Inner diagonals of convex polytopes (Q1296758)

For \(1\leqslant i\leqslant d,\) an \(i\)-diagonal of a \(d\)-polytope \(P\) is a segment \([x,y]\) whose ends are vertices of \(P\) and whose carrier in \(P\) is of dimension \(i.\) The number of \(i\)-diagonals is denoted by \(\delta_i(P).\) Thus \(\delta_1(P)\) is the number of edges and \(\delta_d(P)\) is the number of inner diagonals. The focus here is on \(\delta_d(P).\) The paper's main results are as follows: (a) Within the class of \(d\)-polytopes that have a given number \(v\) of vertices, the maximum of \(\delta_d(P)\) is \(\binom v2-dv+\binom{d+1}{2}.\) For \(d\geqslant 4\) this maximum is attained by the stacked polytopes and only by them. (b) Within the class of \(d\)-polytopes that have a given number \(f\geqslant 2d\) of facets, the maximum of \(\delta_d(P)\) is attained by certain simple \(d\)-polytopes. (c) For each simplicial 3-polytope with \(f\) facets \(\delta_3(P)=(f^2-6f+8)/8,\) while for simple 3-polytopes \(\delta_3(P)\) ranges from \(f^2-9f+20\) to (when \(f\geqslant 14\)) \(2f^2-21f+64.\)