I never thought making those foam blanks would be this hard. Anyway, it's getting there.
After 3 days of pretty hard work I have something that looks a little bit like a windmill blade in front of me. It's not yet fit for consumption but it has recognizeable twist and taper, as well as a decent airfoil shape. Hopefully tomorrow I'll have it perfect.
One of the more nasty little headaches in 3D cam is that you don't really have an infinitely small bit that traces infinitely small paths. Real world router/milling bits are anywhere from .125 to .750 " in diameter for this kind of work. The best ones are 'Ball' shaped, uand have a long shank.
In order to compensate for the cutter diameter you have to perform an operation called radius compensation. The cutter software I wrote already does that, but only in the 'X/Y plane'. That's where the plasma cutter operates, Z isn't an issue at all, except for warp control and obstacle avoidance.
In order to do true 3D sculpting you have to compensate for the cutter radius in the 'Z' axis as well. This took quite a bit of fiddling but I think I have that licked.
The Z axis problem I had yesterday had nothing to do with the weight of the router, I messed up in my 3d adaptation of Bresenhams Line algorithm. That's fixed now and the cutter head now keeps perfect Z position. No need for springs to compensate for step loss because there wasn't any.
btw I *hate* routers... what an unbelieveable mess and noise. I got a tip to reduce the RPM on the router to cut down on the 'snow' (thanks again, Rob) that comes of the foam, I'll definitely look into that tomorrow, this is getting out of hand.