English

Self-dual quasiperiodic percolation

Statistical Mechanics 2023-03-08 v3 Disordered Systems and Neural Networks

Abstract

How does the percolation transition behave in the absence of quenched randomness? To address this question, we study two nonrandom self-dual quasiperiodic models of square-lattice bond percolation. In both models, the critical point has emergent discrete scale invariance, but none of the additional emergent conformal symmetry of critical random percolation. From the discrete sequences of critical clusters, we find fractal dimensions of Df=1.911943(1)D_f=1.911943(1) and Df=1.707234(40)D_f=1.707234(40) for the two models, significantly different from Df=91/48=1.89583...D_f=91/48=1.89583... of random percolation. The critical exponents ν\nu, determined through a numerical study of cluster sizes and wrapping probabilities on a torus, are also well below the ν=4/3\nu=4/3 of random percolation. While these new models do not appear to belong to a universality class, they demonstrate how the removal of randomness can fundamentally change the critical behavior.

Keywords

Cite

@article{arxiv.2206.11290,
  title  = {Self-dual quasiperiodic percolation},
  author = {Grace M. Sommers and Michael J. Gullans and David A. Huse},
  journal= {arXiv preprint arXiv:2206.11290},
  year   = {2023}
}

Comments

17 pages (17 figures) + 3 appendices (6 figures, 3 tables). v3 revisions: integrated supplement into main text; added analysis of a second quasiperiodic model, determination of fractal dimension of hulls, and investigation of universality; revised estimates of $D_f$ and $\nu$ for original model using nominally infinite methods; removed discussion of energy correlator and red bonds

R2 v1 2026-06-24T12:00:41.201Z