|Position||Water Bottles||Rear Rack?||Back- pack?||Weight (lb)||Start (mph)||Max (mph)||30mph Drag (lbs)|
|brake hoods (comfortable)||2||y||y||203||30||32.5||10.0|
|brake hoods (comfortable)||2||y||y||203||30||34.0||8.8|
|brake hoods (comfortable)||2||y||y||203||30||33.0||9.6|
|drop bars (comfortable)||2||y||y||203||30||34.0||8.8|
|drop bars (comfortable)||2||y||y||203||30||33.0||9.6|
|drop bars (ducking), slight crosswind||2||y||n||195||25||32.0||9.1|
|drop bars (spinning), slight crosswind||2||y||n||195||27||32.0||9.5|
|aerobars (pistol grip)||2||y||y||203||30||35.0||8.0|
|aerobars (pistol grip)||2||y||y||203||30||35.5||7.6|
|aerobars (one hand over/one under)||2||y||y||203||30||36.0||7.3|
|aerobars (one hand over/one under)||2||y||y||203||30||35.5||7.6|
|aerobars (over/under), slight crosswind||2||y||n||195||25||33.0||8.3|
|aerobars (over/under/spinning), slight crosswind||2||y||n||195||25||32.5||8.8|
As a first test, I went out on a calm day and climbed Whiskey Hill a bunch of times. I chose Whiskey Hill because it's straight, low traffic, and short enough that I could climb it multiple times. For each run, I'd accelerate to around 30mph, start coasting at the mailbox at 550, and coast until the bottom of the hill. My cyclocomputer recorded the max speed hit.
I carefully calibrated my gradiometer before measuring whiskey hill, and hope that each sample was within 0.5% grade. From the grade, I computed the elevation. This seemed more accurate than my altimeter. Here's the data file, with distance (in miles), grade (in percent), and altitude (in feet):
0.00 1.0% 350.0 ft 650 mailbox 0.01 1.5 350.5 0.02 1.25 351.3 0.03 3.0 352.0 0.04 1.75 353.6 0.05 3.5 354.5 0.06 4.0 356.3 0.07 4.5 358.4 0.08 4.5 360.8 0.09 5.5 363.2 0.10 5.75 366.1 0.11 6.5 369.1 0.12 7.0 372.6 0.13 6.5 376.3 0.14 6.5 379.7 0.15 7.0 383.1 0.16 6.5 386.8 0.17 7.0 390.3 0.18 6.5 394.0 0.19 6.0 397.4 0.20 6.5 400.6 560 mailbox 0.21 5.5 404.0 0.22 5.5 406.9 0.23 4.5 409.8 0.24 3.75 412.2 0.25 2.75 414.2 0.26 1.75 415.6 0.265 2.0 416.1 550 mailboxI computed the drag by simulating each run with a simple C program , which assumed that total drag was equal to some K * velocity squared. It computes K, and then figures out how many pounds of drag that would be at 30mph.
I believe my grade measurements to be within about 0.5%, which would make a difference of up to about 8% in the final drag numbers. Winds also affect the drag -- I have to take many samples to average out the effects due to wind.
It's interesting to note that Jim Martin (see links below) lists 8 pounds of drag at 30mph as "typical", and 6 pounds of drag at 30mph as "excellent". Thus, even with aerobars, I'm still little better than what he considers "typical". Perhaps I should remove the rear rack and try it again... :-)
Although I didn't take a large number of samples, the general trend of
the above table seems to indicate that aerobars reduce drag by roughly
20%. This means that you can cruise at the same speed at 20% less
effort, or ride 8% faster at the same effort. These numbers match
closely with my experience.
Aerobars are (in my opinion) the most effective single thing you can do to improve flatland speed. In my tests, the aerobars reduced drag by about 1.4 pounds at 30mph, compared to a comfortable (non-crouching) position on the drop bars. The next best improvement would be aero wheels, which (according to Jim Martin) can reduce drag by about 0.4 pounds. I'll have to acquire some aero wheels someday, in the interest of science... :-)
Links to Other Aerodynamics Pages