MajorMantra

A mechanic at local repair shop told me yesterday that with the tight tolerances of 10 speed systems, it was more trouble than it was worth to try and use multiple chains with a cassette. I had commented that my chain was very near 1/16" wear and that I was planning to change it soon and he said it was better to run them into the ground together.

Thoughts? This goes against everything I've read on the subject, and I'm keen to get the most out of my cassette.

BCRider

Multiple chains of different brands/models perhaps. But if you like to spread out the gearing wear by using a regular rotation of two or three of the identical make and model chain I can't see where that would be an issue.

canopus 

I disagree. Maybe, if you have all the money to blow on cassettes, sure, run them into the ground, if not, get a new mechanic.

powers2b

Your "mechanic" is an idiot and should not be let anywhere near your bike.
You should buy a chain tool and replace your chain regularly (they are disposable wear items) if you want your cogs to last.
Changing a chain rarely requires any readjustment.

Enjoy

MajorMantra

I wasn't thinking of rotating chains (can't really be bothered) but rather replacing them before excessive wear (>1/16" over a foot) makes this impractical, as recommended by Sheldon Brown. This way I'd expect a cassette to see maybe 2 or 3 chains in its lifetime rather than just one.

BCRider

Ah OK... Still, I'd likely want to use the same brand and model anyway.

Even riding them until I notice a drivetrain noise and begin to suffer from shifting issues I typically get two chains per cassette anyway. If you do it your way and change when it wears to maybe 1/32 I don't see why you can't make your other components last for up to maybe 4 chains before you need to replace the whole lot.

Kotts

Opinions differ on this one. Here are the arguments each way:

Changing Chains

• Replacing the chain at the first sign of significant wear extends the life of the casette, and to a lesser extent, the chainrings.
• Replacing the chain regularly reduces the cash outlay for maintainance at any one point.

Running chans and cassette until trashed

• A slightly worn cassette will cause rapid wear of a new chain, until it is "broken in" to the wear on the cassette, so the new chain has a shorter life.
• A new chain on an old cassette accelerates wear on the cassette, because the load isn't spread out the same way as it was with the chain that created the wear in the first place, and has thuse "worn into" the wear patern.
• A new chain on a worn cassette doesn't engage properly, possibly leading to shifting problems.
• The total cost expenditure over time is greater with swapping chains than it is with running a chain and cassette set into the ground, then replacing both together.

All that being said, I change chains, mostly because the idea of running anything mechanical into the ground bothers me.

operator

Rotating chains is more effort than I have time for and more effort than any customer will want to have. Those that have their bikes serviced at a decent shop exclusively will have already been told at the appropriate times to replace chains before it wears a $400 dura ace cassette out. And if not, usually it only needs to happen once before they learn.

DMF 

The point is, the rationales Kott just outlined are as valid for 10-speed as for 7-speed.

The mechanic is perfectly entitled to his opinion, but it is just that - an opinion.

dabac

But I don't see what he's getting at. Yes, the more gears, the more precise the derailer adjustment has to be. But why would replacing the chain upset the RDs ability to position the chain over the right sprocket?

Admittedly, if I'm going to the effort of replacing the chain I'm quite likely to give the whole drivetrain a once-over. But my commuter has seen at least 3 chain swaps w/o me touching the RD adjustment.

Now, if he was talking about pulling the wires, cleaning, regreasing and rethreading them, then I'd be entirely on his side...

DaveSSS 

A lot of incorrect statements have been posted regarding chain and cog wear.

If you replace a chain with a new one, but leave the old cassette, it will NOT wear the cassette any faster. That notion is totally false. The used cassette will also not wear the chain any faster - another common notion that is false. The only thing that does wear the cogs faster is continuing to use a chain with excessive elongation.

If you get new-chain skip on one or more of the most-used cogs (of a used cassette), while pedaling with a high torque (like pedaling standing produces), then the cassette is trash.

The most cost effective method for managing chains and cassettes does not have a single black and white answer. It depends on the cost of the cassette, the material used for the cogs and the cost of the chain.

If you're using a $30 cassette with $15 dollar chains, then running a single chain into the ground and trashing the cassette in the process, might make economic sense. If you're paying $40-50 for a top of the line chain and using a cassette that cost $200-300, then there are better options.

Even is you toss chains when they have .5% elongation (properly measured with a precision rule, not a chain checker). There will come a point when a new chain skips on some of the worn cogs. That might happen when the third chain is installed or perhaps the fifth or sixth, but it will happen. It depends a lot on how well the use is distributed among the cogs. If only 2-3 cogs see most of the use, those will cogs will produce new-chain skip earlier.

I avoid new-chain skip with a 3-chain rotation. It may be more trouble than most people want to go to, but it's cheaper than using 6 chains for the same mileage and tossing those chains prematurely in an effort to protect the cassette. Even with 11 speed, I expect at least 12,000 miles from a cassette and 3 chains.

Ti cogs will last about half as long as steel, so keep that in mind when considering the purchase of a cassette that costs 2-3 times more, with some Ti cogs.

FBinNY 

The beauty of rotating three or four chains is that it combines the benefits of the use it till it dies and the replace chains while there's still time approaches. Since the relative age gap between the chains and the cassettes is kept relatively narrow, the chains will run smoother, and all the chains can be run until they die, a process that will take roughly 2-3 times longer than if only one chain were run to that point.

The drawback, unless your chains have reusable master links is the cost of the links, and in some cases the reliability of pin closures after too many replacements.

Alternate approaches for minimum cost per mile include increasing the allowable stretch before chain replacement with the first going at .5%, the second at .6%, the third at .7% and the last until the cassette dies (my preference) Or replacing chains at about .5%, but saving them and putting one back on again when to squeeze those last few miles out of a worn cassette.

The thing that makes the least sense is to replace a chain that's running OK with maybe 1% stretch, and then find you have to replace the cassette too. At that point, you might as well leave the old chain on and run both together for as long as possible.

One thing to consider is that the rate of cassette wear increases as the chains stretch, but the rate of chain wear generally decreases, because wear increases the size of the bearing contact area in the pins.

As DaveSSS said, no one approach is absolutely best, or worst, it depends on the relative cost of chains vs cassettes and chainrings, and labor for those who don't do their own work. There are also other considerations, for instance chains develop more side flex with age, and that can degrade shifting especially on smaller sprockets. This is less of a problem for those the RD set up close with narrow cassettes, but worse with wide range cassettes.

In defense of the run til it dies approach, chains and cassettes can often run until the chain is stretched as much as 2%, roughly 4 times longer than the .5% replacement point.

So the mechanic wasn't wrong, just voicing one of a bunch of equally valid opinions. Whatever approach you adopt, for what ever reason, the key is to reduce chain wear per mile as much as possible with good chain lubrication and care.

furballi

Another method to extend chain life is to avoid jack-rabbit start. Don't apply a large peak load on the crank arm. Anticipate shifting to a lower gear. Finally, stick to those gears that do not put excessive twist on the chain.

DaveSSS 

FBinNY made a good point about not tossing those .5% elongated chains. If you put a third or fourth new chain on a cassette and get new-chain skip, you will find that any of those previsouly used chain can be reinstalled and they will not skip. If nothing else, cassettes and chains like these would be great for a bad weather bike.

Another point I need to mention is that well maintained Campy chains will rarely elongate like the other brands. A Campy chain can be totally shot, but have only .2% elongation. After 6,000 miles, that's what one of my Campy chains had, but the rollers were shot and the side clearance was about .013 inch, or nearly twice the original amount. That chain was ready for the trash. That chain also wore my 19T cog enough to cause new-chain skip. That worn cog would not skip if I used a chain that had only 300 miles of break-in wear on a new cassette.

Booger1

Please explain how a used chain wears slower.The pin is smaller,the hole in the side plates is bigger and the inside of the roller is bigger.How does this give it more contact surface?

I don't get it.

DaveSSS 

Who said a used chain wears slower? All chains will measure a little short of the exact .500 inch pitch when brand new and quickly wear-in after only a few rides. Once good pin to bushing contact is established, the wear rate should not change much.

FBinNY 

Draw two circles of different diameters on a piece of paper, touching on one side. That's the fit of the pin and "bushing" on a chain. The point of contact is just that a point, or in the case of cylinders a line.

With wear, the pin doesn't shrink in radius, nor does the bushing enlarge in diameter, but they lap together to matched curvature in the contact zone broadening that theoretical line of contact area to a band.

Booger1

Is that the difference between a cheap chain and an expensive chain? They lap the pins and side plates?

So the hole in the side plate is not flat to begin with?

Somethings wearing,the chain IS longer.

BCRider

Well he didn't say that nothing wears. The elongation is due to wear. As FB described the pin and bushing situation you've got two cylinder forms in a contact line. The pressure and grit eventually wears the bushing into an egg shape with the pin sitting in the "point" of the egg. It's that elongation of the holes in the bushings and a corresponding reduction in the diameter of the pins that we call "stretch".

In a new chain where there's a thin line contact the initial wear will be quicker until the pin has lapped in a saddle that perfectly matches the diameter of the pin. So there will be some rapid initial stretch. Once lapped in there is more contact area and further wear will be slower.

I've always liked the idea of two or three chains and rotate them between cleanings. But I never got around to trying the idea. I just found it easier to rotate bikes instead...

FBinNY 

Maybe I wasn't clear. The wear rate doesn't continually improve, but the fastest wear is when they're new, so wherever you draw the line of half it's life, the first half will be shorter in miles than the second.

Like new cars who lose value fastest in the first year of use, new chains wear fastest at the beginning, and slower later.

DaveSSS 

[QUOTE=Booger1;11000721]Is that the difference between a cheap chain and an expensive chain? They lap the pins and side plates?

So the hole in the side plate is not flat to begin with?

[QUOTE]

All bike chains are made in the same general manner. The bushing for the pin is stamped into the inner side plates and the pins are made to a very precise diameter, perhaps by centerless grinding, before being cut to length on a screw machine. The stamped bushing is very short, and likely to be slightly conical. Cheap chains are most likely made of steel with a lower hardness and will wear faster. Campy chains are the only brand I've used that elongate at such a low rate. That makes the recommended method of using elongation to judge chain wear to be worthless. I go by roller wear with a Campy chain. When new, a Campy chain will have a space between the rollers of about .200 inch (Shimano and KMC will measure more like .210). When that spacing gets into the .235-.240 range, I consider the chain to be shot, even if the elongation is small. Rollers with that much wear will have many times the wear that the pins and bushing do. The OD may be .005 inch smaller and the ID .010 inch larger. In comparison, the pin and bushing may only have .001 inch of wear.

BCRider

Dave raises a good point that many of us forget about or just don't know about.

There are TWO points of wear in any roller chain. First is the pin to bushing that we've been obsessing about in this thread. Second is that there's a separate roller that fits on over the bushing. As he's very correctly noted this roller to bushing wear is also an issue and adds to the problem in an old chain but won't be seen in a purely link distance measurement test for wear.

MajorMantra

Thanks for all the views, very informative. A question for those who rotate a number (say 3) chains: at what mileage interval do you like to swap to the next one?

DaveSSS 

The most critical part of rotating chains is getting all of them into use without encountering new-chain skip. If you've changed chains at .5% in the past and the third chain never skipped, that would be a good interval. With a Campy chain that doesn't elongate very much, I change when the roller spacing increases to about .220 inch, which is also about the time that Campy says to toss a chain.

After all three chains have about the same mileage, I'd alternate a little more often. The idea is keep all of the chains similarly worn.

DMF 

That seems counter-intuitive.

Doesn't a worn cog effectively have wider-than-unworn spacing between the teeth? If so, the spacing would be wider than the link length (of a new chain), concentrating the load on a few teeth and links, and wearing both faster than if the load were evenly distributed. Likewise a new cog with worn chain.

A chain and cassette worn together would have the same spacing and distribute the load evenly.