Non-linear and non-local transport processes in heterogeneous media: from long-range correlated percolation to fracture and materials breakdown

Abstract

We review and discuss recent progress in modelling non-linear and non-local transport processes in heterogeneous media. The non-locality that we consider is caused by long-range correlations that either exist in the morphology of the media, or are caused by the transport processes themselves. The interplay between the non-linearity and non-locality is discussed in depth with the aim of establishing that, often non-linearity and non-locality are “two sides of the same coin”, such that one may have no meaning without the presence of the other one. First, we discuss linear and scalar, but non-local transport processes and, in particular, consider those in percolation systems with long-range correlations. It appears that there are significant differences between percolative transport processes in which the long-range correlations (or the covariance function) decrease with the distance r between two points, and those in which they increase as r does. Application of this problem to flow and transport in geological formations is discussed. We then consider linear vector percolation, one type of which, the rigidity percolation, provides an example of a non-local vector transport in heterogeneous media. Applications of vector percolation to modelling elastic properties of glasses, composite solids and rock, mechanical and viscoelastic properties of polymers, and vibrations and dynamical properties of heterogeneous materials are discussed. Non-linear and non-local scalar transport processes are discussed next, including various breakdown phenomena in disordered composites, power-law transport, piecewise linear transport characterized by a threshold, and non-linear processes that arise as a result of imposing a large external potential gradient on a heterogeneous medium. Their relevance to flow of non-Newtonian fluids in porous media, to electrical currents and dielectric breakdown in composite solids and doped polycrystalline semiconductors, and several other problems is then discussed. Finally, we discuss non-linear and non-local vector transport, the most important example of which is mechanical fracture of disordered solids. We discuss exactly solvable models of fracture, quasi-static and dynamic lattice models, molecular dynamics simulations, and continuum formulation of the problem. In all cases, the predictions of the models are compared with the relevant experimental data.