Originally Posted by
wphamilton
Think of a ribbon flapping in the wind. It behaves differently at different wind velocities, thinking of turbulent motion. Mass, density, length, how stiff they are, wind speed all go into mix.
Re=rho U L / mu. The Reynolds number over normal cycling speeds doesn't change enough to affect CdA, therefore power scales as the cube of velocity. The questions you raise were answered over a hundred years ago. As to wind tunnels and real word testing, well, the referenced paper and subsequent validations should put that question to rest for good. What real world variables* would you suggest are not included in the model?
*as opposed to input parameters