The data you get out will not be better than the data you put in.
Here is a simple method I used to make a model of my Insight in Pro/E. I
then added a tail extension to the model and got an idea of the size and shape of that tail extension.
I tried running the model through a few commercial CFD packages. Programs that output just a streamline illustration tend to be very simple to use, and I was able to get results (of unknown validity) from them. I also tried ANSYS 10, which can spit out a CdA for direct comparison between the established value for the car, the "before" model, and the "after" model. Unfortunately, I was unable to generate a valid mesh in ANSYS CFX.
Some programs can produce pressure and velocity distributions, which would be helpful for identifying problem areas on the model,
IFF the model does not deviate from the car in aerodynamically significant ways. The underside of the car is a great example of something that's hard to model. The wheels ought to turn, the road ought to move, and the underbody should be fully modeled with brake cables, suspension, and exhaust hanging down. I would hesitate to make changes to the car based only on the result of a CFD simulation that uses an incomplete model of the car.
BotSapper took a 3DStudioMax model of an Insight, available on the Internet, and did this with it:
http://ecomodder.com/forum/showthrea...tml#post168928
If you can convert a .3ds to a mesh for CFD, I'm sure I'm not the only one who would love to see Botsapper's model (with an aero tweak or two) run through the simulation with and without a boat tail.
If you'd like a copy of any of my .dxf files, I'll make them available to you.