From a post to BentRiderOnline about one of the coastdown tests we have performed
It does appear that there is a wide range of rider experiences out there when it comes to fairings of any kind. Heck, when it comes to bikes of any kind!
A couple years back, we did a fairly extensive test of our TailSok rear fairings with the help of folks from the Oregon Human Powered Vehicles Association. The goal was to get some real data, and we pretty much succeeded in that. It turns out that testing is a complicated thing.
The test we did was a coastdown test, no pedalling, and we recorded the max speed. We had a long enough, steady enough hill that we were sure we reached terminal velocity. We got a lot of data, some of which was very useful and some of which was not.
Some of the runs were with full body socks, and some runs were orphaned in that we didn't get the second configuration tested in time. A few runs were invalidated because people changed jackets or tires. The hill was steeper than we wanted, but was the only one of about 20 candidate locations scouted that met the criteria for steadiness of grade, course safety, and proximity to Portland.
Here's what we did: we found a long, steady hill with good pavement and little traffic (steady 6% grade for 2 miles). We invited people from the Oregon Human Powered Vehicles Association to come and test their bikes. People coasted down the hill, and we recorded their max speeds in various configurations of fairing/no fairing. We went for at least 3 runs per person/bike/configuration. Two vehicles ferried people back to the top. We got in about 68 runs in 4 hours.
We had three course marshals, a bike safety officer, a photographer, and a data clerk. There was quite a turnout, and everyone had a good time and vowed to come back. The OHPV gang is a great bunch!
In amongst the data, we got 22 directly comparable tailsock off/tailsock on runs. Here are the numbers:
- Average of 22 runs, 7 bikes, without sock: 39.67mph
- Average of 22 runs, 7 bikes, with sock: 42.00mph
- Difference: +2.33mph for sock
This is from 7 different bikes, with an equal number of runs with/without from each bike. Most bikes had 3 runs each; one bike had 4 runs each. Bikes included a Django, a Giro 26 (with front fairing), a Corsa, an Aero, a Zox clone, and a couple of Easy Racers.
The improvement for each of the 7 bikes ranged from +1.23 to +3.83 mph.
We learned a lot, and, next time, we plan to do some things a little differently. For one, we'd like to find a course with a 3-4% grade so max speeds won't be almost 2x typical cruising speed. In subsequent tests on a different hill, the speed percentages held true. It was good to have data at 20mph instead of 40, but we'd like to have tested more riders there. Second, we'd like to test fewer bikes but do more runs per bike. Third, we'd like to record not only max speed, but the overall time as well. We were going to do this, but the radios didn't show up. The other thing we'd like to record is bike+rider weight. We also plan to have one bike that goes down frequently without changing configuration, to serve as a control.
We'd really like to understand why some bikes/riders show big improvements and others only small improvements. One thing everyone does find is that the socks are highly visible, particularly the neon ones. We get at least as many comments about visibility as we do speed.