Longley,
Rapid prototype (RP) techniques like stereolithography (SLA) of polymer resins or direct laser sintering (DLS) of metal powders are not used as much as you might imagine. While parts produced by these methods require no tooling and can be made fairly rapidly, they still have many practical limitations. They are mostly used for "show-and-tell" sub-scale samples or models. And with cheap computing power combined with sophisticated modelling and analysis tools, the need for actual testing grows less and less every year.
The primary limitation of these RP techniques is part size. The largest rapid prototype machines can only produce parts of about 18 inches in any dimension, and larger parts must be made in pieces and glued together, giving inferior results. Most big-budget race teams (like F1) utilize near-full or full scale wind tunnels to ensure maximum accuracy. So making SLA wind tunnels models is not really practical.
DLS of metal powders (like titanium) can actually produce a fairly good part, but it is slow. With the advent of lower-cost, high-speed, multi-axis machining centers, sophisticated CAM softwares, and high fidelity CAD modelling, it ends up being just as cost effective (and quicker) to machine metal parts from billet.
As for composite tools, due to the low production rates required for racing, the tools are composite molds laid-up over five-axis machined foam plugs.
Conceptual,
I believe the highest performance matrix (resin) material that is readily available are thermoplastics. They are difficult to work with since they require heat and pressure to apply, and must be heat cured. But the cured laminate has amazing properties, both with regards to strength and temperature.
Good reading:
http://www.fiberforge.com/documents/Fib ... 006ppt.pdf
Regards,
Terry
