quick link, and wooha, its thick side plate on the closed side was broken in the middle. 3 mm thick steel.
This must be some stress cracking, likely due to the open side, where you shift the other plate over the pins, then a clip over the groved ends. Because there is no riveting there, so there must be some play / tolerance, which allowed those pins to move abit away from eachother, which in turn bends the closed side plate a little, until metal fatigue does it break.

So that advice new chain new quicklink, that I'm gonna follow next time. I reused this one 3 times and that was beyond it. I reused it because I couldn't see any signs of wear, and also measured the thicknesses of the pins end several times on the place where the side plates go over.
A luck that it broke upon chains near end wear life, otherwise I wouldn't have had a spare chain with me. I couldn't get the closed part, now parts, of the quicklink off, because they were bend open, jammed, and I must have ridden 30 km with a link with a broken plate, that held up only because it was ment for motorcycles so way overdimensioned for bicycles. The spare quick link alone wouldn't have saved my day.

I first thought the still present noise was due to the 1 chainring tooth that broke off, somehow that the rollers build up force to then slip over it, so I grinded the tooth remainder off, but as proved later, it wasn't that.

After replacing the chain, the drivetrain all of sudden not any sound, despite chainring now a 46T.

 

[classic33]

Just because you ground the remains of a chain wheel tooth, leaving one less doesn't resize the chainwheel.
I've a couple of chainwheels where every second tooth has been removed, by myself. The chainwheels are still the same physical size.

You'll have replaced the chain wheels and rear cog with something that the motorcycle chain can sit on, without excessive sideways movement?

Uncertain what you're trying to prove or to who.

 

[silva]

Just because you ground the remains of a chain wheel tooth, leaving one less doesn't resize the chainwheel.

Of course not, didn't say that either, I still heard noise after replacing the cog with 10 of 16 teeth broken, with 9 of them very recent, that day and/or the day before, and the drivetrain alot irregular noise, although I noticed a noise I thought related to sprocket teeth, since it was every 5 or 6 rotations, with the 5 the noice when pushing left pedal down and the 6 when right pedal or vice versa not sure.
So, I first thought the wear profiles of specific teeth of the rear cog were responsible but since the cog replacement ended in precisely the same, that explanation was ruled out.
That left the chainring as an explanation and I first thought rollers of the chain had to overcome the remainder of the single (of 47) broken tooth each time, which could have been solved by removing that remainder completely.

And finally, when the noise again increased, just a day later, it was clearly a chain issue, which was proved by the quicklinks broken plate. Snapped in the middle, yet the bike still rideable, due to the equally thick other plate which basically held together the other side of the chain for the moment being of course, since the leverage would wear those pin ends quickly out of their holes.

I've a couple of chainwheels where every second tooth has been removed, by myself. The chainwheels are still the same physical size.

That's an interesting idea, was that to make a very worn chain run as easy as a new one since the widened "valleys" between teeth are simulated?

You'll have replaced the chain wheels and rear cog with something that the motorcycle chain can sit on, without excessive sideways movement?

Uncertain what you're trying to prove or to who.

The motorcycle chain is compatible with bicycle drivetrain parts, same pitch, same roller diameter, only one "internal" difference, that is, the pin diameter is with 3.97 mm some hundreds of mm bigger, as I discovered when I found the pin jammed in my Velosolo YC-324 chain tool. I solved it by hammering it out and drilling out its hole with a 4 mm drill to avoid a repeat.
Sideways movement is not a problem, even a benefit in my case since it can compensate for the eccenter module / clamp in the frame eventually not precisely in the middle anymore after tensioning.

What I (think) was proved here was that a quicklink has as weak point (alot tech specs refer it as having half the strength of a common link) being the closing plate able to just shift over the pins, without any resistence (thus "easy", no chaintool needed). That implies bigger play, and thus allows the pin ends there to move abit away from eachother every load cycle, and since the other pin ends are peened, less play, the there then leveraged force bends the other plate abit every load cycle, until metal fatigue develops a crack in the middle, that grows until the remainder of the local cross section of the plate cannot doesn't suffice anymore to bear the load.
See, it's not like that I had swallowed a can spinach to then have Popeye Powah that breaks tank chains eh.