Posted this in response to a thread in a time trialing discussion forum; the poster wanted to do aero testing without having to go to a tunnel. The Chung method refers to a protocol that enables very accurate analysis based on power meter data (do a Google search on "Chung Method"). As the poster wanted to keep things simple, I offered the following:
A PDF on the Chung method can be found here:
Also, look at the one of the late chapters in Coggin's Training with a Power Meter
The Chung method can yield data that's comparable to that of a wind tunnel--I use this method regularly in-season and valid with an off season tunnel session.
Very important point above--focus on one factor at a time (helmet, or arm position, etc.).
If you don't want to do the math, then a velo and PM will provide you with decent trend data. For instance, 2km at 70% race speed, monitor wattage (or, keep wattage consistent and measure time). The key is being able to do multiple efforts during the same session.
Five 2km trials at 26mph with arms flat (average results/discard outlier data/etc.)
Five 2km trials at 26 mph with arms angled up (i.e., Landis)
Assuming that you can hold the same speed for each of the trials and that the temperature, wind, and humidity doesn't vary during your testing session, you'll get some interesting trend information ("I appear to be using less watts to ride at a given speed with my arms angled up than when they are flat").
The more trials per
variable, the better; this will minimize noise.