Maybe a way to make rotary toolpaths from complex 3D models.

Post by **tahustvedt** » Sun Jan 04, 2009 3:37 pm

Normally I use Discreet Plasma for my 3D modelling, but it can't extract the unwrapped surface geometry. I had an old trial version of 3D Studio MAX 7 which I installed to have a look after I found an unwrapping tutorial on the interweb, and I actually got it to work. Max and Plasma work almost the same way, but Max has a lot more features (and costs three arms, a leg, and a couple of gransmas). I have been looking for an unwrapping software on the internet periodically for a year now but have not had any luck.

I'd like to try milling a propeller blade from round stock in one finishing pass with this technique. Starting from the tip of the blade and working inwards while the rotating axis rotates the stock around. The first picture shows a 3D model of the blade I designed.

1. I start by aligning the axis of the blade so that it goes through the blade and exits at the center of the tip, and not under the bottom surface for example.

2. Then I applied a cylindrical UV map, and rotated it so that the split (the green line) is at the bottom. I think it's important that the cylinder is circular and not elliptical in crossection.

3. Then I open the geometry channel info and copy the UV map channel and paste it over the mesh geometry channel, overwriting the originalmesh with the new UV mesh generated by the cylinder projection. In Max the result is a tiny, rotated, unwrapped model. I'm not sure why it's so small and rotated, but it's perfectly unwrapped.

4. I rotate it back until it's level again and scale it up to the right size. The height should be the same as the distance from the axis to the widest point on the original propeller model. Then I clean up the edges by removing any redundant vertical faces, and move all the stray faces over to the opposite side to make it look cleaner and avoid problems in Aspire. In the fourth picture I have also copied the model and made it couble to make the finish better when I machine it. I lifted the edges so that the tool ramps on and off the blade when it mills.

5. The model in Aspire after simulating a finishing pass looks like it'll work. I use the normal technique for 3D models, except I did not offset the vectors (fit vectors to bitmap) around the 3D model because it would gouge the blade at the edges. I want it to stop exactly at the edge of the model. since I will try to raster from left to right, starting in air just beyond the stock, I haven't bothered making a roughing toolpath as the flutes are long enough on my tool. The height should be zeroed at the right distance from the center of the rotating axis, and the number of steps per rotation will need to be calculated in Mach 3 to fool Mach3 into thinking it's milling a flat axis when it's actually rotating the piece twice for each pass.

I haven't tried it yet as I'm waiting for a new motor for the rotating axis and a T-nut table, but I think it should work great. I'm excited to try it out. Until now I have milled the blades top first and then bottom, which isn't problematic, but it would be much quicker (not to mention cooler) if I could mill it in one finishing go on the rotating axis. there are other benefits as well that I won't go into.

If anyone knows of a reasonable software that can unwrap the UV map to a 3D geometry then please tell me about it.