As per my previous post, several nasty bugs were discovered and [easily] 
fixed.

This now works pretty well...

One issue that *isn't* fixed yet is that under certain conditions, the 
knot can actually "unknot itself". This despite some fairly complex code 
specifically to detect and prevent such occurances. Hmm...

In case it's unclear...
  * Simple bead-and-stick model. (Only the beads are manipulated.)
  * Each bead is subject to two forces:
    * The string applies a linear constricting force, pulling 
consecutive beads toward each other.
    * Every bead is repelled (inverse square) by every other bead.
  * In this universe, force = velocity. (i.e., no inertia, no momentum.)

Also, before moving each bead, the code checks whether the triangle 
formed by this movement is intersected by any other line segment. If it 
is, the move isn't performed. However - for reasons yet to be determined 
- this does NOT prevent knots unrevalling. More research needed...

Andrew.

Post a reply to this message

Attachments:
Download 'knots5-post.m1v.mpg' (393 KB)