Topic: chain and sprocket change: how much truth there is in conventional wisdom?

[eldar]

Here is my thought. There is no way to change the wear on a chain. No matter what, every link will hit every tooth on both sprockets the same number of times. So changing the position of your chain will do nothing.

Same goes for flipping it around. It is still going to contact the same number of times EXCEPT this time, it will be going against the wear pattern which usually has the result of speeding up wear, even more so if the chain has already been worn.

[Jonesy]

Chain life is related to sprocket size. As was mentioned in the beginning, the chain doesn't physically "stretch"- the bushings in the links become worn and you multiply the slop in each link over the number of links in the chain and you get the increased overall length.

How does the wear result? If you factor out lubrication and the abrasive effects of dirt (assuming the O-rings between the plates do a perfect job), the one factor left to wear out the bushings is the mechanical movement of the link pins twisting back-and-forth within the bushings while the chain bends to go around the sprocket and then straightens back out. So, the larger the sprocket, the less the chain has to bend to engage the sprocket. Hence, Honda going form a 16t to a 17t front sprocket on the CB750.

One could also argue that chain tension is also a factor, as with a very loose chain, the chain can flop and flex more, causing more wear at the link bushings. Too tight a chain can put more pressure on the interface between the pins and bushings, causing what little back-and-forth movement there is to do more damage in the from of accelerated wear.

However, the source of debate comes from 2 things- 1) In order to gain any mechanical advantage, one sprocket must always be smaller (unless you eliminate the chain altogether with shaft or belt drive), and 2) there must always be some slack in the chain to allow for the movement of the suspension (unless you run a hardtail bike).

[Raul CB750K1]

Now there's a challenge for the bored on this board...

There you go. Right one is the standard connecting link. Left one is the "half" connecting link.

[eldar]

Dang, I would not be using that half link with a cotter pin!

[Tower]

Well dang!  a half-link chain link - who would have thought was possible.

The argument for 4X lifespan is this:  Unless the sprocket has a prime number of teeth, the same link repeatedly hits the same tooth (in my example, 4x more often than if the numbers were primed), and that this over-repetition causes that specific link-tooth combination to wear into a pattern much faster than would happen if the wear was spread more evenly across all link-tooth combinations.

In other words, all links do not necessarily hit all teeth unless there is a prime number of links or teeth involved.  Since chains are (conventionally) even numbers, it suggests that its up to the sprockets to have a prime number of teeth.

[scondon]

 Would it least be safe to assume that my 106 link chain has 105 gaps? It would help me greatly if I could hold on to this safe, familiar number as a reference point to sanity as the rest is slowly stripped away by these mind numbing equations

    Switching from and 18t front to a 17t front has me thinking my chain is going to last forever. Just seat of the pants stuff tells me this. haven't had to adjust the chain ever, so far, between tire changes. Wasn't even necessary at time of tire change but got done anyway due to wheel removal. Same sprocket manufacturer/chain type/lube as my other 836 F2 running 18/48 which is the comparison in this observation. The 18/48 chain actually has been lubed and cared for at more frequent intervals. Both chains operate on the same roads under same conditions.

[DRam]

Well dang!  a half-link chain link - who would have thought was possible.

The argument for 4X lifespan is this:  Unless the sprocket has a prime number of teeth, the same link repeatedly hits the same tooth (in my example, 4x more often than if the numbers were primed), and that this over-repetition causes that specific link-tooth combination to wear into a pattern much faster than would happen if the wear was spread more evenly across all link-tooth combinations.

In other words, all links do not necessarily hit all teeth unless there is a prime number of links or teeth involved.  Since chains are (conventionally) even numbers, it suggests that its up to the sprockets to have a prime number of teeth.

Sooooo, one could explain why my CBR drive sprocket, a 17 tooth, appears to be in better condition that the driven sprocket, a 42 tooth, using that argument.  Interesting, interesting.  The fly in the ointment though, is that this is the first time since purchasing the bike I have replaced chain and sprockets, and there is no way of being sure the PO replaced both sprockets at the same time.  Knowing him though, it's a fair bet everything was replaced together.  Perhaps I should have gone with a 17/41 combination rather than a 17/42.  I'll have to increase my riding so I can wear this chain out in a short time and try the prime number theory on the next one.

[Raul CB750K1]

Well dang!  a half-link chain link - who would have thought was possible.

The argument for 4X lifespan is this:  Unless the sprocket has a prime number of teeth, the same link repeatedly hits the same tooth (in my example, 4x more often than if the numbers were primed), and that this over-repetition causes that specific link-tooth combination to wear into a pattern much faster than would happen if the wear was spread more evenly across all link-tooth combinations.

In other words, all links do not necessarily hit all teeth unless there is a prime number of links or teeth involved.  Since chains are (conventionally) even numbers, it suggests that its up to the sprockets to have a prime number of teeth.

Does it really matter which tooth contact which link? With every sprocket revolution, each tooth will contact a link. Does it really matter that they will meet again in four revolutions or in four hundred? The chain is closed and is an endless cycle, so in 30.000 miles you will find that some teeth has contacted a link 3.000.000 times and another teeth has contacted a link 3.000.001 times. The advance is not cumulative: some teeth will go ahead, the others will get closer, and the first ones will get ahead again -in "contacts" count- but in the long run it won't make any difference.

[scondon]

 Enter a single grain of dust, or a tiny imperfection, or anything that moves the the system out of a perfect and sterile environment. Imagine that that single grain of dust resides on link #63 and every time it comes in contact with a sprocket tooth it removes X amount of microscopic material from said tooth.

  Now imagine that the link/sprocket combination is such that this link hits the same tooth, and only that tooth on EVERY rotation of the wheel. #63 link and the unfortunate tooth would wear at an accelerated pace compared to their neighbors. Now change the sprocket/link count so that the combination comes around only every third rotation. The wear is distributed among 3 teeth now and the wheel would have to revolve 3x more to match amount of wear on a single tooth compared to the first example.

 "A chain is only as strong as its weakest link". Conventional wisdom? Is "overall" wear the exact same as wear to a specific area? Does the amount of material removed make any difference whether it is in one spot or distributed among many? I have no idea, just food for thought

[Raul CB750K1]

Enter a single grain of dust, or a tiny imperfection, or anything that moves the the system out of a perfect and sterile environment. Imagine that that single grain of dust resides on link #63 and every time it comes in contact with a sprocket tooth it removes X amount of microscopic material from said tooth.

  Now imagine that the link/sprocket combination is such that this link hits the same tooth, and only that tooth on EVERY rotation of the wheel. #63 link and the unfortunate tooth would wear at an accelerated pace compared to their neighbors. Now change the sprocket/link count so that the combination comes around only every third rotation. The wear is distributed among 3 teeth now and the wheel would have to revolve 3x more to match amount of wear on a single tooth compared to the first example.

 "A chain is only as strong as its weakest link". Conventional wisdom? Is "overall" wear the exact same as wear to a specific area? Does the amount of material removed make any difference whether it is in one spot or distributed among many? I have no idea, just food for thought

That would make sense, as long as only that link gets worn and the rest gets isolated in its perfect, sterile world. But the chain moves so fast and the roads are so full of sand that, just by random events, they all will get an average amount of grit. That explanation works also when you want to explain why both tires in the same axis of your car, get worn by approximately -to the 1/1000 of an inch- the same amount, even when they ride over completely random asphalt surfaces in the course of thousands of miles.


The challenge is not to find a "half" link as Steve suggested. I challenge anybody who can show me a chain where one of the links has worn way more than the others, to the extent that it can be measured with a simple caliper.

[scondon]

The challenge is not to find a "half" link as Steve suggested. I challenge anybody who can show me a chain where one of the links has worn way more than the others, to the extent that it can be measured with a simple caliper.

 I don't think that combination of links/teeth exists on a functional machine, only in the world of theory. The example is used only to illustrate how and where wear occurs and to make one ponder whether changing the amount of combinations of contact before the chain/sprockets return to the original starting position can increase or decrease the "average" the highest point of wear.

 An "average" is just the sum of of a group of measures divided by the number of entries, right? reducing the upper and lower extremes of the #'s that are used to achieve that average would mean that you can have less "wear" but still have the exact same "average" if the "chain is as weak as its weakest link" is an acceptable postulate for determining "wear". The "weakest link" being the highest number entered in the group of measures determining "average".

 Not sure a car tire can be used as a comparable measure in this discussion, but am certainly open to the possibility

Edit: Changed the struck through line because it didn't make much sense

[scondon]

 Speaking of tires, maybe it's time to throw one on the tracks, Eldy

[Tower]

The pattern-wear problem is very real. 

By the way, it won't be one tooth or link, but rather a multiple of the lowest common denominator of teeth and links.  e.g. every 6-th tooth will show a specific wear pattern that is worse (or at least significantly unique and possible excessive) when compared to the other teeth.  And a matching multiple of links e.g. 24-links will show that same pattern.

In many cases, as was with the chain on my bike when I first got it, every 6th or 12th link had a missing or cracked bushing and every 6th tooth on the rear sprocket had a scalloped edge.  The other teeth and links were well worn also, but the failure points were quite obvious.

As @scondon described, that combination could be excessive and although the other links are also worn, this particular pattern is excessive.  Being excessive, it deteriorates faster than the others....weakest link,,,blah, blah, and its time to replace the chain.

The 4X statement is mere theory.  Obviously the whole chain wears.  But, its possible that prime-number sprockets could at least double the lifespan of both components due to leveling-out or "averaging" the wear.

In the tyre example, it would be like tyres on oval track cars.  The inside tyre gets more wear, so by switching them to the other side, longevity is increased (not withstanding other aspects).  It may not double the lifespan, but will certainly lengthen it.

[HondaMan]

Here is my thought. There is no way to change the wear on a chain. No matter what, every link will hit every tooth on both sprockets the same number of times. So changing the position of your chain will do nothing.

Same goes for flipping it around. It is still going to contact the same number of times EXCEPT this time, it will be going against the wear pattern which usually has the result of speeding up wear, even more so if the chain has already been worn.

...which brings up this point: don't ever reverse the direction of your chain, if worn, unless you also flip it inside out. The plastic deformation of the insides of the pins will begin to shear otherwise, and your chain will suddenly grow SHORTER for a time (which will drive you crazy, wondering what's going on), then it will suddenly get very uneven, and slowly develop a whole lot of slack (because after it tightened up, you backed off on the adjusters, while wondering what was going on....).

The first time I did this, I was REAL puzzled, on my way thru Texas on a long trip. I had just flipped the sprockets before I left, and had just as purposefully reversed the chain, leaving the same side in, thinking that next time, I would flip it out, then reverse it again, so I could wear it out all 4 directions on the pins. (Smart, huh?) I left home, running about 65 MPH from 7 AM until about 10 AM (speed limits were 55 then), and began noticing the bike pulling to the right (had to push a little on the throttle side to keep it straight), and the bike had a sudden drop in power, and felt strange. I pulled off to the side of the Interstate and standed the bike, found the rear wheel almost locked tight. "Great", says I, "a wheel bearing is freezing, 150 miles from nowhere". Except, they were new bearings a few months ago. Then, I noticed the chain looked and felt like a banjo string.    Thinking I was losing my mind, I loosened the adjusters TWO FULL TURNS to get the 3/4" slack that I run, and hit the road again. After lunch, it was tight again! I loosened it again, made it to Odessa that night, and slept until 10 AM.

When I packed up next day and dropped it into gear, CLANK-CLANK went the chain on the guard.        I looked under, and there was about 3" of slack and the paint was worn off of the top of the swingarm.    Back on the stand, adjust to just 1" of slack (I'm feeling like Rod Serling is right behind me, now...), spray a little lube, hit the road. The bike is shaking its head on every corner, on slight downhills, like I've got a bent wheel. 

I get to lunch (2 PM-ish), wondering if I'm gonna finish this trip with a whole bike. Stand the bike, run it in 1st gear to watch the chain: it's jumping so bad from uneveness that it finally loads up and kills the engine. So, it's gotta be this chain. I find a farm implement store and buy 100 links of #50 combine chain, put it on and hit the road again. It's a noisy chain, but gets me home 1000 miles later.

Back home, I contact Diamond Chain to have a rep come see me, under the guise of a machine I'm building (so they don't send some saleman). I tell him the REAL story, just like above, and he starts laughing at me. Then, he promises to send me some tech info, which he does, and it clearly states my opening comment: "Once a chain is run in one direction and the rollers and pins are burnished, do not reverse the chain direction. Under some circumstances, it is permissible to invert the chain, then run end-to-end reversed, as this will then engage the non-worn pin surfaces... Reverse loading of worn pins will cause plastic deformation and disintegration of the bearing surfaces inside the rollers, resulting in shortened chain length and cracked and locked bushings."

No kiddin. 

[eldar]

The other question is in what way orings affect this wear. Also, does a larger tire change the wear. Speaking of that, how about if the tire is not aired up well?

Wait, weren't we just on the other track?

[eldar]

Did I run the train off the track?