Subscribe
Subscribe
MY AMERICAN SCIENTIST
LOG IN! REGISTER!
SEARCH
 
Logo IMG

COMPUTING SCIENCE

The Science of Sticky Spheres

On the strange attraction of spheres that like to stick together

Brian Hayes

At Sixes and Sevens

2012-11HayesFCa.gifClick to Enlarge ImageUp to this point, each value of n has had a unique cluster that maximizes Cn. Furthermore, in each case the best-connected cluster with n+1 spheres can be assembled incrementally by sticking a new sphere somewhere on the surface of the max(Cn) cluster. These properties come to an end at n=6. With six spheres, two cluster shapes both yield the same maximum contact number, C6=12. (Note that a hypothetical six-sphere clique would have 15 contacts.) One of the max(C6) clusters is built incrementally from the five-sphere triangular dipyramid. But the other max(C6) cluster is a “new seed”—a structure that cannot be created simply by gluing a sphere to the surface of a smaller optimum cluster. The new seed is the octahedron (which might also be described as a square dipyramid).

2012-11HayesFDa.gifClick to Enlarge ImageBeyond n=6, the problem of finding all the maximum-contact clusters becomes more daunting. For n=7, the incremental approach of adding another sphere to the surface of an n=6 cluster yields four solutions that have 15 contact points. Three of these C7=15 clusters consist of four tetrahedra glued together face-to-face in various ways. The remaining product of incremental construction consists of an octahedron with a tetrahedron erected on one face. (One of the seven-sphere solutions has both left-handed and right-handed forms, but the convention is to count these “chiral” pairs as variants of a single cluster, not as separate structures.)

Finding this particular set of structures is not especially difficult. If you spend some time playing with Geomags or some other three-dimensional modeling device, you are likely to stumble upon them. But having identified these four clusters with C7=15, how do you know there aren’t more? And how do you prove that no seven-sphere cluster has 16 or more contacts?

2012-11HayesFEa.gifClick to Enlarge ImageAs it turns out, 15 is indeed the maximum contact number for seven spheres, but there is another C7=15 cluster. It is a new seed, called a pentagonal dipyramid. With its fivefold symmetry, it has no structural motifs in common with any of the smaller clusters. The novelty of this object again raises the question: How can we ever be sure there aren’t still more arrangements waiting to be discovered?

A successful program for answering such questions was initiated about five years ago by Natalie Arkus, who was then a graduate student at Harvard University. (She is now at Rockefeller University.) In a series of papers written with her Harvard colleagues Michael P. Brenner and Vinothan N. Manoharan, she enumerated all the max(Cn) configurations for n=7 through n=10. The results were later extended to n=11 by Robert S. Hoy, Jared Harwayne-Gidansky and Corey S. O’Hern of Yale University. (Hoy is now on the faculty of the University of South Florida.) All of the results I describe here come from the work of these two groups.




comments powered by Disqus
 

EMAIL TO A FRIEND :

Subscribe to American Scientist