Given the same seat angle, hiracers are just as aerodynamic as lowracers

So, the title sorta says it all: this thesis (that if you lower the seat angle on a hiracer, it'll go as fast as a lowracer at the same agnel) has been articulated by various folks at various times over the years. Usually the person positing this goes on to say how hiracers are 'faster over real-world roads' because they have a larger front wheel (or in some cases, both wheels are larger; dual 20 inch wheel lowracers exist, although I've never ridden one). Usually some claims about 'being able to see your front wheel better' or 'seeing road obstacles (e.g. potholes) better' get thrown in there, and finally, usually the poster points to some impressive race results by either a person or a team that rides hiracers. RAAM results are often used to support this contention; i.e., some person or persons on a hiracer turned in a better race than some person or persons on a lowracer.

Now.....I would think that most people can see that using race results as an indicator of the superiority of one bike platform over another is obviously inherently utterly devoid of any logic. In the upright world, they do this sort of thing all the time; I remember Colnago (whose bikes I adore, by the way) claiming various 'wins' for the C40. But really, folks....this is completely illogical. Riders ride bikes, not the other way around. If you put me on the fastest bike in the world next to the slowest rider in the pro peloton on the slowest bike you can imagine, I'm going to lose, most likely (well, maybe if you put him on a BikeE and me on my TiCa, I might have a shot, maybe). The variance of riders, even elite ones, in any given race, is just too great. In short, using anybody's race result as evidence of a bike's superiority vs. another one is just not very useful; it's a well intended fallacy at best, and sycophantic hero worship at worst. The truth is, it's just not interesting to know how well (name your favorite racer) did on a given bike, for the purposes of determining how well that bike will do for you or me.

Then there are wind tunnel tests. You ought to be able to measure the varying drag signatures of respective bikes this way, right? Except....no recumbent manufacturer can afford this kind of thing, with rare exceptions (I seem to recall John Schlitter getting some time one year on a Ti Aero in a wind tunnel, but still....). And there is some controversy as to how valid this is: David (don't know the last name), the owner of Nimble wheels (www.nimble.net), a maker of high quality boutique aerodynamic aftermarket wheels, maintains that wind tunnel testing is completely invalid for bikes and bike gear, because they (wind tunnels) were designed for airplanes.

Well, how about rolldowns? Except....what are we measuring? How much does the respective rolling resistance of bikes with varying wheel size play a part? What about the weight of the bike/rider package?

It just seems hard to get a grip on this stuff, and really give a fair test. I know which bikes I think are the fastest that I've ridden, but....often my perceptions are at variance with others', and heck, I'm at least as subject to bias and prejudice about these matters as anyone.

So, I set out to try to generate some data. It seemed to me (and still seems to me) that the nearest thing to getting any sort of 'truth' about how well a given bike platform will serve me is to test both bikes with power measurement hardware, over the same course, at a constant wattage average. Ideally, we'd do lots and lots of runs, over different days, with different wind conditions, trying to equalize the bike setups and conditions as much as possible. It'd be great to do this on lots of different courses, with varying degrees of road roughness and climbing, etc. Ideally, I'd be able to put enough wattage into these bikes so that we really get into the area where aerodynamic drag is the big limiter. M5 likes to use the 250W value, as some sort of constant for comparing different platforms (I'd like to know where they come up with those values for different platforms e.g. http://www.m5-ligfietsen.nl/uploads/Itemizer/1591.1.gif

However, life is short, and so is my patience for this kind of thing. Also, I had to borrow the highracer, from a very kind and patient acquaintance (DuncanJames on Bentrideronline), a Bacchetta aluminum Corsa. Once I had gotten the bike from the good Duncan, I had to figure out how to equalize (as much as possible) things between the Corsa and one of my M5 TiCa's (the lowracer in question) so as to conduct a fair test. With that in mind, I did the following:

• Both bikes have a Zipp 404 clincher rim equipped with a PowerTap wireless hub. The Corsa wheel is 650c, the TiCa 700c.
• Both bikes needed to have their seats set to the same angle. This proved problematic. The TiCa has a basically unadjustable seat angle (you can change it, but only by swapping in a different set of Delrin seat risers; since I only have the one set, it's effectively unadjustable). The Corsa seat angle is adjustable, but.....I couldn't get Duncan's bike anywhere near as low as my TiCa. Finally, I called Dale Clark at AngleLake Cyclery, asking him if he shorter seat struts were available; he confirmed they were, and was good enough to give me a pair; he dropped them off at my front door, gratis (wow!). But then I needed to get a more rigorous measurement than just eyeballing. How does one do this? It turned out the LBS around the corner (Montlake Cycles) had an 'angle-finder' device. I brought the TiCa over and they measured it at 23 degrees (this bike is set up at the highest angle). But they also told me about an iPhone app that does the same thing. I downloaded it and re-measured the TiCa and got the same result (23 degrees). So I set up the EuroMesh seat on the Corsa to read the same, fiddling with the interactions between seatstays and the seat slider on the bottom until I got the same 23 degree number but still had proper extension for my legs. I jumped on the bike, and whoa, highracers are not all that easy to ride at 23 degrees with a 30 inch inseam. Still...just for one test....and after 5-10 miles, I started getting used to is. I have a CarBent coming from BentUp Cycles that's going to be set to 23 degrees; this experience removed some uncertainty about my ability to ride the bike at that angle; I know the CarBent is a little lower to the ground than the Corsa, so it should be good.
  
• Both bikes used a Schwalbe Ultremo on the rear (the only thing that was available from my LBS in both wheel sizes).
• The Corsa used a 'conventional' paired-spoke wheel on the front with a Vittoria Corsa Evo tire; the TiCa has the proprietary M5 front wheel; it's 18 spokes, and not particularly aerodynamic, and has a Schwalbe Stelvio 451 on it.
• I used my Garmin Edge 705 computer to read the PowerTap readings for both bikes. It uses the 'ANT+' protocol to talk to the hubs. For each bike, I completely re-detected the rear wheel, and calibrated the hub to the computer.
• I had a conversation with customer support at Saris (PowerTap) this morning, who confirmed that I had done everything reasonably possible to equalize the values between the wheels.
  

So, with all that done (what a hassle!), I was ready to test. I rode the Corsa out to a spot on Lake Washington Blvd that I had chosen; it's a run from the intersection of Dearborne St. and Lake Wa Blvd . to Seward Park and back. I chose it for the following reasons:

• While is has motor traffic, it's a favorite spot for cyclists; the motorists are used to cyclists time-trialing and doing intervals and threshold work there.
• It has a mix of smooth roads and washboard ones.
• It's mostly flat, but has a little climbing (maybe 150 ft. gain over 7.8 miles).
• It goes mostly north and south, but wends east and west for shorter periods. The wind usually blows either north or south there, and the fact that I was doing an out and back tends to equalize varying wind conditions, although optimally this would all be done in dead calm wind (fat chance).
  
• It's hopefully long enough (7.8 miles) to show a significant difference between the bikes (assuming there is any difference), and not so long that it's beyond my meager athletic abilities to hold a substantial wattage average. From past experience, I knew I could hold 225 watts over this course without blowing up; of course, I was going to have to do this 4 times, since I wanted at least 2 runs from each bike to reduce the chance of something anomalous screwing up the data. Of course, 2 runs is a paltry sample, but....I also had to compromise with the tensions of time, life, my athletic ability, logistics, and the fact that I need to get the Corsa back to it's owner within a reasonable length of time. It ate up a bunch of time just preparing for all this, getting the seat struts, etc. In addition, it was about 5-7 miles from my house to the test spot; before the test spot were too many stop signs, etc.. I lack a vehicle which can haul both bikes, so I'm going to be getting an extra 20-30 miles of riding just setting things up; this all had to happen in one day, so....

So yesterday (April 18, 2010) everything came together. I got the seat angles equalized, it was a sunny day, and the wind was a steady 5 mph.

OK, enough setup; here's the results:

Corsa:

Split	Time	Distance	Avg Speed	Avg Power
1	0:21:56	7.78	21.3	225
3	0:22:03	7.82	21.3	225

TiCa:

Split	Time	Distance	Avg Speed	Avg Power
2	0:21:16	7.79	22	225
4	0:21:18	7.82	22	225

So, the TiCa was faster. Both bikes turned in basically identical performances from one run to the next, and the TiCa was somewhat faster, although not dramatically so.

So, what does this "prove"? Ummm.....it's pretty murky, actually. I would say that it is evidence that lowracers are somewhat faster than hiracers at the same seat angle over 'real world' roads. It isn't 'proof', by any means. If it proves anything, it proves the TiCa (a very fast lowracer) was faster than the Corsa (which is not Bacchetta's flagship, although it has been ridden very successfully to multiple race wins) on this day, with this rider, with every reasonable attempt made to make the tests fair.

Some may ask about climbing. Well.....climbing is difficult to quantify; you can do multiple hill repeats on each bike, but....you can't equalize out the variables of rider exhaustion, varying fitness from day to day (not to mention weight fluctuation), and the inherent advantage of whichever bike gets to go first in the test. Doing a constant average wattage would probably just favor the lightest bike; there's also the question of which bike allows the rider to deliver the most wattage climbing (since climbing is more of a 'peak output' activity, this seems more part of the question than on the more varied course I picked for the test).

Anyway, all I can offer is anecdotal 'evidence' (my experience). I seem to climb much better on the TiCa than the Corsa. Most of my climbing yesterday was on the way to and from the test locations (also complete with lots of stopsigns; not a good area to test). The Corsa had the advantage of being first, but.....it was significantly easier to climb on the TiCa. This probably says more about my adaptation to that bike (thousands of miles) than any inherent climbing advantage to the bike.

That's it for now. Since it turns out my TiCa is 23 degrees, and the CarBent I've got coming is also going to be 23 degrees; I may have more opportunities to do some comparative testing. And it'll be a somewhat better test, since I can use the very same wheel for each bike (I'm getting a dual 700c Raven). Also, since I have a Zipp 404 front wheel for the Raven, I'll be able to contrast that with the HED Jet 40 wheel I have on my other TiCa, which also has a non-aero Ti from fork, etc. It'll be interesting to see if any of these things are significant.