6
$\begingroup$

A spherical polytope is the intersection of some closed hemispheres which is non-empty and does not contain a pair of antipodal points. A spherical complex is a tiling of the whole (d−1)-dimensional sphere by spherical polytopes. Equivalently, it is the complex obtained by intersecting a complete polyhedral fan with the a (d−1)-dimensional sphere centered at the vertex of the fan.

We know that every combinatorial information of a spherical complex can be realized as a convex polytope in dimension 3 by considering the so-called canonical polytope. The proof from Wikipedia seems to depend on the circle packing theorem and the midsphere theorem.

So my question is:

Is any spherical polytope of dimension $n$ isomorphic to a convex polytope as abstract polytope for $n > 3$?

According to This question the midsphere theorem is still unknown for higher dimension, so I suspect there is a counterexample. Any idea or reference is appreciated.

EDIT:

(1) The definition of spherical polytope is edited several times (see the discussion on comment)

(2) I just saw this wiki where it is stated that "convex polytopes in p-space are equivalent to tilings of the (p−1)-sphere." which seems to give a positive answer of my question, but no reference is provided there.

$\endgroup$
5
  • $\begingroup$ please give a link to your definition of spherical polytope. $\endgroup$ Feb 5, 2022 at 8:12
  • $\begingroup$ @DimaPasechnik I don't have the reference for higher dimension, but my intention is to generalize the definition for dimension 3 defined in Wikipedia as tiling of sphere. I edited the OP for a more formal definition. It seems the same question is discussed in this question. $\endgroup$
    – hyyyyy
    Feb 5, 2022 at 9:18
  • $\begingroup$ your definition of a spherical polytope seems to be wrong, at least wrt. the stated result: as a counter example consider the great dodecahedron {5, 5/2}: it well is contained within a sphere, the origin of which is contained in its interior, but you cannot provide a convex sphere packing with 5 pentagons per vertex. $\endgroup$ Feb 5, 2022 at 17:35
  • $\begingroup$ @Dr.RichardKlitzing You are right. I am new to the theory of abstract polytopes and when I write the term "polytope" I have some additional conditions in my mind (like two faces intersect on their common face). I edited the question so that the terminology is same as the other question cited in my previous comment. $\endgroup$
    – hyyyyy
    Feb 6, 2022 at 1:04
  • $\begingroup$ I am doubtful that the answer is yes, but I have nothing to back this up for now. This question seems very related, but I think the current answer is not very helpful for you. It might hint to some relevant terminology though. $\endgroup$
    – M. Winter
    Feb 19, 2022 at 22:21

1 Answer 1

7
$\begingroup$

Another way to phrase your question is "whether every complete fan is (combinatorially equivalent to) the face fan (or the normal fan) of a convex polytopes". The answer is No in dimension $n\ge 4$ and I will provide an example derived from the following non-polytopal 4-diagram:

This picture is taken from Example 5.10 in Günter Ziegler's book "Lectures on Polytopes". The book explains why this diagram is not the Schlegel diagram of any 4-dimensional polytope. It is important to note that the "outer cell" of the diagram is a simplex.

Now, consider this diagram $D$ embedded into a 3-dimensional affine subspace of $\Bbb R^4$ that does not pass though the origin. Let $|D|$ be the simplex that forms the "outer cell" of $D$. Let $x\in\Bbb R^4$ be a point for which the convex hull $\Delta:=\mathrm{conv}(|D|\cup\{x\})$ contains the origin in its interior. This convex-hull is a 4-simplex and $|D|$ is one of its facets. Take the face fan of the simplex $\Delta$ and subdivide the cone over the facet $|D|$ into the cones over the cells of $D$. This is a complete fan in which no cone contains antipodal points. You can intersect it with a sphere to obtain a spherical complex.

Suppose now that $P$ is a polytope combinatorially equivalent to this complex. Let $y\in P$ be the vertex that corresponds to the vertex $x$ of the complex. Its four neighbors in $P$ span a simplex $\Delta'$ (that corresponds to $|D|$ in the fan above). But deleting $y$ from $P$ (i.e. taking the convex hull of all vertices of $P$ except for $y$) leaves a polytope for which $\Delta'$ is a facet (here we use that $\Delta'$ resp. $|D|$ is a simplex) and whose Schlegel diagram based at $\Delta'$ is exactly $D$. This is a contradiction.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.