Percolation of random compact diamond-shaped systems on the square lattice

Abstract

We study site percolation on a square lattice with random compact diamond-shaped neighborhoods. Each site $s$ is connected to others within a neighborhood in the shape of a diamond of radius $r_s$, where $r_s$ is uniformly chosen from the set $\{i, i+1, \ldots, m\}$ with $i \leq m$. The model is analyzed for all values of $i = 0, \ldots, 7$ and $m = i, \ldots, 10$, where $\overline{z}(i,m)$ denotes the average number of neighbors per site and $p_c(i,m)$ is the critical percolation threshold. For each fixed $i$, the product $\overline{z}(i,m)\,p_c(i,m)$ is found to converge to a constant as $m \to \infty$. Such behavior is expected when $i=m$ (single diamond sizes), for which the product $\overline{z}(i)\,p_c(i)$ tends toward $2^d\eta_c$, where $\eta_c$ is the continuum percolation threshold for diamond-shaped regions or aligned squares in two dimensions ($d=2$). This case is further examined for $i = 1, \ldots, 10$, and the expected convergence is confirmed. The particular case $i = m$ was first studied numerically by Gouker and Family in 1983. We also study the relation to systems of deposited diamond-shaped objects on a square lattice. For monodisperse diamonds of radius $r$, there is a direct mapping to percolation with a diamond-shaped neighborhood of radius $2r+1$, but when there is a distribution of object sizes, there is no such mapping. We study the case of mixtures of diamonds of radius $r=0$ and $r=1$, and compare it to the continuum percolation of disks of two sizes.