The circle centers are supposed to remain vertically aligned, and I believe horizontally aligned as a result aswell. But the math seems to change dependent on whether the oblongs touch on the round or the straight sides?
I am trying to rotate a oblong through 360 degrees and tile it as above. An oblong is essentially a rectangle with semicircles at opposing ends of one direction of the rectangle. Like the shape of a horserace track.
I am close to getting the vertical displacement down using a piece-wise function. When I have that piece of code done, I’ll post it.
I’d love to have a crack at this a bit more in depth but currently bogged down with deadlines.
But the way I see it is as if it has varying degrees of semiregular tile patterns (when I’d initially expect it to be regular by default since it is a singular piece) depending on the rotation of the tiles past 90 degrees. what It’d be curious about is if it could be solve through laying down an elongated triangular tiling frame and seeing if the pieces tie together?
We would treat them all as triangles/squares but we could then overlay the arc on our “triangular” tips, quick sketch of what I mean below (missing the “tie in” triangulation in the rotated options as my hand pulse is currently failing me…):
I didn’t look extensively at your code but the math is there.. just copy the row in x direction by the distance 2Rcos(theta) and rotate them (180 - theta)
Edit: maybe I got the sign convention wrong, however something like this should work