ArtCAM Not Turning Indexer a Full 360

[jdervin]

I was doing a test cut for a fairly simple turned table leg on the indexer and encountererd something new. There are two individual turning zones on the leg (otherwise it's square) and the cutting file generated by ArtCAM is turning a portion of one zone less than a full 360°. In the roughing file, the first two slices in that zone raster between 178.984° and -179.941°. Starting with the third slice, the raster rotates between 159.229° and -178.984°, leaving an uncut "wall" along the back side of the turning (see photo below).

I've often wondered why my files don't simply raster between 180° and -180°, but assumed it had something to do with bit diameter or imperfections in calculating PI to convert a circumference to a length. This time though, I can't explain why the cutting file suddenly changes at a new depth, and also wonder why the problem only happens in one of the two zones (both were created on the same model).

The finishing file behaved similarly, but the same problem zone had additional steps at the end of each rotation: after stopping at the "wall", the bit lifted off the piece, the B-axis finished rotating to the 180° point, then rotated back (after stepping over), dropped back down and continued cutting.

I've poured through the data in the .sbp files and scrutinized the ArtCAM model, but can't see any reason for these results. Any suggestions would be appreciated.

[bob_dodd]

John ; I think it may have something to do with the size of your relief & size of your model and the bounding vector if your using one , Artcam likes to Z up if your using a bounding box (Vector ) to select your maching area and the bounding vector is the same size or larger than thr releif , if your using one try making it a couple of thousands smaller , also your releif needs to be 3.14 X dia. in your wrapping direction Hope this helps

[jdervin]

Bob--

I'm pretty sure I've got the 3.14 thing in place. And yes, I'm using a bounding box that is exactly the flattened circumference of the wrapping cylinder. I'll try reducing it to see what happens.

Thanks for the tip.

[jdervin]

I think I've got it solved now. It was a combination of two problems: one of my bounding boxes was no longer centered (not sure how I accomplished that...) and Bob's suggestion of reducing the bounding box slightly. It's counter intuitive as I would think that decreasing the size of the box would only increase the uncut area, but I checked it both ways (larger and smaller) and it absolutely works with the slight reduction.

These corrections not only fixed the problem, but reduced a ~3500 line file by close to 2000 lines of code (lots of unnecessary Z moves). I haven't had the chance to cut another sample yet, but I'm guessing I'll be shaving off a nice chunk of cutting time.