Freesteel Blog » Version zero of valley machining

Version zero of valley machining

Thursday, December 11th, 2008 at 12:40 pm Written by:

This machining strategy appeared in Vero’s machining strategist a couple of years ago, though it bears a lot of similarities to a Work NC strategy I was shown way back in 1994 when I was driven from NC Graphics all the way over to Depo in Germany to see the product with a view to making our own system that had as good a capabilities. In 1994 Work NC it was a lace pattern, with the toolpaths zigzagging perpendicular to what you see in the above image, but the idea and implementation is the same, bar a surprisingly small change in the code.

Years later we had something in machining strategist that found the complex boundaries of the rest material and ran a constant scallop stepover cycle from them, which was very hard to program and suffers from all sorts of problems of sharp changes in direction and wide collapse areas half the time. This more morphed result was what users really wanted, but of course I always knew better, so I didn’t make it back then.

The attraction with the scallop stepover is that it theoretically works everywhere, no matter how complex the area. Users could draw their ideal results on something straightforward like in the picture, but they didn’t seem to be concerned that the idea didn’t generalize very effectively when things got complicated, as they always do.

(I don’t know that, actually. It’s possible that 99.99% of parts were as simple as the above picture, but then they don’t produce bugs, and then they don’t get sent in, and then I don’t see them. I only got to see an exponentially increasing sequence of utter bastard parts as the years progressed and CAD systems could sprew out double the number of surfaces every 18 months, which would break everything.)

So here goes. New machining strategy under very early development which someone else has managed to make work. Lots of work to be done projecting the small tool to the surface where it belongs, planning the structure (in high level python), and smoothing it, and I’ll probably try and do it for toroidal cutters instead of only spheres. That will at least represent progress, rather than merely being copycat. Of course, 99.99% of users are probably perfectly satisfied with doing this machining with a ballnosed cutter, though I don’t get to see this, because I’ll hear about the 0.01% who’d quite like to try it with a toroidal cutter and are unaware of the 10X more grief it takes to code for this tiny unpopular case.

2 Comments

  • 1. lezdep replies at 14th January 2009, 3:26 am :

    I’ve done my first machining using toolpath generated by your Adaptive Clearing demo. I was
    trying to do 2 side machining, which demo does not support and of course had few scary moments, because algorithm assumed it approached stock from outside, while in fact it was
    plunging right trough. I was machining Denril, so machine and stock survived somehow. Beside
    that algorithm done good job. Finishing toolpath was generated using pycam. That’s open source
    project I contribute to. Imho, it has good potential and would love to help integrate Adaptive Clearing into pycam. Let me know if you would be willing to work with us – lezde at rogers dot com.

    Following are pictures of results. I’ll put some videos on youtube.com later.

  • 2. Freesteel&hellip replies at 4th February 2009, 6:19 pm :

    […] the fastest most parallizable CAM algorithms are being written, we’re attempting to code up valley machining which is going to rely on the not-yet-written non-axial tool projection […]

Leave a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <blockquote cite=""> <code> <em> <strong>