Just a quick question: Is it possible to align circles made by the CNR component onto a curved surface such that it flushed with the surface curvature. much like tiling a curved surface with circles.
i think you may first look out how to tile your surface. because if you want to maintain same sizes, i assume that also your tiling has to be made out of equal parts.
but then you are limited in getting the overall shape, because with bigger curvature your tilings also have to adapt. also its a challenge te be sure that all cirles are touching each other.
but im sure thats a problem thats already solved with grasshopper...
since im myself no pro, i cant help you really much...
look in the def i attached.
you see that if you dont have same tiling it wont function because inner circles of the triangles dont have the same tangent points.
so you have to look out for a more sophisticated tiling for your needs. one way i think would be to have same tilings with even triangles all edges same lengh.
other could be to look out for dense packaging of circles with different radiuses.
Just saw your file and tried to figure out what you intend to have. Here comes a straight way of getting circles on the triangle panels (not onto the surface) -- see the attachment. Once you increase the number of divisions you get smaller triangles, and circles will approach to the surface. Hope this help.
Your example shows that three mid-points moving inward/outward in equal distance and reside on a circle centered on the triangle's center of mass -- that's because they are on an equilateral triangle. For non-eqilateral triangles, just like those in the first set of files you uploaded, it's not the case -- they won't move in equal distance. Instead, they move in equal parametric range, i.e. from 0 to 1/3.
The attached example does not perfectly deal with the edge length, but it at least shows the proper motion as it fold/unfold. Maybe we should try Kangaroo as it is more feasible to deal with such problems.