Geometrical characteristics of the pore space in a random packing of equal spheres

Tetrahedral tessellation is applied to an existing, computer-generated, random packing of nearly equal spheres.The vertices of each tetrahedron are the centers of four neighboring spheres.The objective is to identify the tetrahedral pores and then determine the geometrical characteristics of the pore space. The tessellation procedure begins with a seed tetrahedron. Then, in a crystal growth-like manner, tetrahedra are added one at a time to the seed until a cluster of non-overlapping, space-filling tetrahedra is formed. Each tetrahedron has a pore chamber and four constrictions, one on each of its four triangular faces. Statistical analysis allows the determination of pore size distribution, constriction size distribution as well as various correlations between (among) pores and (among) constrictions. It is shown that pores linked by a common constriction are rather close in size, while constrictions of the same pore tend to have more different sizes.