A finite volume method for continuum limit equations of nonlocally interacting active chiral particles

Nikita Kruk, José A. Carrillo, Heinz KoepplJournal of Computational Physics 2020 · DOI

Introduction

Active matter systems—comprised of individual agents like bacteria, birds, or synthetic micro-robots—are defined by their ability to consume energy to generate autonomous motion. While individual agent-based models provide granular insights, simulating millions of interacting agents is computationally prohibitive. To address this, physicists employ "continuum limit" equations, which use Partial Differential Equations (PDEs) to describe the evolution of the swarm's density as a whole.

This paper focuses on a particularly complex subset: chiral active particles. Unlike standard active matter, these particles exhibit spontaneous rotation as they interact. The resulting PDEs are high-dimensional (involving both spatial position and orientation), nonlinear, and nonlocal, meaning the state at any given point depends on the configuration of the swarm across a finite interaction radius. The authors present a robust Finite Volume Method (FVM) designed to solve these equations efficiently while preserving the physical integrity of the solution.

The Numerical Framework

Solving the continuum equations for chiral active matter requires balancing computational efficiency with numerical stability. The authors' FVM framework introduces several key technical innovations to handle the 3+1 dimensional problem (2D space, 1D orientation, and time):

1. Positivity Preservation

Since the PDE describes a probability density function, the solution must remain non-negative to be physically meaningful. The authors implement a piecewise linear reconstruction combined with a "minmod" slope limiter. This approach prevents numerical oscillations that could lead to negative density values, ensuring the scheme remains stable even in regions with sharp gradients.

2. Dimensionality Splitting

To manage the high dimensionality, the framework employs Strang splitting. This technique separates the spatial movement (transport) from the angular rotation (alignment). By decoupling these components, the authors can apply specialized solvers to each part without sacrificing second-order accuracy in time.

3. Efficient Nonlocal Interactions

The nonlocal nature of the interactions is mathematically represented as convolutions. A naive calculation of these terms would scale at $O(N^2)$, making large-scale simulations impossible. By utilizing the Discrete Fourier Transform (DFT), the authors reduce this complexity to $O(N \log N)$, allowing for the simulation of complex, large-scale collective behaviors.

4. Gradient Flow Structure

For the angular component of the equations, the authors exploit the underlying "gradient flow" structure of the Kuramoto-Sakaguchi model. This ensures that the numerical scheme respects the fundamental physical properties of the system's alignment dynamics.

Key Insights and Collective Phenomena

The numerical framework allowed the authors to explore the rich phase space of chiral active matter, leading to several significant findings:

  • Spontaneous Symmetry Breaking: The model demonstrates that collective rotation can emerge even if individual particles are not inherently "hard-coded" to rotate. Instead, rotation arises spontaneously due to the phase lag in alignment interactions between particles.
  • The "Chimera State": A highlight of the study is the recovery of localized chimera states. In these configurations, a high-density, ordered "cloud" of particles rotates and migrates through a background of completely disordered, chaotically moving particles. This coexistence of order and disorder is a hallmark of complex dynamical systems.
  • Phase Transitions and Bistability: The researchers mapped a detailed phase diagram identifying three distinct regimes: Spatially Homogeneous Disordered Motion (SHDM), Spatially Homogeneous Ordered Motion (SHOM), and Spatially Inhomogeneous Motion (SNM). They also discovered regions of "bistability," where the system's final state (ordered motion vs. a chimera state) depends entirely on its initial conditions.

Validation and Performance

The proposed FVM was rigorously validated against analytical solutions and synthetic benchmarks. The results confirmed second-order accuracy across $L^1$, $L^2$, and $L^\infty$ error norms for both stationary distributions (such as the von Mises distribution) and traveling wave profiles. This level of precision ensures that the complex patterns observed—such as traveling bands and vortices—are genuine physical phenomena rather than numerical artifacts.

Conclusion

This work provides a powerful new tool for the study of active matter. By developing a stable, positivity-preserving FVM that efficiently handles nonlocal interactions, the authors have opened the door to more detailed investigations of biological and synthetic swarms. While the current model focuses on alignment-based interactions, the framework is extensible. Future research may incorporate excluded volume effects or 3D spatial environments, further bridging the gap between theoretical models and the complex reality of active systems in nature.

Summary updated

BibTeX
@article{kruk2020finite,
  author = {Nikita Kruk and José A. Carrillo and Heinz Koeppl},
  title = {{A finite volume method for continuum limit equations of nonlocally interacting active chiral particles}},
  journal = {Journal of Computational Physics},
  year = {2020},
  doi = {10.1016/j.jcp.2021.110275},
}
Figure