Quadric Flat Torus Theorem

• The Quadric Flat Torus Theorem is a result that shows proper actions of non-cyclic free abelian groups on quadric complexes yield an invariant isometric embedding of the Euclidean square tiling.
• The proof uses combinatorial torus maps, disc diagram techniques, and dual-curve methods to lift local isometries to a global flat embedding.
• A key implication is the rank-bound corollary, which restricts proper free abelian group actions on quadric complexes to Z², contrasting with higher-dimensional CAT(0) settings.

The Quadric Flat Torus Theorem characterizes the algebraic and geometric structure of free abelian group actions on quadric complexes, establishing that proper actions by non-cyclic free abelian groups force the presence of a group-invariant isometric embedding of the standard square tiling of the Euclidean plane. This theorem provides a precise analogue of classical flat torus results for the class of quadric complexes, strengthening the connection between group properties and combinatorial nonpositive curvature in two-dimensional settings (Munro et al., 2024).

1. Quadric Complexes: Foundational Concepts

A quadric complex is a simply connected square complex with a local structure governed by immersion, no folds, and specific replacement properties for small cycles. Formally, a square complex $X$ consists of 2-cells attached along combinatorial circles of length four; it is locally quadric if every square attaches by an immersion, back-to-back squares with the same attaching map are forbidden (no folds), and any immersed boundary cycle conforming to prescribed 2- and 3-cycle shapes can be locally replaced, with these rules ensuring that minimal-area disc diagrams model $\CAT(0)$ square complexes.

A group $G$ acts metrically properly on a quadric complex $X$ if, for every $R < \infty$ and $x \in X$, the set $\{g \in G \mid d(x, gx) \le R\}$ is finite, making properness automatic for free actions on locally finite complexes. The significance of quadric complexes lies both in their combinatorial tractability and their capacity to support dual-curve machinery, providing control over isometries and geodesics.

2. Precise Statement of the Quadric Flat Torus Theorem

The core result of Hoda and Munro (Munro et al., 2024) is the following:

Quadric Flat Torus Theorem.

Let $G$ be a non-cyclic free abelian group acting metrically properly on a quadric complex $X$. Then $G \cong \Z^2$, and $\CAT(0)$0 contains a $\CAT(0)$1-invariant flat, i.e., an isometric embedding $\CAT(0)$2, where $\CAT(0)$3 denotes the standard square tiling of the Euclidean plane by unit $\CAT(0)$4 squares.

This theorem explicitly prohibits the existence of metrically proper actions by higher-rank ($\CAT(0)$5) free abelian groups on quadric complexes, sharply contrasting with the situation in higher-dimensional CAT(0) or cubical settings (Munro et al., 2024, Wise et al., 2015).

3. Proof Outline and Key Lemmas

The proof unfolds in several structured stages:

1. Constructing the Torus Map. A combinatorial map from the torus $\CAT(0)$6 into the quotient $\CAT(0)$7 is realized by any injective homomorphism $\CAT(0)$8, via covering space theory and van Kampen's lemma. Minimizing the number of squares in such a map yields a locally injective representation away from vertices.
2. Lifting to a Flat. Lifting the minimal-area torus map to the universal cover produces $\CAT(0)$9; each $G$0 grid restriction is a minimal-area disc diagram.
3. Disc Diagram and Dual-Curve Machinery. Minimal-area disc diagrams in quadric complexes are $G$1 square complexes (Lemma 2.7), where dual curves are embedded arcs, and geodesic distance is encoded by the number of separating dual curves.
4. Characterization of Flats. A key equivalence (Theorem 3.7) states that a map $G$2 is a flat (isometric embedding) if and only if it is locally isometric, which further holds if and only if each combinatorial star is a minimal-area disc diagram. This local-to-global principle employs the ladder argument (Lemma 2.11) to exclude shortcuts and cycles violating flatness.
5. Thickening to $G$3-Invariance. If $G$4 acts cocompactly, a thickened subcomplex $G$5 is constructed so that $G$6 stabilizes $G$7, and a $G$8-invariant isometric copy of $G$9 is obtained by compressing fibered orbit representatives (Theorem 4.6 and Lemmas 4.1-4.2).

4. Invariant Square Tiling and Construction

The square tiling invariant under $X$0 arises via a three-step scheme:

1. Tor us Lifting: Lift a minimal-area combinatorial torus in $X$1 to $X$2.
2. Flat Verification: Apply the local-to-global flatness criterion to certify $X$3 as an isometric embedding.
3. Orbit Compression: Form the thickening $X$4, use orbit compression lemmas to ensure that, modulo $X$5-action, a unique representative exists above each vertex. The result is that $X$6 preserves a global square tiling isometric to $X$7.

This construction leverages the structure of quadric complexes to globalize local isometries and underpins all applications of the theorem in the context of free abelian group actions (Munro et al., 2024).

5. Rank-Bound Corollaries and Limitations

An immediate corollary is the rank-bound: if $X$8 acts metrically properly on a quadric complex, then necessarily $X$9. Any free action of $R < \infty$0 on a locally finite quadric complex also forces $R < \infty$1. This restriction is not an artefact of combinatorial methodology but arises from the interplay of abelian rank and the available nonpositively curved geometry within quadric complexes.

The result does not generalize to higher-dimensional cubical or more general systolic complexes without further hypotheses. For instance, the cubical flat torus theorem for CAT(0) cube complexes admits virtually abelian groups of arbitrary (bounded by the dimension) rank acting properly cocompactly, with the convex hull decomposing as a product of quasilines (Wise et al., 2015). Hence, the quadric theory marks a sharp dichotomy between the two-dimensional and higher-dimensional settings.

6. Comparative Context and Applications

The Quadric Flat Torus Theorem both complements and extends analogous statements for systolic complexes (e.g., Elsner’s systolic flat torus theorem), benefiting from the stronger disc diagram and dual-curve techniques unique to CAT(0) square complexes. Notably, for $R < \infty$2-dimensional crystallographic groups (Bieberbach groups) acting freely on locally finite quadric complexes, the theorem implies virtual $R < \infty$3 structure and preservation of an isometric tiling, providing a combinatorial proof of the flat torus phenomenon in that genre.

If $R < \infty$4 contains a normal subgroup $R < \infty$5 acting properly, $R < \infty$6 stabilizes a thick flat subcomplex, with crystallographic and Artin-Tits group examples illustrating the versatility of the result. However, the rigidity imposed by the quadric condition sharply limits the ambient group structure compared to more flexible cubical environments (Munro et al., 2024).