Hi Mårten, "bend the other way" is fine and dandy in 2D, but it's not very well defined in 3D. So be careful there.
I think the arc approach would be the best one, and you're probably quite correct and it will be tricky. I once tried to create a Fibonacci spiral, but couldn't get it to work with a variable number of segments. -- David Rutten [email protected] Robert McNeel & Associates On Feb 26, 7:36 pm, Mårten <[email protected]> wrote: > Dear Grasshopper Group. > > I would like to create a "Curve by Curvature", defined either as a > list of numbers or with some kind of Law Curve that could be > interactively changed. Any thoughts on how I could do this? > This page describes Curvature pretty well I > think:http://www.bio.unc.edu/faculty/harris/Courses/biol166/curvature.html > > For example, if it was a series like "1,2,3,4,5,6,7,8,9,10,11..." and > so on, the resulting curve would be a Cornu Spiral. > > I suppose one way of doing it is to connect a number of arcs with > different radius according to the local curvature, but it seems a > little tricky. Is there a better way? > > If the curvature was negative, the curve would bend the other way. > > Many thanks! > > Mårten > ------------------------------------------------------------------http://thegeometryofbending.blogspot.com/
