Today I tried to get a 1:5 model of the blades out of a chunk of wood to see what kind of troubles I'll run into when I try the full scale one. I wrote a bit of python to give me a 'job' file that the plasmacutter program will accept, ran into a few bugs in the software (I never tried true 3D work yet), and fixed them.
I rigged a clamp to an ancient black & decker router, ran a few tests and then set it up to route the front side of the model.
It works, but the finish is pretty rough (lots of runout on the bearings of the router) especially near the root portion of the blade where it bites pretty deep into the wood. Eventually we'll be routing foam, so I don't think it will be a problem, but right now it looks real ugly in some places. Also the blade is 'backwards', if you were to make a windmill with this one it would spin counterclockwise, I have changed that in the software to get a 'proper' blade.
It's funny how I was totally confused as to how the blade actually sits on the hub, I somehow pictured it with the airfoil portion pointing *into* the wind, but that's not the way it should be. I checked Hugh Piggott's book one more time and found out that I did indeed have it completely backwards...
One big problem that I don't have a good solution for is how to make sure that the blade top and bottom portion will 'register' properly when I flip over the base material to route the other side.
The cutter table looks pretty rough right now (some rust even!), I haven't gotten around to painting it any further because I'm too busy with the windmill, I hope to get to that in the winter. It is in its current state because I recently redid the whole gantry system to reduce vibration (the trusses make it much lighter).
The blade profile is a little wider at the tip than the norm, that's to keep the noise level down. If there are strong reasons for narrow tips other than weight reduction then I'd like to know about it. I have heard some narrow tipped blades and I'm amazed at the noise levels they put out, some really scream (especially plastic injection molded blades). The wider ones seem to be much more quiet.
Another - minor - gripe is that the router bit is flat at the tip, I could not find a ball shaped bit that has a long enough cutting portion to do the whole job (unless I cut the profile in several passes, but in foam the nice thing would be to route the whole thing in one go). This 'flat' portion causes a dimensional error in the steeper portions of the cut.
pictures follow:
The router clamp, it fits the toolmount on the z axis slide of the cutter:

Side view of the router mounted on the slide:

Top view:

Halfway into the cutting:

Finished front side (but reversed):

Semi intelligent vacuum cleaner mount:
