What to learn (if anything) from wind tunnel smoke lines images?
 

Looks like they are becoming marketing tools getting away from aero research.
At the right angle and wind speed any car might look to have a smooth flow.

Besides some obvious and extreme turbulence and separation like this
http://image.corvettefever.com/f/175...l_testing+.jpg
http://image.corvettefever.com/f/175...l_testing+.jpg
http://inter.action.free.fr/images/a...t-aerodyne.jpg
http://images.hemmings.com/wp-conten...ro_04_1200.jpg

What can you learn from it?
Does the height/angle difference before vs after the car for the lower smoke line means something?
What about the upper lines?
http://cdn.greenoptimistic.com/wp-co...ynamic_car.jpg

I agree, but keep in mind that recently taken wind tunnel pictures and videos of +50 year old cars would not be considered research other than in a historic contex anyway.

I think this falls into the "Art" category, similar to having a stone wall, castle, abandoned factory or pretty girl next to the car.

Some of these images are just meant to be cool.

"it's all blue smoke and mirrors"

Different information becomes available when you employ curtains of air bubbles in moving water.

http://ecomodder.com/forum/member-fr...74-mini14l.jpg

They can be used experimentally

http://ecomodder.com/forum/member-fr...25-7-36-55.png

http://ecomodder.com/forum/member-fr...70-imagesa.jpg

That last one might be polarized light.

learn
 

It would be the images never published which would be of greatest value.
If the aerodynamicist is locked into a 'hard styling' protocol,where the finished body must very much resemble what the stylist signed his name to,then the smoke can reveal 'optimization' potential as was done with hundreds of cars now.
A full-scale clay can be altered by master clay modellers right in the wind tunnel and sequential runs made to evaluate any progress.They'll literally camp out at the tunnel,running 24-hours a day until they dial in a shape.
Scale models and CFD aren't able to allow such alteration and evaluation.
Lee Iacocca said that many Chrysler products were saved this way.
What we're allowed are centerline flow images of dubious value.But hey,in the immortal words of Bob Parsons,"they're way out ahead of nothing."

Waht about using CFD to design an optimum shape for, say, a small 2 seat, side by side sportscar?

In other words, start with the side by side seating, then design around that. Whatever shape works best for side by side seating, mid engine is fine.

Thoughts?

Who needs CFD or an actual wind tunnel? Just fire up a car photo in a graphics program, load a fuzzy spray brush and paint your own. Make it look as cool, fast, and efficient as you want. I could probably make a brick look sexy fast by drawing my own lines.
How is anyone going to prove this wrong?
This is just eye candy....some are real but the results don't mean anything as they only present the photo that looks the coolest. No raw data, no raw or failed photos, no raw or failed CFD results. No time averaged info....
The visuals are not proof. Raw data can be proof. The data is rarely shown and even the real published data has disclaimers.
Data is not very nice to look at (which is why they draw/select the pretty pictures).

I think it's been done. What about CFD models of two vehicles, say the Volkhart-Sagitta and the Lambroghini Egoistica, with equal numbers of vertexes so they can be morphed into all the intermediate shapes.

Big time : Excellent question that I too have wondered about.
Also, why have they done away with tuft testing these days ? Tuft testing shows so much just by looking at the direction of the tufts as well as the flutter.

Good point about the smoke test making any car / truck / car-truck ( SUV ) / brick etc look aerodynamic.

The only time that a smoke test seems useful to me is when they have a situation like this one :http://www.steveohalloran.com/blog/archives/F150.jpg

Just look at this truck : http://0.tqn.com/d/trucks/1/0/C/F/1/...ind_tunnel.jpg

The flow looks almost as smooth as this one ! : http://www.electric-bikes.com/better...kgroundhtc.jpg

what about
 

I suspect that today,this is exactly how one would proceed.
I'd start with freebeard's Saggitta and the 1939 Porsche 60K10.