Race Mirrors claims that their
Pro Aero Extreme (!) model has 55% less drag thanks to the transparent cone extension.
I was pondering something similar to this. I'm glad I wasn't going crazy (on my own). Imagine a short rugby ball shape with the mirror just to the rear of the widest point. Then cut away the minimum at the rear to give the sight lines needed for the rider's eye. Maybe some of the "shell" does need to be transparent. Maybe not.
Obviously the shell diameter needs to be larger than the mirror, but the Cd (cw) can come down. It's trying to enable closure behind the mirror, reducing the separated volume there. This shell shape idea would also improve the interaction with flows over the car. The negative synergy, the whole being less than the sum of the parts, is going to be the issue with mirrors. So reducing the penalty from disturbed flow over the VM body is worthwhile.
Testing in the wind tunnel, u sing mirror shapes with well known drag (flat plate, half sphere) and streamlined struts with known Cd, tests with/without the mirror assembly will show the penalty from mirror body interaction. The mirror and strut drag is calculable so the difference can show.
Like
rash in post #165 above, I was pondering the relative size of the drag from simple mirrors. Comparing them to the total drag (cwA) values from Jan's tunnel tests, looking at the lowest Milan values there....I calculate two half sphere mirrors with CD
0 (cw)=0.38 giving drag about 3.2% the size of the tested Drag of the Milan at low speed. Was 3.6% at high speed.
They sound small values, especially if one is not chasing drag in the first place. I hope that the tunnel tests will give a fuller picture, values will be higher.
For my calculations I referred to Hoerner's Fluid Dynamic Drag, about section 3-17 and Fig 33 for the common Cd
0 or cw values.
Considering the variation of Cd
0 with Re, I followed his notes, graphs that Cd
0 was fairly constant 10e3<Re<10e5. (section 3-8, Figs. 10, 12, 13).
At Jan's tunnel test Vs, the Re for a 50mm diam mirror was 34e3<Re<64e3.
Edit: Trying to have good habits and follow Hoerner's nomenclature...Cd
0 meaning drag coef with cross-sectional area as the reference area. The space between the d and the 0 subscript comes somehow automatically when I upload, and is not correctable (moderators please note).
Gregg....