But I'm talkin' about Shaft!

July 17, 2003 1:20 PM

While shaft-driven motorcycles have been around for nearly a hundred years (and in bicycles even longer), they've only recently been re-introduced to bicycles. Chainless.com is a company promoting a new shaft drive system for use in low-end bicycles (looks like mountain, road, and bmx are offered). Without chains, it's easier to repair, safer, and they claim it is more efficient, though gearing is limited to a few hubs that feature up to 7 internal gears. If it really is more efficient, how long until Lance Armstrong is sporting a shaft-driven Trek?
posted by mathowie (39 comments total)

They look way cool, and I'd love to try one out, but there's no way I could think of springing $400 for a bike I've never ridden. The net can be quite frustrating like that. The whole 'never get that nasty grease stain on your right pantleg again' thing is quite appealing.
posted by woil at 1:30 PM on July 17, 2003

Hmm. The concept itself is really cool, but their mountain bikes look like total shit. Would be awesome if someone else developed a usable shaft mtb...
posted by COBRA! at 1:36 PM on July 17, 2003

I forgot to mention I found out these existed from Kevin Kelly's review of his bike. Looking around, it seems that some Taiwanese company makes the shaft drive system and is including it on pretty low-end bikes. They don't look too sturdy to me and the parts look pretty shoddy.

I'm hoping a serious manufacturer like Shimano tries their hand at it and perfects it into something impressive.
posted by mathowie at 1:39 PM on July 17, 2003

Has anybody here ever ridden one of these things? I'd be interested to hear a first-hand opinion of their performance.
posted by mathis23 at 1:41 PM on July 17, 2003

I wonder if weight's an issue. Is the overall amount of metal in the shaft about the same as in a chain?

I suppose a titanium shaft would be pretty light and strong, but expensive.
posted by COBRA! at 1:42 PM on July 17, 2003

well, the say the current system weighs 4 lbs, which sounds hefty.
posted by mathowie at 1:44 PM on July 17, 2003

If motorcycles are a guide, it's likely to be the weight of the shaft (one that's strong enough for real work) that limits it's appeal. Shaft drive for motorcycles is still limited to heavier less agile machines - sports bikes still have chains even after 100 years (sadly).
posted by grahamwell at 1:52 PM on July 17, 2003

I would imagine as they develop they could get lighter, much like the mt. bikes and BMX themselves. I'm with mathis23 on this, anyone tried it? Why didn't this compete all along? I wonder if the limited gearing would have that much of a competitive edge for the chain drive back in the 3 speed beach comber days anyway?

As for "serious manufacturers" didn't folks say the same thing about Shimano 20-30 years ago? I totally buy what your saying, I just think its funny how one name bubbles to the top every once and a while.
posted by Pollomacho at 1:57 PM on July 17, 2003

shaft drives in motorcycles also reduce torque being produced at the rear wheel.
posted by jbelshaw at 1:58 PM on July 17, 2003

i'd be totally into getting my hands on just the drive to replace the gearing on an old trek hybrid that's completey wrecked. i'm not a avid-enough cycler to really care about the added weight that this might bring (recently i've been riding around a tank of a purple mountain bike that i bought for $25)

it looks like they manafacture it separately, but doesn't appear that any of the distributors are making it available.

i guess i'll try putting in some emails later on today.
posted by fishfucker at 2:08 PM on July 17, 2003

Whoa, I saw a very ancient shaft-driven bike in a transportation museum once. That's pretty cool that they're bringing them back...
posted by beth at 2:13 PM on July 17, 2003

Can I still run over these in my SUV?
posted by xmutex at 2:17 PM on July 17, 2003

the store down the street from me has had these for about 3 years (don't know the brand but they're always in the window). i've never really been tempted to try. i always think that if one's on the road and it breaks the chances the local fix it shop would be able to do anything are remote. i should pop in and lift one though and report back on the weight.
posted by dobbs at 2:20 PM on July 17, 2003

from zerocycles FAQ:

"Question: Can my existing bike be fitted with a shaft drive unit?
Answer: No. The shaft drive system needs a wider and shorter bottom bracket shell and special dropouts."

bah. it's just a novelty until i can make it work on *any* bike.

(either that or i get a hammer and go to work on my frame.)
posted by fishfucker at 2:29 PM on July 17, 2003

Anyone know how easy these thigs are to fix if something breaks? The FAQ had something about the low maintenance, but I ride my bike like I stole it... If it could be fixed trailside easily enough, it might be worth investigating (anything to bring an end to the heartbreak of chain suck!)

Oh yes, and fishfucker; Apparently this system requires special dropouts at the rear and a non-standard bottom bracket, so retrofitting is out, I guess.
posted by Jughead at 2:30 PM on July 17, 2003

sigh. my youth was spent riding pro bmx, so I'm all for this. i ruined many pants by getting grease stains on them or getting caught between the sprocket and the chain.

Now, here's a question.. would power transfer be as smooth as it is w/chains? i wonder if it would be awkward at all. i'm not a physicist, so please no flames.

and would this still have the traditional freewheel?

*looks at his scar from a skipped freewheel incident* I miss those days.
posted by shadow45 at 2:48 PM on July 17, 2003

This is a non-starter. It's been re-invented every few years as long as there's been a bicycle industry - I remember seeing a shaft drive design at a bicycle trade fair around 1991, and as mattowie points out, they've been around a lot longer than that. Given the history, if this were a better design, it would have been recognized and widely adopted by now. The big problem with bicycles, from a marketing perspective, is that they're a mature product - pretty much unimprovable, other than through the application of new materials.

Problem with shaft drives is severalfold:

Shafts are way heavier than chain drives.

They're much less efficient than chain drives. Their FAQ is simply wrong, unless they're using a very unorthodox definition of 'efficiency' that I can't imagine. In fact, the bicycle chain drive, when clean and well adjusted, is pretty much the most efficient power transfer mechanism known to mechanical science - we're talking in the region of 99.5%. The design hasn't changed in generations (other than better materials) because it's pretty much unimprovable.
The problem with shafts is that they suffer from gear lash (momentary loss of contact between surfaces when transmission force cycles - like it does when pedalling). The only effective way to overcome lash is to beef up the components to make them very stiff and give them a greater inertia... more weight.

They're much more difficult to maintain. Encasing them in a gear case is smart - protection - but that case then needs an oilbath: more weight and makes maintenance even more of a hassle. Cases also need excellent sealing from the environment to prevent contamination of the transmission. Good seals are hard to design and expensive (c.f. Phil Woods hubs for bicycles) - I'd be surprised if this version was well sealed, given the price.

By low maintenance, I suspect they mean "get a whole new unit when this one wears out".
posted by normy at 2:51 PM on July 17, 2003

Various points:
-Racing motorcycles still use a chain, due to the weight and efficiency of same.

-"Reducing torque" doesn't sound right, as torque is a production of the engine, not the drive line. Shaft drive bikes do have a problem with driveshaft torque raising and lowering the rear end, and I understand it's a bugger to engineer around it.

- I've seen this shaft drive bike ad around. I'm not sure how well it would work on racing bikes, as changing gear ratios and fixing impact damage would be a hassle.

- I'm also concerned w/ the number of moving parts in a shaft drive as compared to the number of moving parts in a chain drive.

- Still, I suspect for a neighborhood cruiser it'll work great, as it'll be a "set it and forget it" operation.
posted by Elvis at 2:55 PM on July 17, 2003

trharlan: what normy said... basically its an efficiency issue.
posted by jbelshaw at 2:56 PM on July 17, 2003

Every time shaft-drive has been tried on a bicycle, it's never gone anywhere. The problem? Torsion.

Chains don't twist, they simply pull -- the major strain is tension. Steel and other metals have amazing tensile strength.

Shafts, however, spin. Now, rather than putting the strain in tension, you're putting it in torsion -- your "chainwheel" turns the shaft, which then turns the "cogs" (scare quotes, because you wouldn't have conventional chainwheels and cogs.)

There aren't many materials that are strong in torsion, compared to tension, period. They'll be using a pretty large billet of Cro-Moly steel, most likely -- Ti and Al wouldn't work at all -- way to weak in torsion.

Worse, building a shaft drive for a fixed single speed would be one thing -- it's very hard to gear. The only easy way to do it is to use an internal hub gearing system. While there are a couple of nice things about internal hub gears, they are heavy. Add the weight of that hub to the total weight of the system (or deal without gears.) The largest range I know of on an internal hub is 12 speeds, but I think that was a one-off. 7 speed hubs are common -- and it appears they use the Shimano Nexus 7 speed hub, which is a pretty nice unit.

So, forget about this going onto racers -- they'll swap out butyl tubes for latex tubes to save 60 grams -- they aren't bolting an extra kilogram or so onto a bike. For commuting, though, it could be a win, since it appears that the system is sealed. Take this on a serious mountain trail, however, and it'll be a race to see wether the shaft drive or the internal hub blows out first.


As for "serious manufacturers" didn't folks say the same thing about Shimano 20-30 years ago?

True. 1986 was the year that Shimano took over from Sun Tour as the largest selling component manufacture in the US, by 1992, they had for all intents won everything. Why? Simple.

In 1982, Shimano introduced Posi-shift. Afraid that the notoriously conservative race set wouldn't trust index shifting, they shipped it on the low end components. Being low end, they weren't reliable. SunTour kept making killer friction shifts, and Posi-shift went away after 1984. The fact that Posi-shift had the index points in the derailuer, not the shifters, made adjusting them nightmarish, even worse on cheap components.

In 1985, Shimano introduces two new technologies -- Biopace chainwheels, and SIS index shifting. This time, however, both are launched not on the bottom end, but on the Dura-Ace group of components -- Shimano's very top line group. (Still is, for that matter.)

SunTour went and talked to every dealer they could, asking what they thought about this. The answer was universal. Index shifting failed before, it'll fail again, good luck getting racers to change -- but that Biopace ring looks like a winner, indeed.

In 1986, the song had changed. "If it doesn't click, it doesn't sell." Shimano outsold SunTour by a big chunk. Racers loved index shifting. It took SunTour two more years to come out with Accu-Shift, but they didn't sell it as a group, and it was intolerant of poor adjustment -- basically, it sucked. Thus ended SunTour, they kept shrinking, and were gone by 1992. Biopace went away too, but SIS rules the roost.

Theoretically, that could happen again -- but I doubt it. Shimano always spent far more on R&D than SunTour or Campo -- and they haven't stopped spending.
posted by eriko at 3:08 PM on July 17, 2003

14 speed internal geared up by a very reputable company in this area. Too rich for my blood though...
posted by daver at 3:16 PM on July 17, 2003

up = hub. oosp.
posted by daver at 3:17 PM on July 17, 2003

The bikes look kind of naked without the sprockets and chain dangling, but perhaps that is just because we have got so used to seeing the mass of external bits in a derailuer set-up. I like the look and the idea of not having pants eaten by the chain, but the inherent complication with the need for precision machined parts for the system to work effectively seems to be the drawback to me. The market here would be "the MTB that never strays from the street" buyer, I feel.

Under normal use, the gears will not fray for a long time.

What is "a long time"? Is that a new engineering term?
posted by dg at 4:07 PM on July 17, 2003

damn, eriko's droppin science!
posted by shadow45 at 4:13 PM on July 17, 2003

xmutex: That's the funniest thing I read all day.
posted by sharksandwich at 5:28 PM on July 17, 2003

Thanks for the excellent info, eriko.
posted by mathowie at 5:48 PM on July 17, 2003

I think for motorcycles the main benefit of shaft drive is ease of maintenance. A 50+HP bike will stretch that chain something wicked, and then you have to adjust it to take up the slack.
I've always fancied the idea of a shaft driven push-bike, but I don't suppose it's economically competitive. But, assuming it worked OK and wasn't too dear, I'd love one.
posted by Joeforking at 6:50 PM on July 17, 2003

i thought suntour lost it because their patent(?) on the parallelogram in the rear derailleur expired and they'd done no development for years. is that just an urban myth? (i must admit it doesn't make much sense, because campy produced rear derailleurs...)
posted by andrew cooke at 6:52 PM on July 17, 2003

I'd love one of these bikes.

I've had conventional chain-drive bikes and motorcycles, and now own a shaft-driven MC. It doesn't need readjustment, alignment, tensioning, constant oiling or cleaning. It won't fly off at 130 km/h and take my leg off, and it doesn't pick up dust and sand from my forays in the great outdoors.

It may be true that a well oiled, perfectly aligned and newly-maintained chain is more effective (but I can't see why), but honestly, how often is your chain in that condition? How often have your chain jumped off? How often have you gotten a twig in the derailleur and had to spend 15 minutes readjusting it?

/rant
posted by spazzm at 7:35 PM on July 17, 2003

The rise and fall of SunTour is described in detail by Frank Berto in an excellent article.
posted by normy at 7:52 PM on July 17, 2003

Nah, SunTour had lots of clever things -- first barcons out -- and a very trick leverage wheel that unloaded the cable tension when you shifted -- a bitch to dial in, but man, there was nothing smoother -- they were *the* shifter in the Tour de France for at least a decade. They figure out how to put 6 cogs on 120mm, and 7 on 126mm. Clever guys, and they were very consistent -- the bad parts were bad, but the good parts were always good.

But SunTour was basically a couple of very clever people in R&D, where Shimano was lots of people. When the clever people at SunTour made an error, it intended to be dramatic -- the Superbe Tech (IIRC) derailers jump to mind -- great shifting, but they broke easily. Ok on a midline part, a disaster on a component that's supposed to live on your top tier of components. Never mind the brake failure in 1991 that lead to the end of SunTour.

Shimano, by just throwing people at the problem, figured out the real problem with index shifting. It wasn't the derailers, and it wasn't the shifters -- it was the cable! Conventional Bowden cable would compress under load. For brakes, where you are transmitting force, no big deal, for friction levers, you'd just work the lever until it was on the gear. On index, it was a disaster -- between thermal expansion and force compression, you never knew the length difference between the housing and the cable, so, you never knew where the derailer would end up.

So, Shimano fixed that, with the "SIS cable" -- more commonly known as "compressionless housing." Instead of the spiral of Bowden cable, they used lengths of steel fibres, laid parallel to the main cable. This used steel in compression (yeah, I know, but that's what they call the cable) steel's good at that, and this let them control the difference in length between the cable and the housing -- which meant they could put the indents in the levers, and indexing worked.

Note Well: Never, Ever, Ever, EVER use this cable for brakes. It will fail, and it will fail exactly when you need the brakes.

My current ride is a 1988 Canondale Criterium 400, fitted with SunTour componets (alas, the lousy Accu-Shift system, which is anything but.) I love the De Compi licensed brakes, and the Sugino crank, and the Suntour Freewheel (6 cogs, 13-24 -- good match for St. Louis) but those shifters? Bah. They suck large rocks.

I really need to find the time/money to drop a decent Shimano shifter onto that bike.
posted by eriko at 10:22 PM on July 17, 2003

Hey, I hadn't seen that 14 speed hub. Sheldon Brown could drop that onto his crazy 63 speed bicycle Instead of the 3x7x3 gearing, he'd have 14x7x3 gearing -- 294 gears!

Sheesh. You'd grow old and die just figuring out the shift pattern.
posted by eriko at 10:35 PM on July 17, 2003 [1 favorite]

I like the chain. It is absurdly simple. Everything on my bike, including the Shimano shifterifficus, can be groked by a plus-side moron. It is all delightfully to the point, unhidden, and innocent. I like it that way. Leave us alone.

Interesting stuff, eriko.
posted by Opus Dark at 1:45 AM on July 18, 2003

*bows down before the cycle-history deity that is eriko*
posted by humuhumu at 2:19 AM on July 18, 2003

Compressionless cable housing was only one part of a reliable indexing system. In recognition that it's impossible to make a completely compressionless system (even if the housing is compressionless, the cable itself will still strain under load), Shimano developed the floating jockey wheel, which allowed the top derailleur pulley to adjust its lateral position in compensation for slight inaccuracies in the chainline.
posted by normy at 5:54 AM on July 18, 2003

Instead of one of these shaft-drive bikes, what you could do now is use a full chaincase. They don't work with deraillers—the bike would have to be singlespeed, fixed, or use an internal geared hub—but that's the case if you had the shaft drive anyway, right? The chaincase will keep the chain clean, and the simple drivetrain from it falling off, without actually replacing the chain and losing its advantages.

All the bikes in the Netherlands are like this, I hear.
posted by Utilitaritron at 6:57 AM on July 18, 2003

tharlan/elvis: when you put a step-up gear in your transmission, you divide the "downstream" torque by the the step-up ratio. On my road bike, in a 52x13 gear, I may be exerting easily 100 foot-pounds of torque at the pedals, but only about 25 pounds at the rear hub. If you check out the Rohloff website (home of the 14-speed hub), you'll see they require a certain minimum gearing to prevent too much torque from reaching the hub.

Shaft drive is much less efficient than chain drive for two reasons. 1. You lose power whenever you change the drive axis (and with shafties, you change it twice). 2. Chains have rollers on every link, which are an important reason for their low friction. To achieve this with bevel gears, each gear-tooth would need to have a tiny roller embedded in it, mating to the trough on the opposite gear. It's been done, but it's incredibly expensive. Plus, in order to tolerate the torque a cyclist can produce, the shaft needs to be quite sturdy, as others have pointed out.

Bikes are tough to engineer: a strong rider can generate 200 foot-pounds of torque, but less than one horsepower. So there's a premium on efficiency and lightness, but you need a lot of strength, and you need a lot of gears to keep the rider's cadence in the sweet spot. A car's engine may have a sweet spot from 2000 to 5000 rpm. A good cyclist's sweet spot may be from 75 to 150 rpm.

There have been shaftie commuter bikes in Japan all along. Not very popular, but they do keep your pant-leg clean. It is extremely unlikely that shaft drive will catch on with serious cyclists, just as two-wheel drive bikes have not (it's been done).
posted by adamrice at 10:00 AM on July 18, 2003

thanks, eriko, normy.
posted by andrew cooke at 5:09 PM on July 18, 2003

The now defunct BikeE used SRAM internal three speed hubs and a rear freewheel with 7-8 speeds. If one were to put three chainrings on the front you'd have a lot of gears to play with.

The Recumbent people are often playing around with Rohloff hubs. There's also "The Behemoth" a huge portable bicycle winnebago thing that had a bazillion gears:

The Behemoth.
posted by mecran01 at 10:25 PM on July 28, 2003