finite element method trial 001

Hey guys,

i've been trying to attempt my own 'finite element analysis' by getting the curvature of curves, and rounding the radius of the resultant circles up to certain values (eg, x<50, a=50, 50<x<100, a=70, etc). I've got up to this part. However, i am unable to put back the new 'standardised' circles back into the curve to limit the curvatures in the curve.

(what i'm trying to do have been done more than 10 years ago by FOA in the yokohama project where they limited their curvature to 7 circles!!!)

I've got the standardised circles, but have failed as trying to turn these circles back into the curves. (i tried using closest point to curve, but that didn't seem right, and then tried dividing the circles into points and finding the closest point between the original curves and the circles?)

aaaaaahh...so here's my GH definition and rhino file...PLEASE HELP~~~~~~~HELP HELP~~~~~