Dynamic Processes in Disordered Media

Anders Malthe‐Sørenssen, Anders Malthe-SorenssenLecture Notes in Physics 2024 · DOI · OpenAlex

Abstract In this chapter we start to address dynamical processes that generate percolation-like disordered structures. We will address the scaling behavior of diffusion fronts for diffusion in disordered media, and develop a theory for the front width. We will address the slow displacement of fluids in porous media with the invasion-percolation model. For this model, we will map the model onto a percolation system and demonstrate how the model changes with the introduction of an additional aspects, gravity.

Introduction

While traditional percolation theory often examines static disorder—such as the connectivity of a fixed porous medium—many physical phenomena involve dynamical processes where the disordered structure evolves over time. This paper explores how spreading particles and invading fluids generate complex, fractal structures that can be mapped onto percolation theory. By integrating analytical scaling laws with computational simulations, we demonstrate how external forces like gravity reshape these fronts, transforming stochastic growth into predictable physical patterns.

Diffusion Fronts and Scaling Behavior

When particles diffuse from a source into a medium, they do not form a smooth concentration gradient. Instead, they create a jagged, irregular boundary known as a diffusion front. A primary challenge is predicting the width and scaling of this front.

Figure 1: (a) Simulation of a diffusion front showing the average position $x_c$ and the local width $\xi$. (b) A magnified view of the front's fractal structure.

Using the diffusion equation and Taylor expansions, we derive a self-consistency equation for the front width ($w$). This analysis links the geometry of the front to the universal correlation length exponent ($\nu$) from percolation theory. A key insight is that the "edge" of a diffusing cloud behaves like a percolation system at its critical threshold ($p_c$). Interestingly, as the front advances, its relative width decreases, causing the interface to appear sharper over time. The width scales as $w \propto x_c^{\nu/(1+\nu)}$.

Invasion Percolation

Invasion percolation describes the slow displacement of one fluid by another in a porous medium, a process central to oil recovery and groundwater remediation. This process is a prime example of self-organized criticality; the invading fluid naturally "finds" the path of least resistance, creating a fractal structure similar to a spanning cluster without the need for manual tuning to a critical probability.

Figure 2: Schematic of invasion percolation on a lattice, where the invading fluid (blue) occupies pores based on pressure thresholds.

To model this, we map the invasion process onto a standard percolation problem using a random lattice of pressure thresholds. In 2D, the "trapping" of the displaced fluid reduces the fractal dimension to $D \approx 1.82$, whereas in 3D, the dimension remains $D \approx 2.5$.

The Role of Gravity

The introduction of gravity, characterized by the Bond number (the ratio of gravitational to capillary forces), fundamentally alters the invasion front. Gravity can have a dual nature:

  • Stabilization: When a denser fluid is injected from below, gravity suppresses the formation of large "fingers," resulting in a more compact, flattened interface. In such systems, the front width scales with the gravity gradient $g$ as $a \propto g^{-\nu/(1+\nu)}$.
  • Destabilization: Conversely, injecting a less dense fluid at the bottom of a denser one causes the front to collapse into a one-dimensional finger, a phenomenon known as secondary migration in petroleum geology.

Methodology and Computational Implementation

The study utilizes Monte Carlo simulations on synthetic datasets, typically $400 \times 400$ random lattices. Each site is assigned a threshold value from a uniform distribution. We employ the scipy.ndimage library to efficiently identify connected clusters within these digital laboratories. While the provided Python implementations are educational, they successfully validate the theoretical scaling laws against simulated results.

Conclusion

This work demonstrates that dynamical processes in disordered media are not merely chaotic but follow rigorous scaling laws derived from percolation theory. Future research should address the limitations of the current models, such as the assumption of uncorrelated disorder in the medium and the neglect of viscous effects or injection velocities, which play significant roles in industrial-scale fluid dynamics.

Summary updated

Abstract

Abstract In this chapter we start to address dynamical processes that generate percolation-like disordered structures. We will address the scaling behavior of diffusion fronts for diffusion in disordered media, and develop a theory for the front width. We will address the slow displacement of fluids in porous media with the invasion-percolation model. For this model, we will map the model onto a percolation system and demonstrate how the model changes with the introduction of an additional aspects, gravity.
BibTeX
@inproceedings{malthesrenssen2024dynamic,
  author = {Anders Malthe‐Sørenssen and Anders Malthe-Sorenssen},
  title = {{Dynamic Processes in Disordered Media}},
  booktitle = {Lecture Notes in Physics},
  year = {2024},
  doi = {10.1007/978-3-031-59900-2\_12},
}
Percolation Theory Using PythonLecture Notes in Physics · 2024
Diffusion in Disordered MediaLecture Notes in Physics · 2024
Flow in Disordered MediaLecture Notes in Physics · 2024
Geometry of ClustersLecture Notes in Physics · 2024
Introduction to PercolationLecture Notes in Physics · 2024
Finite-Dimensional PercolationLecture Notes in Physics · 2024
Figure