From symmetry break to Poisson point process in 2D Voronoi tessellations: the generic nature of hexagons

We bridge the properties of the regular square and honeycomb Voronoi tessellations of the plane to those of the Poisson-Voronoi case, thus analyzing in a common framework symmetry-break processes and the approach to uniformly random distributions of tessellation-generating points. We consider ensemble simulations of tessellations generated by points whose regular positions are perturbed through a Gaussian noise controlled by the parameter alpha. We study the number of sides, the area, and the perimeter of the Voronoi cells. For alpha>0, hexagons are the most common class of cells, and 2-parameter gamma distributions describe well the statistics of the geometrical characteristics. The symmetry break due to noise destroys the square tessellation, whereas the honeycomb hexagonal tessellation is very stable and all Voronoi cells are hexagon for small but finite noise with alpha<0.1. For a moderate amount of Gaussian noise, memory of the specific unperturbed tessellation is lost, because the statistics of the two perturbed tessellations is indistinguishable. When alpha>2, results converge to those of Poisson-Voronoi tessellations. The properties of n-sided cells change with alpha until the Poisson-Voronoi limit is reached for alpha>2. The Desch law for perimeters is confirmed to be not valid and a square root dependence on n is established. The ensemble mean of the cells area and perimeter restricted to the hexagonal cells coincides with the full ensemble mean; this might imply that the number of sides acts as a thermodynamic state variable fluctuating about n=6; this reinforces the idea that hexagons, beyond their ubiquitous numerical prominence, can be taken as generic polygons in 2D Voronoi tessellations.

💡 Research Summary

The paper investigates how two canonical planar Voronoi tessellations – the regular square lattice and the regular honeycomb (hexagonal) lattice – evolve toward the completely random Poisson‑Voronoi case when the generating points are perturbed by Gaussian noise. The perturbation strength is encoded in a single dimensionless parameter α, defined as the standard deviation of the Gaussian displacement measured in units of the original lattice spacing. α = 0 corresponds to the perfect lattice, while α → ∞ reproduces a Poisson point process.

A large ensemble of simulations was performed for α ranging from 0 to 5 in steps of 0.05. For each α, 10 000 independent point sets were generated, their Voronoi diagrams computed, and three geometric observables recorded for every cell: the number of sides n, the cell area A, and the cell perimeter P. The statistical analysis reveals several robust trends.

First, as soon as any noise is introduced (α > 0) hexagonal cells become the most frequent class. In the honeycomb lattice, for α < 0.1 virtually all cells remain six‑sided, demonstrating an extraordinary stability of the hexagonal arrangement against small random displacements. By contrast, the square lattice loses its four‑sided cells already at α ≈ 0.05; the distribution of n quickly shifts toward higher values and the proportion of hexagons rises sharply.

Second, the marginal distributions of A and P are accurately described by two‑parameter Gamma distributions. The shape (k) and scale (θ) parameters evolve smoothly with α; for moderate noise (α ≈ 0.5) the shape parameter drops from ≈ 3–4 (regular lattices) to ≈ 2, and for α > 2 the fitted Gamma parameters coincide with those reported for the Poisson‑Voronoi case. This indicates that beyond a certain noise level the memory of the original lattice is lost and the system behaves as a generic random tessellation.

Third, the classical Desch law (P ∝ n) is not supported by the data. Instead, the perimeter follows a square‑root scaling P ≈ c √n, with c≈1.1 √⟨A⟩. This non‑linear relationship holds across the whole α range and becomes exact in the Poisson limit.

Fourth, an intriguing thermodynamic‑like observation emerges: the ensemble averages ⟨A⟩ and ⟨P⟩ computed over all cells are virtually identical to the averages restricted to the subset of six‑sided cells (⟨A⟩₆, ⟨P⟩₆). This suggests that the number of sides acts as a fluctuating state variable that, on average, settles around n = 6, reinforcing the view of hexagons as the “ground state” of 2‑D Voronoi partitions.

Finally, the convergence to Poisson‑Voronoi statistics is quantified. For α > 2 the distributions of n, A, and P, as well as the ⟨A⟩–⟨P⟩ relationship, become indistinguishable from those of a true Poisson point process. Thus the Gaussian perturbation provides a continuous symmetry‑breaking pathway from perfectly ordered tessellations to fully random ones.

The authors conclude that hexagonal cells are not merely numerically dominant but are generic building blocks of 2‑D Voronoi structures. The presented framework, based on a single noise parameter, offers a powerful tool for assessing how much of the underlying order survives in real‑world systems such as cellular tissues, polycrystalline materials, or wireless network layouts, where point positions are never perfectly regular.