Memory formation in dense persistent active matter

Elisabeth Agoritsas, Peter K. MorsearXiv 2024 · DOI · arXiv

Abstract

Protocol-dependent states in structural glasses can encode a disordered, yet retrievable memory. While training such materials is typically done via a global drive, such as external shear, in dense active matter the driving is instead local and spatio-temporally correlated. Here we focus on the impact of such spatial correlation on memory formation. We investigate the mechanical response of a dense amorphous packing of athermal particles, subject to an oscillatory quasistatic driving with a tunable spatial correlation, akin to the instantaneous driving pattern in active matter. We find that the capacity to encode memory can be rendered comparable upon a proper rescaling on the spatial correlation, whereas the efficiency in memory formation increases with motion cooperativity.

Introduction

Amorphous solids, such as glasses and foams, possess the unique ability to encode memories of past deformations within their disordered structures. This phenomenon is typically studied through global driving protocols like oscillatory shear. However, in active matter—ranging from biological tissues to synthetic self-propelled particles—driving is inherently local and often spatially correlated. This study investigates whether the principles of memory formation in passive glasses can be extended to these active systems by examining how spatial correlation in local driving influences memory capacity and efficiency.

Methods: Bridging Global and Local Driving

The researchers utilized 2D computer simulations of athermal bidisperse Hertzian disks ($N = 2048$, packing fraction $\phi = 0.95$) to compare two distinct driving protocols. The first, Athermal Quasistatic Shear (AQS), represents the standard global drive. The second, Athermal Quasistatic Random Displacement (AQRD), introduces local displacement vectors for each particle, drawn from a Gaussian field with a tunable correlation length $\xi$ ranging from 1.0 to 9.5. This $\xi$ parameter allows for the control of "cooperativity" in particle motion.

Comparison of AQS and AQRD protocols and their stress-strain responses.

To compare these protocols, the authors introduced a rescaling factor, $\kappa(\xi)$, derived from the material's apparent elastic modulus. This factor maps local displacements onto an equivalent "effective strain," providing a common framework for analysis.

Results: Statistical Equivalence and Rescaling

A key finding of this work is the statistical equivalence between global shear and local correlated driving. By rescaling stress and strain using the initial ratio $\kappa_0(\xi)$, the complex stress-strain loops observed in AQRD simulations collapse onto the master curves defined by AQS. This collapse holds across various observables, including maximal stress ($\sigma_{max}$), zero-stress strain ($\gamma_0$), and energy dissipation per cycle ($I$).

Rescaling of limit-cycle observables demonstrating statistical equivalence between global and local driving.

The study identifies a critical yielding threshold at an effective strain of approximately 0.04 ($\sqrt{\kappa_0} \tilde{\gamma}_{max} \approx 0.04$). Below this threshold, the system settles into stable limit cycles, effectively "remembering" the driving amplitude. Above this point, plastic rearrangements lead to "amnesia," where the material's state becomes independent of its training history.

Efficiency and Cooperativity

While the capacity for memory is equivalent after rescaling, the efficiency of memory formation depends strongly on the spatial correlation length $\xi$. As $\xi$ increases—meaning particles move more cooperatively—the system reaches stable limit cycles more efficiently, requiring less energy dissipation per cycle. In the pre-yielding regime, energy dissipation follows a universal cubic power law ($I \propto \gamma_{max}^3$), regardless of the driving protocol. Using Mean-Square-Displacement (MSD) between cycles, the authors further confirmed that memory training follows a universal decay curve when viewed through the lens of effective strain.

Conclusion and Future Directions

This research establishes a robust link between global and local driving in amorphous solids, showing that spatial correlation is a key determinant of memory efficiency in active matter. Future work will aim to move beyond the quasistatic limit to incorporate finite driving rates and particle persistence times, more closely mimicking real-world active systems. Additionally, exploring the energy landscape's topology could reveal how to engineer materials capable of storing multiple, localized memories simultaneously.

Summary updated

Abstract

Protocol-dependent states in structural glasses can encode a disordered, yet retrievable memory. While training such materials is typically done via a global drive, such as external shear, in dense active matter the driving is instead local and spatio-temporally correlated. Here we focus on the impact of such spatial correlation on memory formation. We investigate the mechanical response of a dense amorphous packing of athermal particles, subject to an oscillatory quasistatic driving with a tunable spatial correlation, akin to the instantaneous driving pattern in active matter. We find that the capacity to encode memory can be rendered comparable upon a proper rescaling on the spatial correlation, whereas the efficiency in memory formation increases with motion cooperativity.
BibTeX
@misc{agoritsas2024memory,
  author = {Elisabeth Agoritsas and Peter K. Morse},
  title = {{Memory formation in dense persistent active matter}},
  year = {2024},
  eprint = {2403.11701},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2403.11701},
}
Figure