Chain maintenance is one of the most controversial aspects of bicycle mechanics. Chain durability is affected by riding style, gear choice, whether the bicycle is ridden in rain or snow, type of soil in the local terrain, type of lubricant, lubrication techniques, and the sizes and condition of the bicycle's sprockets. Because there are so many variables, it has not been possible to do controlled experiments under real-world conditions. As a result, everybody's advice about chain maintenance is based on anecdotal "evidence" and experience. Experts disagree on this subject, sometimes bitterly. This is sometimes considered a "religious" matter in the bicycle community, and much vituperative invective has been uttered in this regard between different schismatic cults.
This article is based on my personal and professional experience and my own theories. If you disagree with them, I won't call you a fool or a villain, you may be right. I hope you will extend me the same courtesy.
The past fifteen years have seen many revolutions in bicycle design. Some of them have been conspicuously visible, such as the development of mountain bikes, disc wheels, new handlebar designs. Some have been less visible, but just as important, including clipless pedals, cyclecomputers, indexed shifting.
There has also been an invisible and little noted revolution in the construction of bicycle chains. People don't pay much attention to bicycle chains. After all, they all look basically alike. Derailers are much more interesting to look at and talk about, but it is still the chain that has to do the shifting. An old-style bicycle chain has ten parts per link. The typical 57 link chain used on the average multispeed bike has 570 parts, more than the whole rest of the bicycle put together. There are 114 outer plates, 114 inner plates, 114 rollers, 114 rivets, and 114 bushings.
The major revolution in chain design has been the introduction of the bushingless chain. The first of this type was the Sedisport, and it has acquired such a good reputation that other manufacturers have copied the design.
Bushingless chains have only eight parts per link. You cannot tell by looking at a chain on a bike if it is of bushingless design, because the bushings are hidden by other parts of the chain. The bushings can only be seen if you disassemble the chain.
In a conventional chain, the two inner side plates are held together by tubular bushings, which are like hollow rivets. If you open a link of conventional chain, you can see the ends of the bushings flush with the outsides of the inner side plates. When the chain is assembled, the outer side plates hide the ends of the bushings. The middles of the bushings are not visible because they are surrounded by the rollers.
The rivets that hold the chain together run through the middle of the bushings, and the rollers roll around the outside of the bushings, so both the inside and the outside surfaces of the bushings are subject to wear.
The inner side plates of a bushingless chain are three dimensional. Instead of having a simple hole at each end with a bushing pressed through it, each inner side plate hole has a protruding shoulder that amounts to half of a bushing. Since the side plates have an inside and an outside determined by the existence of the shoulders, they can also have bevels on the inside edges without further complicating the manufacturing process. These bevels permit the chain to run more smoothly when it is not perfectly lined up with the sprocket than a conventional chain with flat inner plates. They probably also improve shifting performance.
Since the "bushing" of a bushingless chain is made up of two halves that don't connect directly with each other, this type of chain is more flexible sideways than a conventional chain. This is because the two halves of the "bushing" have a bit of "wiggle room" with respect to each other.
Cyclists often speak of chain "stretch", as if the side plates of an old chain were pulled out of shape by the repeated stresses of pedaling. This is not actually how chains elongate. The major cause of chain "stretch" is wearing away of the metal where the rivet rotates inside of the bushing (or the "bushing" part of the inside plate) as the chain links flex and straighten as the chain goes onto and off of the sprockets. If you take apart an old, worn out chain, you can easily see the little notches worn into the sides of the rivets by the inside edges of the bushings. With bushingless chains, the inside edge of the side plate hole that rubs against the rivet has a smooth radius instead of a sharp corner. This probably contributes to the greater durability of bushingless chains.
You can see how the rivet of this unusually badly worn chain has been worn away. |
Note also how the roller has flopped out of position.
|The reason the roller flops around is that the "bushing" part of the chain has been eroded away.
No doubt the inner surface of the roller has become enlarged as well.
In my estimation, a major reason for the greater durability of bushingless chains is the improved flow of lubricant to the vulnerable parts of the chain.
There are three points where a chain needs lubrication. First, and most importantly, the rivets need to be lubricated where they move inside the inner links as the chain bends and straightens. Second, the insides of the rollers need lubrication to let them revolve freely around the bushings as they engage and disengage the sprocket teeth. If the rollers don't roll, they slide along the sprocket teeth, causing accelerated sprocket wear. Third, the surface where the outer side plates overlap the inner side plates can benefit from lubrication as well, although this contact surface is much more lightly loaded than the first two.
When a conventional chain is oiled, before oil can reach inside of the bushings to lubricate the rivets, it has to pass between the inner side plates and the outer side plates. With usual oiling techniques, such as sprays, the oil tries to get into both ends of the bushing at once. Air bubbles can get trapped in the space between the rivets and the bushings, and with oil at both ends of the bushings there is no place for the air bubbles to escape. In addition, the cracks between the inner and outer side plates are highly exposed to road dirt, and are often quite grungy. Thus, even if you are able to get oil into the bushing, it is likely to be contaminated.
The air bubble problem may also exist with lubricant flow into the inside of the roller to let it turn freely around the bushing, but the shorter length and larger diameter of the roller, compared to the inside of the bushing, probably make this a non-issue. The contamination problem here is also probably less severe, because the sprockets tend to clean the rollers off automatically.
With bushingless chains, the lubricant flow is entirely different. If oil is applied to the rollers, it can easily flow into both sides of the rollers, because air (and oil) can flow through the gap between the "half bushings". If a bushingless chain is only oiled on the rollers, for instance by a narrow-spout oil can, the oil is able to flow into both sides of the rollers, through the gap and onto the middles of the rivets. The oil then flows out along the rivets to the side plate junctions. Since the side plates are oiled from the inside, there is a natural self flushing action that brings dirt and sand out of the chain instead of into it.
The rollers themselves are cleaned by contact with the sprockets.
One of the great controversies of chain maintenance is whether you should oil a chain or not. The downside of oiling a chain is that the oil may carry grit into the interior of the chain, and that this grit-mixed-with-oil will act as grinding compound, causing accelerated wear. Many experts whose judgement I highly respect hold this opinion, but I do not believe that this is always the case. I have no doubt that this "grinding-compound" effect can occur, but the severity of the risk depends upon the sort of dust/soil prevalent in a given area, and, particularly, the oiling technique used.
Many cyclists oil their chains by spraying, usually aiming the spray can at the rear of the derailer cage, because this reduces the risk of overspray getting on their rims or tires. In my opinion, spray lubricants should not be used on bicycle chains, because they get too much oil where you don't need it and/or too little where you do.
Most of the schmutz that contaminates a chain is stuff that is thrown up by the front tire. This lands on the outer circumference of the chain. For this reason, you should always oil a chain on the inner circumference, the side that faces the sprockets. This decreases the tendency of the oil to carry crud into the interior bearing surfaces of the chain. The best way to apply the oil is with a drip-type oil can, along the top of the lower run of the chain. This applies the oil to the cleanest part of the chain.
I just run the pedals backwards while applying a line of oil down the rollers. It takes 15 or 20 seconds to oil a chain this way. In the case of old-fashioned bushing-type chains, I usually run the oil along the left side only, in hope that the clean oil will be able to flush through the bushings from left to right. I have used many different oils, my favorite is Phil Wood Tenacious oil.
The problem with lubricating conventional chains is that thick lubricants can't penetrate into the inaccessible crannies where they are really needed, but thin lubricants don't last long enough. There is a family of popular chain lubricants that deal with this by mixing a thick oil-type lubricant with a volatile solvent. The resulting mix is thin enough to get some penetration by capillary action, then the solvent evaporates and leaves the thick oil behind. This type of lube is easy to apply, and is very popular for that reason.
An alternative approach to chain lubrication is to immerse the chain in hot wax. This is a variation on the oil/solvent approach. The hot wax is of a thin enough consistency that it can theoretically penetrate into the private parts of the chain, then when it cools off, you have a nice thick lubricant in place where it can do the most good. The major advantage to this approach is that, once cooled off, the wax is not sticky, and doesn't attract dirt to the outside of the chain as readily. Downsides of the wax approach include the fact that it is a great deal of trouble, and that wax is probably not as good a lubricant as oil or grease.
New chains come pre-lubricated with a grease-type lubricant which has been installed at the factory. This is an excellent lubricant, and has been made to permeate all of the internal interstices in the chain.
This factory lube is superior to any lube that you can apply after the fact.
Some people make the bad mistake of deliberately removing this superior lubricant. Don't do this!
The factory lubricant all by itself is usually good for several hundred miles of service if the bike is not ridden in wet or dusty conditions. It is best not to apply any sort of lube to a new chain until it is clearly needed, because any wet lube you can apply will dilute the factory lube.
There are several ways that people try to clean their chains, none of them very satisfactory. Two of these ways may actually work.
The traditional way to clean a bicycle chain is to remove it from the bike, then soak and scrub it in solvent. This is a problem with newer chains, however. Improved sprocket design, such as Shimano's "Hyperglide" system have made it possible to shift under full power, which is very stressful to chains. (Older derailer systems with plain sprockets required the rider to ease up on the pedals while shifting.)
To withstand these high stresses, modern chains have rivets that are tighter fitting into the chain plates. The new rivets are difficult to remove and reinstall without damaging either the rivet or the side plate.If you wish to make a habit of cleaning your chain off-the-bike, the best approach is to buy an aftermarket master link, such as the Craig Super Link or SRAM PowerLink . These permit removal and re-installation of the chain without tools. The PowerLink is standard equipment supplied with SRAM chains, but it also works on other chains of the same width.
I used to use a parts cleaning tank and a toothbrush to clean chains, but Zaven Ghazarian, an excellent mechanic I used to work with came up with a better system: drop the chain into a plastic Coke bottle with a couple of ounces of un-diluted citrus degreaser, cap it, and shake thoroughly. Fish the chain out with a spoke, rinse in water, and you are all set! (I am told that Pepsi bottles also work, and are easier to remove the chain from, because they have a wider mouth...but I'm a Coke guy, not a Pepsi guy.)
The other major way to clean chains is with an on-the-bike cleaning machine. These are boxes which clip over the lower run of chain. They contain brushes and rollers that flex the chain and run it through a bath of solvent.
The off-the-bike approach has the advantage that it usually uses more solvent than will fit into an on-the-bike cleaning machine. Thus, it can dilute away more of the scuzz from the chain.
The on-the-bike system has the advantage that the cleaning machine flexes the links and spins the rollers. This scrubbing action may do a better job of cleaning the innards.
When a new chain meshes with a new sprocket, every roller that is in contact with the sprocket is pressing more-or-less equally against the corresponding tooth of the sprocket, so the load and stress are shared out equally, among 10-11 rollers/teeth, in this case. From the center of each roller to the center of the next is exactly 1/2" (12.7 mm). This dimension is known as the "pitch" of the chain. The sprocket teeth are made so that the center the curve that makes up each "valley" is 1/2" from the next. The diameter of the sprocket is determined by the pitch and the number of teeth.
This chain and sprocket have worn together. You can see daylight under the chain in some places. The worn chain has "stretched" so that it no longer matches the original pitch of the sprocket. The sprocket has worn so as to effectively increase its pitch to match the worn chain.
The illustration above shows two formerly identical sprockets, viewed from the right side. The one closest to us is badly worn. On a new sprocket tooth, the surface that the roller presses against is perpendicular to the pull of the chain. The worn teeth have become ramps, causing the chain to ride up under load.
The rollers ride up on the sloped teeth until they reach a radius that corresponds to the longer pitch of the worn chain. The effective diameter (and thus, the effective pitch) of the sprocket has become larger, since the chain is no longer riding in the bottoms of the valleys.
A new chain on a worn sprocket. Most of the driving is happening on the left side, where the chain first engages the sprocket. Due to the mismatch in pitch, the rollers in the part of the chain that is about to leave the sprocket is doing very little to push their teeth forward, instead, they are being forced upward by the slope of the teeth.
Since the teeth/rollers on our left are doing most of the work the load is concentrated on them.
In addition, as the roller follows its tooth around the sprocket it rolls up the "ramp," while under load. This promotes wear to the insides of the rollers and to the "bushings" they roll on. With a properly meshed chain, the roller only turns a tiny bit as it rolls onto and off of the chain.
Due to the pitch mismatch, the chain will not reliably mesh with the sprocket under load, and will tend to jump forward, skipping over the teeth.
A worn chain on a new sprocket. Due to the pitch mismatch, the load is only carried by the teeth/rollers on our right, the chain hangs slack on our left. The new sprocket will wear rapidly to match the pitch of the worn-out chain.
The standard way to measure chain wear is with a ruler or steel tape measure. This can be done without removing the chain from the bicycle. The normal technique is to measure a one-foot length, placing an inch mark of the ruler exactly in the middle of one rivet, then looking at the corresponding rivet 12 complete links away. On a new, unworn chain, this rivet will also line up exactly with an inch mark. With a worn chain, the rivet will be past the inch mark.
This gives a direct measurement of the wear to the chain, and an indirect measurement of the wear to the sprockets:
- If the rivet is less than 1/16" past the mark, all is well.
- If the rivet is 1/16" past the mark, you should replace the chain, but the sprockets are probably undamaged.
- If the rivet is 1/8" past the mark, you have left it too long, and the sprockets (at least the favorite ones) will be too badly worn. If you replace a chain at the 1/8" point, without replacing the sprockets, it may run OK and not skip, but the worn sprockets will cause the new chain to wear much faster than it should, until it catches up with the wear state of the sprockets.
- If the rivet is past the 1/8" mark, a new chain will almost certainly skip on the worn sprockets, especially the smaller ones.
This issue is discussed in my Article on Derailer Adjustment
See also Jobst Brandt on Chain Maintenance
|Articles by Sheldon Brown and others|
Copyright © 1996, 2008 Sheldon Brown
If you would like to make a link or bookmark to this page, the URL is: