Multiscale Energy Spreading in Hard-Particle Chains
Arkady Pikovsky · 2025
Read Multiscale Energy Spreading in Hard-Particle Chains by Arkady Pikovsky (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-2504-03511.txt.
Introduction
Pikovsky studied how energy spreads in a chain of particles in this physics paper from twenty twenty-five. The particles sit in a row and interact through an endless-well rule, like hard balls that bounce. At the start, only a small group moves while the rest sit still. The question is how the active zone grows over time.
Pikovsky tracks the growth with averages and entropy measures of the energy spread. In two special cases the spreading follows a single clean scale. One is the plain hard-particle gas. The other fixes the spacing at half the well width. In all other cases the spread mixes several scales at once, a pattern called multiscale anomalous diffusion. The result sorts when simple laws of spreading apply.
Catalog record
| CREATOR | Arkady Pikovsky |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 19 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-03511
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Multiscale Energy Spreading in Hard-Particle Chains Arkady Pikovsky Abstract We consider a one-dimensional array of particles interacting via an infinite well potential. We explore the properties of energy spreading from an initial state where only a group of particles has non-zero velocities while others are resting. We characterize anomalous diffusion of the active domain via moments and entropies of the energy distribution. Only in the special cases of a single-well potential (hard-particle gas) and of the distance between the particles being half of the potential width does the diffusion have a single scale; otherwise, a multiscale anomalous diffusion is observed. Introduction. One-dimensional classical nonlinear lattices posses quite peculiar transport properties. On the one hand, many such systems demonstrate anomalous heat transport, in which the finite-size heat conductivity diverges as a power of the lattice length Lepri 2016 ; Livi 2023 ; Benenti et al. 2023 . In a typical setup, one attaches a lattice to two heat baths with different temperatures and measures the energy exchanges with them. Because according to the Green-Kubo theory, the conductivity can be expressed in terms of the equilibrium fluctuations of the heat current, anomalous transport manifests itself in a non-integrable power-law decay of the correlation function of the energy current Lepri et al. 1998 . Another manifestation of anomalous heat transport is a superdiffusive spreading of local energy perturbations on top of the equilibrium state with finite temperature (finite energy density) Cipriani et al. 2005 . Theoretical and numerical findings about anomalous heat transport have been confirmed in experiments Chang et al. 2008 ; Upadhyaya and Aksamija 2016 ; Crnjar et al. 2018 ; Yang et al. 2021 . Another setup where one-dimensional lattices demonstrate nontrivial transport properties is the spreading of initially localized perturbations on top of a zero-temperature state. Due to the effect of Anderson localization Anderson 2010 , in a linear lattice, already a small amount of disorder leads to exponential localization of eigenmodes, which blocks spreading at large times. However, the nonlinearity of the lattice may result in chaos, which destroys localization and leads to a subdiffusive spreading. One popular example is a disordered nonlinear Schroedinger lattice Pikovsky and Shepelyansky 2008 ; Flach et al. 2009 ; Fishman et al. 2012 ; Flach 2015 . However, even for this widely studied model it still not clear whether spreading persist at very large times, as chaos may degenerate into localized quasiperiodic modes Pikovsky and Fishman 2011 ; Aubry 2023 . Spreading in a nonlinear Schroedinger lattice has been observed in optical experiments Lahini et al. 2009 . While in the context of the Schroedinger lattice, one follows the spreading of the wave packet, for nonlinear lattices of Klein-Gordon or Fermi-Pasta-Ulam-Tsingoi type, one studies the spreading of energy from a localized perturbation on top of a zero-temperature state Mulansky et al. 2011 ; Mulansky and Pikovsky 2013 ; Martínez et al. 2016 ; Ngapasare et al. 2019 ; Pikovsky 2020 . While in many numerical experiments a subdiffusive spreading is observed, asymptotic regimes at large times remain elusive. For the experimental realization of such a spreading in a disordered granular chain, see Ref. Kim et al. 2018 . In this Letter, we explore energy spreading on top of a zero-temperature state for hard-particle models previously studied in the context of anomalous heat transport. These models share basic properties like conservation of momentum with lattices with smooth coupling potentials, but allow for an efficient numerical implementation. We will demonstrate that while in some cases anomalous superdiffusion or subdiffusion, characterized just by one exponent (mono-scale diffusion), is observed, generally a hard-particle chain demonstrates multi-scaled diffusion, where differently defined “lengths” of the spreading domain grow with different exponents. Model formulation. Our basic model is a one-dimensional chain of particles interacting via an infinite well potential Delfini et al. 2007 ; Politi 2011 ; Mejía-Monasterio et al. 2019 . This hard-particle chain (HPC) is defined via the Hamilton function \displaystyle H \displaystyle=\sum_{i}\frac{p_{i}^{2}}{2m_{i}}+U(x_{i+1}-x_{i})\;, (1) \displaystyle U(\Delta x) \displaystyle=\begin{cases}\infty&\Delta x<0\text{ and }\Delta x>a^{-1}\;,\\ 0&0\leq\Delta x\leq a^{-1}\;.\end{cases} The essential dimensionless parameter of the problem is 0\leq a<1 , the ratio of the mean distance between the particles and the potential width. We set the initial distance between the particles to 1 ; thus the parameter a enters the definition of the potential (the width of the well) in ( 1 ). To ensure the dynamics is non-integrable, we set random masses according to a uniform distribution 0.5\leq m_{i}\leq 1.5 . In the case a=0 , the double-well potential becomes a hard-core potential, and the model reduces to a famous one-dimensional hard-particle gas (HPG) model Casati 1986 ; Grassberger et al. 2002 ; Lepri et al. 2020 , where only elastic collisions at x_{i+1}=x_{i} happen. Remarkably, the HPC model is symmetric with respect to the change of the parameter a\to 1-a Mejía-Monasterio et al. 2019 . The value a=1/2 is special because here, the two types of the collisions at x_{i+1}=x_{i} and at x_{i+1}=x_{i}+a^{-1} become in equilibrium equally probably, and the pressure vanishes. Therefore, we restrict our attention below to the interval 0\leq a\leq 1/2 . The HPC ( 1 ) conserves energy and momentum, and in the heat transfer setup has demonstrated anomalous heat conductivity Delfini et al. 2007 ; Politi 2011 ; Mejía-Monasterio et al. 2019 . Below in this letter we study energy-spreading properties at zero temperature. We start with a configuration where positions are equidistant x_{i}(0)=i , and the momenta p_{i} of the particles are randomly set in a small domain (ten sites) around i=0 , other particles are at rest p_{i}(0)=0 . Because of the possibility of time rescaling, without loss of generality, the total energy can be set to 1 . Furthermore, we set the total initial momentum to zero. Figure 1: Examples of the energy spreading process: trajectories x_{i}(t) of particles vs time for three values of parameter a : panel (a) a=0 ; panel (b) a=0.25 ; panel (c) a=0.5 . Energy spreading and its characterization. The phenomenology of the dynamics is simple (Fig. 1 ): more and more particles are involved in the nontrivial dynamics via collisions, and a spreading “active domain” consisting of particles having non-zero energy is formed. At each time instant, only a finite number of particles belong to the active domain, and this number L(t) is a natural measure of the domain width. However, this number is only one of the possible definitions of the domain width. Because we normalize the total energy to one, and the energies of the particles are non-negative, the set of local energies E_{i}=p_{i}^{2}/(2m_{i}) can be interpreted as a probability distribution. Accordingly, we can characterize the distribution width using this distribution’s moments or entropies. In the former approach, we define the “center of energy” \langle x\rangle , the moments M_{p} , and the corresponding moment-based sizes of the domain \ell_{p} according to relations \langle x\rangle=\sum_{i}x_{i}E_{i},\;M_{p}=\sum_{i}|x_{i}-\langle x\rangle|^{p}E_{i},\;\ell_{p}=(M_{p})^{1/p}\;. (2) Here, the indices p>0 need not be integers. Another way to characterize distributions is to calculate their Renyi entropies I_{q} depending on index q\geq 0 , and to use them to define the widths \mathcal{L}_{q} : I_{q}=\frac{\log\sum_{i}E_{i}^{q}}{1-q},\quad\mathcal{L}_{q}=\exp[I_{q}]\;. (3) Note that for q=0 , the entropy is the logarithm of the support of the distribution I_{0}=\log