Electrical conductivity of a two-dimensional two-component medium: discretization and effective medium approximation

Abstract

The effective conductivity of a two-dimensional two-component medium is studied within a discretization scheme and the effective-medium approximation (EMA). A square lattice is superimposed on a random isotropic medium; each cell is assigned the conductivity of its majority component, and the cell is replaced by a star of four resistors. Neighboring cells are connected by an equivalent resistor whose conductivity is $σ_1$, $σ_2$, or the harmonic mean $σ_{12}=2σ_1σ_2/(σ_1+σ_2)$, with probabilities $(1-p)^2$, $p^2$, and $2p(1-p)$, respectively. Applying Foster's theorem and the EMA yields a closed equation for the dimensionless effective conductivity $x$ and, after transformation, a cubic equation whose coefficients depend on $p$ and the conductivity ratio $r$. The EMA predictions are compared with direct calculations on a $200\times200$ resistor network with periodic lateral boundary conditions and Dirichlet conditions on the top and bottom. For uncorrelated disorder, good agreement is found. For correlated disorder, when resistor conductivities are assigned according to the discretization rule from the colors of adjacent nodes, significant deviations appear and the percolation-like inflection point shifts to the right of $1/2$. The results question the validity of the correlated-disorder EMA approach and help explain discrepancies previously reported for random nanowire networks.