Technical FAQ

Technical FAQ: Cross-chaining and friction

In this week's column, Lennard fields questions about friction caused by cross chaining in modern drivetrains.

Have a question for Lennard? Please email us veloqna@comcast.net to be included in Technical FAQ.

Dear Lennard,
While doing an entire ride in the small (34 tooth) ring on my Campagnolo Centaur 10 speed compact chainring set, in order to find out what riding a 1X system might feel like, I was surprised to note that the chain is actually quieter and produces less vibration, even when the chain line is not that straight (e.g., riding in 34-14, 34-13). Against the received view that the most efficient, and least wearing chain line is a straight one, I discovered that when on the large chainring (and upper middle section of the rear cassette), there is greater noise and vibration because the chain is under greater tension as a result of the strong spring of the lower (jockey wheel) cage increasing the force at which the chain hits the lower jockey wheel.

Could it be that having a straight chain line is no longer that important, especially in light of modern chains and cassettes that are designed to be ridden while being twisted and turned? More specifically, can the efficiency gain of having less tension on the lower jockey wheel outweigh the efficiency losses of not only cross chaining, but also having the chain run through the smaller cogs? If so, does this mean that a modern 1X drive train with 11 or 12 speeds as being just as efficient as a 2X system?
— Marc

Dear Marc,
Those are interesting observations about less chain noise and vibration in the smaller chainring. While you have a point about the chain tension being higher on the lower section of the chain span on the bigger chainring due to higher rear-derailleur-cage spring tension, the tension on the top span of the chain is higher when on the inner chainring, and, being so much higher than the tension on the lower span of chain, plays a bigger role in chain friction. The higher the chain tension, the greater the pressure between roller and tooth and between roller and inner-link collar bushing. In short, the answer to your second question is no.

This article clarifies this point as well as answers your other two questions. It shows that, in answer to your third question, a 2X system is more efficient than a 1X system. And it also reveals that the answer to your first question is yes, namely that cross chaining is less important (efficiency-wise) with modern drivetrains; however, this is due to the larger rear cog sizes of modern cassettes, rather than to the design of modern teeth and chains, since it is not the case with smaller rear cassettes.

Regarding your cross-chaining question, please consult the graph in the article. On the 2X system, it shows the drivetrain friction continuing to drop with cross-chaining as the cog size increases beyond when the chain is perfectly aligned between the chainring and rear cog on the 39 X 25-tooth gear. With the 1X system, it also continues to drop with cross-chaining as the cog size increases beyond the aligned ratio of 48 X 18-tooth, albeit only slightly. Conversely, when using the big chainring on the 2X system, cross-chaining friction dominates after the aligned 53 X 19-tooth ratio; friction increases with either increasing or decreasing cog size from there.

― Lennard

Dear Lennard,
On October 8, Louis commented that the angular change in each link was not to be considered by itself because a larger ring had more bent links; the premise is that there is a total 180-degree angular change that is invariant with respect to ring size. Thus a 52-tooth ring would have the same 180-degree “total articulation” as a 46-tooth ring. But I don’t think that’s the whole story.

The way I recall it, drivetrain gurus have pointed out that during each pass over the chainring, every link in the chain bends precisely once per pedal revolution. RPM depends on gear teeth ratios rather than gear teeth and I will assume we’re keeping gear ratio and rpm constant. Thus the “total articulation” for just the chain ring would be equal to the number of chain links X 180 degrees / half the chainring teeth. Then if one is on a bike with a given chain length and one selects a given gear ratio using larger rather than smaller cogs using the existing chain, larger gears have less total articulation because the articulation angle per link is reduced, but the number of links stayed the same.

If one were to redesign a bike from the ground up with all gears 13 percent larger and then made the chain longer to accommodate it, the net increase in chain links would be less than 13 percent, because roughly half of the chain length is fixed according to the there-and-back traverse across the length of the chain stays. For example, assuming that the derailleur pulley tooth count were to remain constant, a 52 X 26 gear would engage 39 teeth + pulleys, while a 46 X 23 gear would engage 34.5 links + pulleys; if you started with a 110 link chain for the 46 X 23 drivetrain, adding an extra five links to accommodate the 52 x 26 drivetrain to arrive at 115 links would only be a five percent increase in links; yet the reduction in articulation angle per link would still be 52/46 – 1 =13 percent. In that case the reduction in articulation angle per link would diminish faster than the increased chain length and one would still have less “total articulation”.

At least that’s what I imagine is true — until someone tells me the RIGHT answer.
— Brad

Dear Brad,
Wow, I love that calculation!
― Lennard


Lennard Zinn, our longtime technical writer, joined VeloNews in 1987. He is also a custom frame builder (www.zinncycles.com) and purveyor of non-custom huge bikes (bikeclydesdale.com), a former U.S. national team rider, co-author of “The Haywire Heart,” and author of many bicycle books including “Zinn and the Art of Road Bike Maintenance,” “Zinn and the Art of Mountain Bike Maintenance” and “Zinn’s Cycling Primer: Maintenance Skill Building for Cyclists.”

Zinn holds a bachelor’s in physics from Colorado College.

Follow @lennardzinn