Are you going to use "conventional" side mirrors, or do you plan on video cameras?
It's not soon in the process design process to be considering something like what Nissan did with the LEAF; an up-stream air-splitter for the morrors. In Nissan's case they say the very strange, sticks-up-above-the-hood/body-panels turn-signal/running lights contribute significantly to reduced wind noise from the side mirrors, if not actual turbulence reduction.
Lest I forget, you're doing an excellent job keeping us informed about your fascinating project. Thank you and keep up the good work. (Are you getting much sleep?)
I probably will use video cameras -- they might be able to be behind the side windows, because the sides taper in, so there is a good line of sight to the rear without protruding out from the car? If I do optical mirrors (which use zero power, naturally) I would try to have a gentle "half teardrop" bubble on the side.
The quarter scale model actually has small bumps where the video cameras could go, but I think the scale and accuracy (or lack thereof) will make it tough to judge how well this will work.
Another video with more tufts (I need a lot more pearl cotton yarn than I thought), and with the blower slightly higher up. (I'll get a second blower later this week.)
I still don't think this is an accurate representation of what would happen when the car is moving through the air, like in a wind tunnel. But, it does show the reversal at the wiper pocket, and the stagnant point is fairly obvious. There is a lot more "lateral" flow than I would have guessed.
I emailed the folks at the MIT wind tunnel today, and they were prompt in the their reply. They have a busy schedule -- late summer is the first opening. :-( And it costs a minimum of $1,000 for 4 hours... I can't wait that long to start building the shell. :-(
I'm going to do some more work with my tuft testing and see what I can figure out to improve the flow...
I think I will have to try a sheet of plywood on t he roof of my car, with a couple of video cameras and a acrylic panel under the model, so the tufts on the bottom can be seen. Pocket HD video cameras are $100 each, and a sheet of plywood is about $50. The speed is not realistic for a quarter scale model, but the general air flow on the surface should be realistic.
A full sized shell will cost about $1,000 or so... I'd rather do that with my money.
Finally, some useful-ish results doing tuft testing with three leaf blowers!
Click on thumbnail for a link to the album. Some photos are with flash, so they freeze the tufts, and some are without, so some of their motion is apparent.