Seems like something a nesting algorithm would help with. Unsure of the precise workflow but maybe something like this:
Input panel to be cut geometry
Define plywood panel sizes
Apply panel geo on ‘tiles’ of plywood
Algorithm objective for smallest area difference between populated panels on the tiled plywood
Result should apply geometry on tiled panels
I believe some mass manufacturing methods use nesting algorithms for similar applications. Might be useful to look into that depending on project scale. Hope this helps.
If the NGon is split across plywood panels how will the individual tiles be joined?
Will they be simply butted together without overlap?
Or will the joints between tiles be scraph joints or lap joints with the tiles overlap? If so and the overlap then the results will not be useful for actual construction.
What allowances are needed for the width of the joints?
If the NGon is split across plywood panels how will the individual tiles be joined? - there will be no joints, the panels will be fastened to a frame, without overlap.
“Algorithm objective for smallest area difference between populated panels on the tiled plywood”
I’m not sure I follow your logic… smallest area difference between populated panels… would this solve the minimal waste condition?
would you sim this in Kangaroo?
my current thinking is to array the tile, move array to the parts bounding box, intersect the curves and output individual parts. I still work on determining the most efficient way to position the tiles for minimum waste. ..
I was thinking maybe some sort of machine learning objective that minimizes the wasted area to solve the minimal waste condition… a lil over my head so that was just some wishful thinking I suppose.
Maybe there is some way to include all the shapes to be cut in a kangaroo sim. I am only really familiar with form finding in kangaroo but I suppose it could work.