KUKA PRC unwanted orientation flip

[David_Zhou]

Even though the tool and toolpath plans are oriented for a simple A2 / A3 rotation downwards from starting position, A4 and A5 perform a really odd 180 turn before reaching their starting positions. Video is attached for clarity. For further context, I’m hoping to figure out some toolpaths from nonplanar 3D printing, but if I can’t even get the planar stuff down, I’m pretty worried . Any help/ general tips on relative plane definitions would be much appreciated!

sample toolpath2.gh (15.3 KB)
toolpath setup2.3dm (10.6 MB)

[inno]

the very first command the arm gets in your definition is to move LIN:

you might want to PTP somewhere closer to working area first,

This is just for showing the final result: the arm moves PTP to the neigborhoods first (honestly… even too close…) then LINs along the final path

sample toolpath2_Re.gh (19.0 KB)

[images taken from KUKA System Software 8.3]

[David_Zhou]

Thanks so much for the help! adding an intermediary PTP move seems to solve the issue for most of the plane orientations EXCEPT the specific one I’m looking to implement(as described in the tool setup). Attached are the files in question with the above suggestion added. The robot still unfortunately twists in the undesirable manner described in the OP. I’ve tried changing how close the initial P2P move gets to the initial position and added more intermediary P2P points before starting the LIN moves, but it hasn’t yielded any results for me yet unfortunately. Once again, any help is much appreciated!

sample toolpath2.gh (20.0 KB)
toolpath setup3.3dm (9.7 MB)

[inno]

the new tool setup in the last posted file looks like a “perfect translation” of the working planes

I would just take the definition you originally posted in your first post, replace the old tool with the new one and use it as it is (only LINs)

sample toolpath2_original_file_new_tool.gh (12.6 KB)