Cycling
This page is not so much here because I’m a hardcore cyclist—I’m very much not—but because I’ve been emailing friends about some of the issues discussed in these links, and needed somewhere to put them all. May as well make it public.I am very excited about a George Reynolds Wishbone that I picked up from a friend. Nothing else I have ridden comes close to the Wishbone for handling, balance, responsiveness, and sheer bent-grin. An incident early in our relationship earned her the name “Trebuchet”: the front tire blew out and flipped around backward, and I came to a rather abrupt stop. Upon consideration of the effects of a front tire blowout on an upright, I was delighted to have lost a little skin off my elbow and hip.
Here is some info that recumbent riders or general bicycle geeks may find useful:
General Info and Links
- Ken Kifer’s Bike Pages. Amazing! Read them!
- Harris Cyclery: articles by Sheldon Brown. Excellent technical info.
- Bent Rider Online: Recumbent discussions, reviews, events, etc…
Safety second!
- Disk Brakes
- Safety Warning! Disk brakes and quick releases - what you need to know. And an older but frightening article testing how much stopping power can disk brakes generate before failure? Few of the disk brakes in the tests could compete with rim brakes, although the failure mode of some disks is safer.
- Lights
- For a graduate student, commuting means riding late at night.
The technology is advancing fast, but almost all manufacturers still
use a ridiculous flashlight-grade round beam pattern, rather than
the vehicle beam pattern used on car headlights and a few fringe (in
North America, at least) bike light manufacturers (for example,
Busch & Müller).
- Sheldon Brown explains Why Reflectors Don’t Work. However, this is not quite right. The car that is about to hit you will not see you, but if any car sees you, you raise awareness that there are bikes on the road. Therefore, reflectors should have an effect similar to increasing the number of cyclists (see below).
- Want to understand batteries? William Burrow’s Bikecurrent FAQ. My reading is this: Bright lights are easy to find. The slightly tricky—and really crucial—part is getting a good charger!
- Lots of good info on Bicycle Lighting Systems.
- Myra VanInwegen’s Bike Articles has lots of good stuff. Regarding lights, see here.
- Another lighting links page, lots of info and links to manufacturers, from Alan Braggins.
- This thread on BikeForums, entitled “Total Geekiness”, is about building 20W (or more, waaaay more, up to about 1400 lumens) bike lights for $40. There’s also some discussion of taillight options.
- Do you prefer HID? Build your own Xenon Bicycle Light and get 3200 lumens!
- Front:
- NiteRider MiNewt X2 Dual mounted too low (recumbents are weird like that). It’s about 300 grams, 300 lumens, has a really aggressive strobe mode (at least 5 Hz), lists for $250, and can be found fairly deeply discounted. This is a great light aside from the idiotic (round) beam pattern, but I had to return one because the charger didn’t work properly and another because the circuitry didn’t work properly. Third one’s a charm—recommended if you buy from somewhere that stands behind its products. NiteRider’s QA sucks, but REI’s warranty is perfect.
- Rear:
- NiteRider UltraFazer 3.0: 110 grams with 2 AA batteries, not sure about lumens but it's brighter than most taillights, excellent strobe mode, lists for $30. I got a $2 amber marker light cover at an automotive store, so I have an amber blinky (much much better than red). I should just replace the LEDs, but these lights have a history of very poor quality control, so I might have to return it someday.
- Helmets
- Surprisingly, there seems to be little or no evidence that helmets are good for anything but some special cases of superficial injuries. [Thompson, Rivara, Thompson 1989] claim 85% reduction in injuries, but they have been thoroughly discredited. "The 85% figure is a litmus test: anybody quoting it is either uninformed or deliberately misleading you." (one of the sources linked below; you get to find it). Furthermore, they can increase the risk of serious injury (little really solid research; most of this is speculation trying to account for the data) through
- Making your head effectively bigger.
- Increasing torque about the neck.
- Making motorists systematically give you less room (this mechanism, at least, has been measured).
- Good old fashioned risk compensation.
Mandatory helmet laws have been conclusively shown to increase the cycling injury rate:
- Helmets do not reduce risk, and may increase it.
- Mandatory helmet laws reduce the number of cyclists on the road.
- As the number of cyclists increases, the accident rate plummets, since motorists are used to anticipating cyclists.
- CycleHelmets.org, in particular the lit review.
- Guy Chapman’s literature review: maxi-FAQ.
- The Vehicular Cyclist has a great FAQ.
- “Bicycling is nearly six times as safe as living!”
- Since this is such an important issue, here’s a direct link to Ken Kifer’s article Is Cycling Dangerous?
- On Encouraging Cycling
- Get more people out on bikes! “If cycling halves, the risk to cyclists doubles.” I’ve also heard that if cycling doubles the risk to cyclists goes down by a third… at least, all studies I’ve seen found a dramatic effect!
- Get SUVs and trucks off the road. They’re not just
intimidating, but extremely dangerous:
When the average large pickup truck collides with a second vehicle, people in the second vehicle die at a rate of 293 for every 100,000 crashes, according to federal crash statistics. By comparison, large sport utility vehicles kill people in the second vehicle at a rate of 205 per 100,000 crashes; minivans kill at a rate of 104 deaths; and large cars at a rate of 85 deaths.
—NYT, July 31, 2003…not to mention the fact that SUVs and trucks are less likely to avoid a collision in the first place, due to both poor handling and, um, psychological factors. - I would love to see statistics on how many cyclist-car accidents involve a serious (ie. commuting) cyclist in the car. I have a hunch that it’s low, but no numbers. Wouldn’t it be wonderful if getting a driver’s license required biking a few hundred on-road miles per year?
- Random Links
- Bicycle Universe
- Please take the Bicycling Life survey!
Performance
- Issue 40 of Huff has a fascinating article comparing the effects of steering, tire size, tire pressure, seat inclination, etc., on wind resistance. The article begins on P. 5.
- Legs Larry’s Unorthodox Bicycle Region is a wealth of information on high-performance recumbents, tire rolling resistance data, fairing information, etc…
- Analytic Cycling has piles of performance data.
- Rolling resistance of bicycle tires from the University of Helsinki.
- Schwalbe technical info
- Rainir Pivit’s article, Rolling resistance of bicycle tires, with links to a few other related subjects.
- For those of us who really want a wind tunnel but who can’t quite find the time to make one, Coastdown Test info might be a passable alternative.
- John Tetz explains how to build a foam fairing.
Are you a Gear Slut too?
- Recumbent shops: Is the Boulder cycling community really too conservative to support a recumbent shop? Luckily you needn’t go far afield. This list thanks to Jim Dettwiler.
- Trail Central: mostly mountain biking stuff, but based in Boulder, and has a survey of the local bike shops that will be useful as soon as some more people submit opinions :)
- Calhoun Cycle has some nifty recumbent-specific stuff.
- Bouré sells a line of
recumbent-specific clothes. Excellent customer service, too.
Highly recommended! Here’s what I’ve used so far:
- Pro recumbent shorts: Great cut, great fabric, well-built, no annoying chamois. The somewhat bulky seams took a few rides to get used to, but they’re very well positioned not to chafe. Bouré claims that in exchange for the slightly bulky seams they get more stretch and vastly more durability, and I believe it.
- Elite thermo tights: Very comfortable down to -5°C, and probably below that. And then you could layer. Slightly loose around my skinny ankles—room for big socks.
- Pro armwarmers: These are truly a wonder-garment. I take them everywhere. They are amazing on backcountry ski days, especially in combination with a long wool scarf enveloping my torso: now I can shed a significant layer in 15 seconds.
- Pro leggings: They keep slipping down my legs when I wear them walking, but my legs are skinnier than those of most cyclists. I can only wear them cycling, but for this, they are warm, compact, stretchy, and appear very durable.
- Wool skull cap: Much quieter than any synthetic I’ve tried. Doesn’t stink quickly, either. This is not in fact made by Bouré, but is a Woolistic Sputnik. A little bulkier than a Smartwool skullcap, but more windproof and cut to cover the ears.
- M5 has the coolest tandem ever: a carbon back-to-back called the Double D. I’ve heard, though, that M5 is a very difficult company to deal with if you live in North America, and there are no rumours of a bike from M5 weighing as little as advertised.
- Bicycle Coffee Systems is a resource page linking to purveyors of fine caffeine accessories, and, of course, much information on why coffee is good for you.
- Bike Snob NYC
The Sport Commuter
The bike I want is not made. I’m convinced that there is a market for it but I have little evidence beyond the fact that a few of my friends have expressed interest in something similar. If this sounds like the bike for you, please send me email, and pester all the bike manufacturers and dealers you know. I just want to prove that there’s a market niche.- Lowracer (or maybe Highracer) (importance: 0.9)
- Speed is not the only advantage of a well-designed recumbent,
but it is a compelling one. There are three main factors that make
a bike fast:
- The rider’s ability to produce power.
- Low aerodynamic drag.
- Low rolling resistance.
The first is not just due to training: it appears that maximal sustained power is achieved with a hip angle of about 125°; peak bursts prefer a hip angle of about 135°.
The second is accomplished mainly by reducing frontal area (unless you want to get wild and tackle a fairing). That means putting the body in the windshadow of the feet. That means that the bottom bracket will be at least on the order of 23cm higher than the ass, for my size 11 feet. Higher than that doesn’t really help, and besides, if you ride with a headlamp, all you will see is your knees.
The third is a function of using high-pressure tires, and when in doubt, larger diameter seems to be better. So unless you have a good reason to use a small wheel, use a big wheel. 559/406 is not bad (especially given that many recumbents carry 70% of the weight on the rear wheel), but 559/559 is better. 406/406 should only be used when absolutely necessary. What do highracers (559/559 or 622/622) have going for them?
- Lower rolling resistance.
- Better low-speed handling.
- More interchangeable parts. Do you carry a spare tube?
- More efficient chain routing.
- Less likely to be fallen upon by a roadie.
- Better visibility.
- Harder to avoid heel strike.
- Harder to put your feet down when stopped.
- To my eye, they just don’t look as cool.
- You have further to fall. Falling off Trebuchet is almost relaxing. This alone may be a good enough argument for a lowracer.
- Under-seat steering (importance: 0.8)
- This slightly increases weight (sometimes) and frontal area, but above-seat stearing (ASS) doesn’t play well with helmet lights and low seat angles. USS also does much to protect your delicate flesh in a fall. More importantly, I have never ridden a bike with ASS that felt right, but USS feels like you’re reclining on a magic carpet, completely relaxed and dialed in as you swoop along. Brilliant!
- Full suspension (importance: 1)
- Besides the obvious and important comfort considerations,
remember that recumbents go very fast, especially downhill.
Forget offroad—suspension is crucial in order to keep the
tires in contact with imperfect pavement. Front suspension is
crucial for maintaining control. Rear suspension is less important
for safety, but hitting a pothole at speed can actually bounce you
off a hardtail.
And of course the bike needs to be gentle on my laptop—or my beer. This demands rear suspension. I don’t demand much travel, but it needs to be adjustable for a wide range of loads.
- Front disk brake (importance: 1)
- An excellent argument can be made for 20″ front wheels: less heel interference (although Bacchetta seems to do a good job with bigger wheels). But on long downhills, brakes that dump energy into the small front rim can cause the tube to melt, whereas some disk brakes can deal with the heat. The rear brake, heating a 26″ rim, is better off (and the consequences of a rear tire blow-out, while still ugly, are less dire) so I’m less concerned about it.
- Chain tubes optional (importance: 0.6)
- I don’t actually have numbers on drag for these puppies (do you?), but obviously they cause some. My main complaint against them is simply that they’re noisy! So I will not buy a bike that uses tubes to guide the chain—that’s what idlers are for.
- Lightweight (importance: real-valued)
- Why are recumbents so heavy? How much should a
good one weigh? There are two considerations. First, recumbents
have more stuff attached to them. Second, the frame shape is just
less structurally efficient.
Recumbents have some heavier components. This is not terribly hard to analyse: here’s a very quick and dirty list of the reasons that a ’bent must unavoidably weigh more than a wedgie. I’ve explained above why I consider a disk brake to be necessary, but I’ve not considered the rear suspension yet.
I’ve overstated the case here slightly, as Aerospoke wheels are not the lightest available (which might bring the difference down to 300g). I haven’t considered the reduced weight of tubes and tires because tubes are almost irrelevant and tires are too variable for good numbers.Part Weight, g Basic wedgie saddle −200 Basic wedgie seatpost −350 Velokraft Carbon seat with pad +800 Roadie caliper brake −150 Disk brake +400 Roadie fork −700 20″ suspension fork +1400 Extra drive chain +450 Underseat steering +600 700c front and rear wheels, Aerospoke −2700 26/20″ wheels, Aerospoke +2210 Total 1760 The upshot is that the frame (less fork) will need about 2kg more stuff stuck to it than will a wedgie. 8kg total stuff should be quite reasonable, 8.5kg trivial, and throwing money at the problem should get you down under 7kg (see Light Bikes for ideas).
How about the frame? The big unknown is how much weight we must gain by going from the structurally efficient rigid diamond frame to a tricky rear-suspended linear frame. I am having a hard time finding data on mountain bike frames, and even that information won’t be vastly useful: on the one hand 5cm of travel should be adequate, but on the other hand the geometry is less convenient. But based on nothing but optimism, I’m hoping that 4kg (almost three times the weight of the lightest road frames) is possible.
In fact, Velokraft has a rear-suspension bike that weighs, apparently, 2.8kg with suspension and seat (and front fork?).
Bikes that come close
- HP Velotechnik Speed Machine USS
-
I borrowed one of these and put somewhat over 100 miles on it. My
review is here.
- Minimum seat angle: 25°
- Lighter for 2006, but still 3kg too heavy.
- Front shock is damped using the friction of the elastomer spring! For chrissake, guys, that’s Wal-Mart technology!
- Poor road feel.
- When I forwarded the Disk Brake safety article to them, described a possibly-related problem I saw on a friend’s bike, and asked for them to refute the argument, they offered nothing but “Trust us, it’s safe.” Faith-based engineering??? Morons.
- Optima Cheetah
-
- Showstopper: Optima does not sell a rack for this bike.
- No under-seat steering.
- Bottom bracket is too low.
- …but damn it’s gorgeous!
- Challenge Fujin
-
- No under-seat steering.
- About 4kg too heavy.
- Chain tubes.
- Rear rack? “Day rack in development”?
- Challenge Mistral SL
-
- Available with front suspension? I do not know.
- The seat is too high, or the bottom bracket is too low.
- The seat only goes down to 31°
- Chain tubes.
- Based on my test-ride, these bikes don’t handle very well.
- Challenge Seiran
-
- The seat is too high, or the bottom bracket is too low.
- Once decked out with all the optional necessities, it is quite expensive.
- Nazca Fuego
-
- No front suspension.
- Seat-BB height is almost as much as I’d like.
- No underseat steering.
- About 3kg too heavy.
- A real luggage rack (that accepts standard panniers) is in development but not here yet.
- They do not have a dealer in North America.
On Drag
Nothing here yet but a couple of pictures of me on Trebuchet. The main reason that recumbents are so much faster than diamondframes is frontal area. Given that frontal area can’t possibly get much lower than it is on a good lowracer, the next thing to work on is decreasing the coefficient of drag. This is best done with some kind of fairing. | Frontal area of the bike. Using a
primitive method, I’ve calculated the frontal area at roughly
0.3m2. Typical values for uprights seem to be a bit over
0.5m2.
I wonder how much faster I’d go with one of those silly teardrop faired helmets… |
 | I just removed this
nasty-looking fairing from Trebuchet, and over my rolldown test course
I’ve noticed no difference in speed.
My arms are in normal riding position. Underseat steering involves more frontal area than the most extreme Crouching Gopher handlebars, but I find it far more comfortable and, well, exhilarating. But the fact that the fairing didn’t provide any measurable advantage is not surprising given that the only place in which it is the right width is behind my neck. In fact, an Ortlieb Large Office Bag or two adds not a whit to my frontal area :) But I’ve moved over to a pair of Ortlieb Bike Packer Plus panniers. They’re easier to open and close, can carry about half again as much stuff (despite Ortlieb’s official information, have external pockets, and have compression straps for better management of small loads. However, with my current rack they do add slightly to frontal area: they hang down a little. |
My next taillight?
It is well known that, as a fraction of cycling accidents, cyclists are rarely struck from behind. But as a fraction of accidents that happen to good cyclists who are obeying traffic laws, the number is much higher. Yes, it’s still low, but since it’s something that a cyclist can only do so much to avoid, you need to trust the traffic behind you to some degree. As Neal Stephenson says, cyclists in Boston ride without lights because if a motorist can’t see you, he can’t aim for you. But in more enlightened climes, let’s assume that being seen by motorists is a Good Thing.The law in many countries and most states of the USA demands a constant (not blinking) red light visible for such-and-such a distance straight behind. But red lights aren’t the most visible. Amber is far brighter, and denotes caution to motorists, leading some to recommend them. Bonus: legally they do not replace the red taillight, but they are legal.
Having just one battery pack for all lights would be convenient, but for a self-contained taillight I would dearly love to see one with, say, 5 bright red LEDs always on, and an amber xenon strobe or really bright (at least a Luxeon 3W?) LED blinky. For bike paths and group rides, run with just the red LEDs. For roads, twilight, rain, fog, etc., turn on the strobe.
Things I’d look for:
- The strobe should have a very wide angle of visibility, like 180° horizontal and at least 60° vertically. Do not underestimate the value of lighting up the pavement behind you. It is much easier to gauge distance to a large area than to a point source!
- Battery life with a 3W Luxeon blinky and 4 AAs will be on the order of 12 hours (depending on flash rate). A xenon strobe could be similar, or less. So the batteries should be very easy to remove for recharging. Speaking of which, it would be ludicrous to design such a light for non-rechargeables (ie. not to run off 1.2V/cell).
- Some claim that the strobe will want to flash at well over 1Hz, perhaps closer to 2. The triple-flash pattern used on many of the pedestrian and streetsweeper lights in Boulder (something like 1Hz primary and 8Hz secondary frequencies) is very effective!
- A remote switch for the strobe would be a very nice option for those of us whose routes switch frequently between bike path and road. No-one likes to stop.