Tuesday, July 10, 2018

Getting Shafted

Chains are such a drag. I mean, what else can you call a transmission system that's only 98% efficient? And yes, they're, simple, straightforward, "proven" technology, and relatively inexpensive - but they also get greasy. So, Yuck.

And that's why innovators keep trying to come up with new and better alternatives.

The latest is a shaft drive transmission from Ceramicspeed - the same people who have been pushing the envelope with pricey lubricants, bearings and derailleur pulleys to eliminate drivetrain friction - in a quest to get drivetrain efficiency up to 99%. Their new shaft system uses carbon fiber, tiny ceramic bearings, and raspy-looking cogs with rows of outward-pointing teeth.

Though the pictures show a multispeed (I count 13) cog set, it's apparently only a mock-up as they haven't actually figured out exactly how to get the system to shift across all those teeth.
Instead of bevel gears, as most shaft drive systems employ, the Ceramicspeed design uses a pinion with tiny rollers to engage the outward-pointing teeth on the cogs. It's supposed to virtually eliminate friction.
. . . And it apparently works - as long as you don't want to shift gears. . . yet.
According to the system's developers, their shaft drive system has "49% less friction than a stock Shimano Dura Ace drivetrain averaged across all gears." That sounds impressive. On the other hand, as already mentioned, the traditional chain-drive transmission is already very efficient. Depending on certain variables, including the specific gear combination being used, as well as lubrication, the efficiency can be anywhere from 93 to 98 percent. For example - the highest friction/lowest efficiency gear combination would be one where you have an extreme "cross-chain" effect, such as the small chainring/small cog combination. So when you're talking about an improvement like 49% less friction, you're really talking about a fraction of what is already a very small fraction. In other words, while the difference might be measurable with a power meter, I'd be skeptical about whether many riders could actually feel the difference. I mean, can you feel the difference in drivetrain friction between your small chainring/small cog as compared to a more sensible gear combo? I doubt it.

Some caveats. First - the system is still very early in its development and there are a lot of issues that remain to be solved. One major issue (mentioned already) is that so far it cannot shift gears. There are some ideas involving a small electric motor to move the roller-bearing pinion forward or backward along the face of those cogs - but how to get it to move across those teeth in a synchronized way at speed and under load is something of a puzzle. Imagine shifting gears in a manual transmission car without use of the clutch, and without (or with worn out) gear synchrosCaarrunch!

Second - there is no compatibility, and likely no way to make it compatible with any existing bikes or frames. That shouldn't be a surprise, since it's a totally new thing, but bikes would definitely need to be built specifically for it. And I cannot even imagine what it would do to wheel changes.

Another thing to consider is the complexity and cost. I simply cannot foresee any way that a system like this could even approach the cost of chains and "normal" cogs. That raspy-looking cog set bears an unlikely combination of looking simultaneously fragile and dangerous, in addition to probably being a real bear to make.

A couple things to remember - shaft drive is nothing new. Columbia Bicycles (Pope Bicycle Mfg.) had a shaft drive bike as early as the 1890s. I don't believe they were the only ones, either. This was before the roller chain we've known for about a century took its current form and established itself as the preferred transmission. In a manner of speaking, one could argue that the chain was the improvement over the shaft. The shaft was overly complex, less efficient, and too expensive to manufacture as compared to chains. But that hasn't stopped numerous companies and "innovators" from trying to re-invent them.

Turn of the century (the 20th century, that is) Columbia shaft drive bike. It would appear that the bike had rear suspension, too.
Modern shaft drive from Dynamic in the UK.
On a related note, there are a number of shaft drive motorcycles out there - mostly touring models. The added cost and complexity, as well as the extra weight, are much more easily absorbed in a motorcycle than on a bicycle. But except for companies like BMW (better known as a car company, by the way) which uses shaft drive on most (maybe all) of its models, the majority of motorcycles still run with chains. I would argue it's for good reason.

So, is there a shaft-drive bicycle in your future? Meh. Probably not.


  1. There are some nice pictures out there of Major Taylor racing a shaft drive bike.

    1. Shhh! Don't let Brooks hear that, you'll get him all kerfuffled!

  2. At this point, an "engineering marvel" but probably not a solution to the "problem" of chains. The shifting problem could be easily solved with an internally-geared hub, but there goes the efficiency advantage. One of the problems with chains is their performance in wetness and dirt. How would those little little rollers, which appear to have very small, exposed bearings, do with a bit of mud in them? I'm not optimistic.

  3. I agree with all your criticisms of this system. And what about weight? I'm thinking this would end up weighing more than a derailleur. But what irked me most when I first saw a post on this was the thing on the rear wheel being called a "cassette".

  4. There are shaft drive bicycles laying around everywhere today: Many municipal bike shares use the design. They are low maintenance, low mess, and importantly - difficult to vandalize. I rode one in Portland a few years ago. Wide range internal geared hub and a shaft drive. It allowed me to pedal the tank up hills without a problem. Designs don't have to be about efficiency.