Hay, all! I've just read this thread, and figured I would post a message with some notes. I've been interested in fractal antennas for wifi purposes (of course), and have also built myself a SBGH out of waste materials. I'm looking to build a system for fractal antenna design - just to see how to do it.
Fabrication - I figured that since the fractals are self-similar, you should be able to print out a design of a smaller subset of the whole piece, and make a bunch of smaller pieces. You should be able to screw them all together onto a substrate to make the whole pattern. This would allow for more uniform pieces.
Having access to a waterjet/laser cutter would probably be the best way to fabricate some of these shapes, such that they are as exact as possible. The more complex shapes would eat a lot of material unless they were broken down into smaller pieces as mentioned above.
(I assume that having an element that is exact in size and position will have more ideal electrical features. I can't help wondering if my homebuilt SBGH is 'lumpy' in some way, and not working as well as it could. Wouldn't this be exacerbated when using fractal antennas?)
Modeling and Mining - I'm also sure there are variations in the shape that may improve with minor adjustments. I was wondering if L-systems may be used to 'grow' a better antenna (http://en.wikipedia.org/wiki/L-systems
). Think the Folding@Home for antennas.
My plan was to write some terrible shell scripts to couple the output of the L-systems generator to whatever 4NEC variant I could get working with a qualifier at the end.
I'll go read the modeling software thread.