Numerical investigation of the Brownian q = 2 q=2 Potts Model
Letian Chen, Gunnar Pruessner · 2025
Read Numerical investigation of the Brownian q = 2 q=2 Potts Model by Letian Chen, Gunnar Pruessner (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-04609.txt.
Introduction
Letian Chen and Gunnar Pruessner simulate a flocking-style model where the particles also wander at random. Each particle carries an inner state, like a heading, and copies neighbors within range. The inner state never pushes back on the motion, which keeps the system out of balance by design. The team ran the model both on a grid and in open space.
Earlier work had shown the random drift leaves the transition's static fingerprints unchanged, and this study agrees. Where it parts ways is on clustering. The drift does not break up ordered groups, the authors argue. Instead, wandering acts like an extra attraction that helps groups form. They tie the two versions of the model to classic model families from statistical physics, which explains both behaviors in one frame.
Catalog record
| CREATOR | Letian Chen, Gunnar Pruessner |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 80 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-04609
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Numerical investigation of the Brownian q = 2 q=2 Potts Model Letian Chen, Gunnar Pruessner Abstract In active matter, such as the Vicsek Model of flocking, particles possesses an internal degree of freedom, such as their director, which is subject to interaction with other particles, provided they are within a certain range. In an effort to understand better the interplay between spatial and internal degrees of freedom, we study numerically a variation of the q = 2 q=2 Potts Model on and off the lattice, where particles are additionally subject to Brownian motion. The lack of a feedback of the internal degrees of freedom to the spatial degrees of freedom renders this model generically non-equilibrium. We confirm previous work that showed that the static exponents of the phase transition are unaffected by the diffusion. In contrast to previous work, we show that the formation of ordered clusters is not undermined by diffusion, but should rather be thought of as an effective form of interaction. We demonstrate how our numerical findings can be understood on the basis of the well-established Model A, B and C: Off lattice, the Brownian q = 2 q=2 Potts Model is Model C. On the lattice, it is Model A with an additional (irrelevant, conserved) Model B noise. I Introduction One major controversy surrounding the Vicsek Model [ 38 ] and other models [ 32 , 31 , 10 , 29 , 18 , 9 , 7 ] of flocking remains the questions whether there is an ordered state in two dimensions, the nature of that ordered state and the characteristics of a possible phase transition into that state. What makes the question so relevant and interesting is that no such ordered state can exist in equilibrium either on nor off the lattice, because of the Mermin-Wagner theorem [ 24 ] and extensions to it [ 35 ] , which imply that in equilibrium no long-range order exists in two dimensions between continuous degrees of freedom subject to short-ranged interaction. The order observed consistently in the Vicsek Model [ 38 ] appears to be an oddity, even when allowed by its non-equilibrium nature which protects it from the spell of the Mermin-Wagner theorem. On the other hand, models with discrete symmetries are known to display long-range order in two dimensions. The Ising model and equally the q=2 Potts Model are particularly well understood in this respect [ 42 ] . In the following, we study the q=2 Potts Model with diffusion, the Brownian q=2 Potts Model [ 41 ] , as a minimal variation of the conventional q=2 Potts Model that brings it closer to the Vicsek Model, but without some of the major complications in the Vicsek Model: The diffusion takes the Potts Model out of equilibrium, but we do not implement any self-propulsion and alignment interaction has no effect on the spatial degree of freedom. As Tasaki ( 2020 ) points out, the non-equilibrium nature might bring about order in the Vicsek and related models, which suggests that anything violating equilibrium might be a relevant perturbation. In the following, we add diffusion to a classical model that displays a well-understood order-disorder transition in the absence of diffusion. We will focus on the question to what extent the diffusive motion of the interaction partners poses a relevant perturbation. The local order parameter of the q=2 Potts Model is scalar and does not obey a continuous symmetry so that in the limit of vanishing diffusion, we expect to recover the standard Ising Model, albeit with disorder when the interaction partners are uniformly distributed. Further, as particles are not self-propelled, but purely diffusive, there is no feedback of the internal state, i.e. the spin orientation, to the movement of the particle. As a result the stochastic equations of motion are not expected to carry an advection term as, say, the Toner-Tu equations do [ 37 ] . there is no advective term present. We ask: How is the q=2 Potts Model modified by the presence of such a non-equilibrium diffusive term? The basic features of the Brownian q=2 Potts Models studied in the literature and further detailed below are: 1) Particles have an “internal” Potts degree of freedom subject to Glauber [ 41 , 23 , 4 , 32 , 33 ] or Metropolis [ 40 ] updating as a form of spin-interaction with a local neighbourhood. 2) Particles are subject to diffusion off-lattice or, in the present work nearest-neighbour swapping on the lattice, irrespective of the internal Potts state. Because the swapping and diffusion occurs without reference to the internal Potts state, the models are generically out of equilibrium. Forces are not reciprocal: Particles move freely, even when their interaction makes them align with their neighbourhood. This lack of mechanical consistency would allow for infinite energy extraction whereby particles are subject to an alignment force when they are close, can be separated without performing work, then flipped without work and re-align once brought close together again. The non-equilibrium nature of the setup can also be observed in the following Gedankenexperiment : If there is no mechanical feedback of the spin orientation to the movement of particles, their trajectories approaching each other are indistinguishable from their trajectories departing from each other. However, as they get close, they will tend to align. As a result, their will be plenty of forward trajectories of the form: approach misaligned, align, depart aligned. As particles align as they get close, they typically depart aligned, so that the reverse trajectory is clearly less likely, breaking time-reversal symmetry. In equilibrium, the same construction cannot be made, because particles attract each other as they are aligned and separate (more easily) once misaligned, boosting the frequency with which misaligned particles are seen departing each other. Solon and Tailleur [ 32 , 33 ] have studied a Brownian Ising Model on the lattice. As this model was inspired by the Vicsek Model [ 38 ] , particles, equipped with a spin, self-propel according to their spin orientation. Without any self-propulsion particles are free to diffuse. Spins are subject to an all-to-all interaction when they reside on the same site. Otherwise, they do not interact. Nevertheless, this model displays long-ranged ordering as the diffusion establishes effectively interaction from site to site. Although observed only in the particle density as the control parameter, Solon et al .’s setup displays in two dimensions the standard Ising phase transition. At first sight, this comes as a surprise, because the local interaction is all-to-all and thus inspired by mean-field models. What is more striking is the rôle of the diffusion. On the one hand, it mediates the interaction between nearest neighbouring sites, as particles with aligned spins “invade” neighbouring sites and “win them over”. On the other hand, the diffusion also results in a loss of local magnetisation as particles leave their aligned neighbourhoods. This mechanism is key in the setup studied by Woo, Rieger and Noh [ 41 ] who studied a q=2,\dots 8 Potts Model off the lattice with particles subject to random motion that amounts essentially to diffusion. Again, there is no spatial force acting on the particles as their motion is unaffected by the state of mutual