Reports of the demise of this Web site are greatly exaggerated! We at thank Harris Cyclery for its support over the years. Harris Cyclery has closed, but we keep going. Keep visiting the site for new and updated articles, and news about possible new affilations.

Search and

find us on FB

More Articles by Jobst Brandt
Next: Electronic Shifting
Previous: The Continuously Variable Transmission


Subject: Chain Care, Wear and Skipping
From: Jobst Brandt
Date: January 10, 2002, revised November 23, 2004

Chain wear and care evoke never-ending discussions, especially for new bicyclists who are not happy with this dirtiest of bicycle parts. This leads to the first problem, of whether there is a best (and cleanest) way to care for a chain. Among several ways to take care of a chain, some traditional methods are the most damaging while others work to prolong useful life.

A myth that is difficult to dispel is the story that grease on a new chain, fresh out of the package, is not a lubricant but rather a preservative that must be removed. This piece of bicycling myth and lore thrives despite its illogic.

Riders often speak of "chain stretch," a technically misleading and incorrect term. Chains do not stretch, in the dictionary sense, by elongating the metal by tension. Chains lengthen because their hinge pins and sleeves wear. Chain wear is caused almost exclusively by road grit that enters the chain when it is oiled. Grit adheres to the outside of chains in the ugly black stuff that can get on one's leg, but external grime has little functional effect, being on the outside where it does the chain no harm.

The black stuff is oil colored by steel wear particles, nearly all of which come from pin and sleeve wear, the wear that causes pitch elongation. The rate of wear is dependent primarily on how clean the chain is internally rather than visible external cleanliness that gets the most attention.

Only when a dirty chain is oiled, or has excessive oil on it, can this grit move inside to cause damage. Commercial abrasive grinding paste is made of oil and silicon dioxide (sand) and silicon carbide (sand). You couldn't do it better if you tried to destroy a chain, than to oil it when dirty.

Primitive rule #1: Never oil a chain on the bike.

This means the chain should be cleaned of grit before oiling, and because this is practically impossible without submerging the chain in solvent (kerosene, commercial solvent, or paint thinner), it must be taken off the bicycle. Devices with rotating brushes that can be clamped on the chain while on the bicycle, do a fair job but are messy and do not prevent fine grit from becoming suspended in the solvent. External brushing or wiping moves grit out of sight, but mainly into the openings in the chain where subsequent oiling will carry it inside.

Never use gasoline because it is explosive and contains toxic light petroleum fractions that penetrate skin. Removing the chain from the bicycle isn't always possible. There are times (after riding in the rain) when a chain screams for oil and good cleaning is impractical. Fortunately, after riding in heavy rain, the chain is fairly clean and in that case rule #1 may be violated for humanitarian reasons. However, only an internally clean chain squeaks, so it isn't as bad as it sounds. Also, water is a moderately good lubricant, but it evaporates as soon as the road dries.

Removing solvent from the chain after rinsing is important. Compressed air is not readily available in the household, nor is a centrifuge. Manually slinging the chain around outdoors works best if the chain is a closed loop but without pressing the pin completely in. The other way is to evaporate it. Accelerated drying methods by heating should be avoided because they can be explosive.

Lubricating the chain with hot 90W gear lube works but it is also efficient fly paper, collecting plenty of hardpack between sprockets and on the outside of the chain. Motor oil is far better, but motorcycle chain and chainsaw lubricants are better yet, because they have volatile solvents that allow good penetration for their relatively viscous lubricant. Paraffin (canning wax), although clean, works poorly because it is not mobile and cannot replenish the bearing surfaces once it has been displaced. This becomes apparent with any water that gets on the chain. It immediately squeaks.

[I have found that motor oil works poorly: it washes out of the chain due to its detergent properties -- John Allen.]

Swaged bushing chains

Sedis was the first with its Sedisport (five element) chain to introduce swaged bushings, formed into its side plates, to replace (six element) chains with full width steel bushings on which the rollers and pins bear. Although stronger and lighter than prior chains, the five element chain achieves its light weight at the expense of durability. These chains, now the only derailleur chains available, have only vestigial sleeves in the form of short collars on the side plates to support the roller on the outside and the link pin on the inside. This design is both lighter and stronger because the side plates need not have the large hole for insertion of sleeves.

Pins inside full bushings of (six element) chains are well protected against lubricant depletion because both ends were covered by closely fitting side plates. Some motorcycle chains have O-ring seals at each end. In the swaged bushing design there is no continuous tube because the side plates are formed to support the roller and pin on a collar with a substantial central gap. In the wet, lubricant is quickly washed out of pin and roller and the smaller bearing area of the swaged bushing for the pin and roller easily gall and bind when lubrication fails. Although this is not a problem for this type of chain when dry it has feet of clay in the wet.

[I don't know whether Brandt's complaint about lubricant's washing out is conjecture, or borne out by experiment. Sheldon points out that swaged-bushing chains are much easier to lubricate -- John Allen.]

Spoke Divider

Spoke Divider

Chain Life

Chain wear depends almost entirely on cleanliness and lubrication rather than being a load problem. For bicycles the effect of load variations is insignificant compared to the lubricant and grit effects. For example, motorcycle primary chains, operated under oil in clean conditions, last as much as 100,000 miles while exposed rear chains must be replaced often.

The best way to determine whether a chain is worn is by measuring its length. A new half inch pitch chain will have a pin at exactly every half inch. As the pins and sleeves wear, this spacing increases, concentrating more load on the last tooth of engagement as the chain rolls off the sprocket, thus changing the tooth profile. When chain pitch grows over one half percent, it is time for a new chain. At one percent, sprocket wear progresses rapidly because this length change occurs only between pin and sleeve so that it is concentrated on every second pitch; the pitch of the inner link containing the rollers remaining constant. By holding a ruler along the chain on the bicycle, align an inch mark with a pin and see how far off the mark the pin is at twelve inches. An eighth of an inch (0.125) is one percent, twice the sixteenth limit that is a prudent time for a new chain.

Skipping Chain

Sprockets do not change pitch when they wear, only their tooth form changes. The number of teeth and base circle remain unchanged by normal sprocket wear. What changes is the diameter at which the worn and lengthened chain bears on the sprockets, making wear indentations at a larger diameter than a new chain requires. In practice, this amounts to a change of pitch, since the chain will no longer ride in the original valleys between the sprocket teeth.

A new chain often will not freely engage a worn rear sprocket under load even though its root diameter has the same pitch as the chain. This occurs because the previous (worn and elongated) chain formed pockets higher on each tooth (a larger pitch diameter) than an in pitch chain describes. This wear occurs because a worn chain rides high on the teeth. A chain with correct pitch cannot enter the pockets when its previous roller bears the previous tooth, because the pocket has an overhang that prevents entry.

Without a strong chain tensioner or a non derailleur bicycle, the chain has insufficient force on its slack run to engage a driven sprocket. In contrast, engagement of a driving sprocket, the crank sprocket, generally succeeds even with substantial tooth wear, because the drive tension forces engagement.

However, worn teeth on a driving sprocket cause "chain suck", the failure of the chain to disengage. This occurs more easily with a long-arm derailleur, common to most MTB's: that is one reason this occurs less with road racing bicycles, which experience a noisy disengagement instead.

In contrast, a worn chain will not run on a new driving sprocket. This is less apparent because new chainwheels are not often used with an old chain. In contrast to a driven (rear) sprocket, the chain enters the driving (front) sprocket under tension, where the previous chain links pull it into engagement. However, because a used chain has a longer pitch than the sprocket, previous rollers bear almost no load and allow the incoming chain link to climb the ramp of the tooth, each successive link riding higher than the previous until the chain jumps. The pockets in a used sprocket are small, but they change the pressure angle of the teeth enough to cause skipping.

Jobst Brandt

Spoke Divider

See also Sheldon Brown's article on chains

Spoke Divider

More Articles by Jobst Brandt
Next: Electronic Shifting
Previous: The Continuously Variable Transmission

Spoke Divider

Spoke Divider

Articles by Sheldon Brown and others
Beginners Brakes Commuting
Fixed Gear
Frames Gears &
Tandems Touring What's
Wheels Sheldon

Reports of the demise of this Web site are greatly exaggerated! We at thank Harris Cyclery for its support over the years. Harris Cyclery has closed, but we keep going. Keep visiting the site for new and updated articles, and news about possible new affilations.

Harris Cyclery Home Page

If you would like to make a link or bookmark to this page, the URL is:

Last Updated: by John Allen