Monday, 16 December 2024
The study of Langevin dynamics within highly non-convex random landscapes has been crucial for understanding fundamental aspects of glassy dynamics such as aging, violations of fluctuation-dissipation relations, the emergence of effective temperatures, and the role of high dimensionality and entropy in slowing down relaxation. A comprehensive understanding of the activated regime of the dynamics, where the system transitions between metastable states by overcoming energy barriers, remains however elusive. In the talk I will consider a prototypical model of a high-dimensional energy landscape with Gaussian statistics, exhibiting plenty of metastable states. After recalling the main reasons why activated dynamics is theoretically challenging, I will consider effective processes in which the system jumps in the landscape, visiting the closest metastable states at given energy. Understanding this effective dynamics requires to analyze the geometry of the landscape, particularly the joint distribution of triplets of metastable states that lie in the same region of the high-dimensional configuration space. I will report on the results of this geometrical analysis, and comment on implication for activated dynamics.
We have used Langevin dynamics simulations to study the effects of activity in a two-dimensional athermal glass-forming system of Lennard-Jones particles. We consider the limit of infinite persistence time in which the self-propulsion forces on the particles have the same magnitude but different directions that do not change with time. This system exhibits a liquid state for large values of the self-propulsion force and a force-balanced jammed state if the self-propulsion force is smaller than a threshold value. The liquid state is found to exhibit long-range correlations. A length scale extracted from spatial correlations of the velocity field increases with system size as a power law with exponent close to one. Spatial correlations of the self-propulsion forces also exhibit a similar length scale, indicating that the particles self-organize to form a steady state in which particles with similar directions of self-propulsion forces come close to one another and move together. This state is “critical” in the sense that it exhibits a correlation length that diverges in the limit of infinite system size. The velocity pattern in the steady state exhibit an intriguing asymmetry. The development of correlations in time, starting from an initial state with random velocities and forces, is analogous to that in domain growth and coarsening in spin systems after a quench from the disordered to the ordered state. However, quantitative features of this process appear to be different from those in domain growth in spin systems with the same symmetry.
This work was done in collaboration with Suman Dutta, Atharva Shukla, Pinaki Chaudhuri and Madan Rao.
Many-body localization is the paradigm for how interacting quantum systems can resist thermalization in the presence of strong disorder. After a brief recap on the main ideas of many-body localization, I will present new results in the limit of weakly interacting systems, where our numerical simulations indicate that below a certain disorder threshold, weak interactions necessarily lead to ergodic instabilities.
Reference: Jeanne Colbois, Fabien Alet, Nicolas Laflorencie, Phys. Rev. Lett. 133, 116502 (2024)
-
-
We explore the dynamics of the simplest worm algorithm for a class of two-dimensional dimer models and argue that the dynamics of the worm head represents an example of fractional Brownian motion whenever dimer correlations in equilibrium have power-law character. Numerical estimates of the corresponding Hurst exponent and persistence exponent are obtained, and it is further argued that the Hurst exponent is completely determined by equilibrium dimer correlations, while the persistence exponent is additionally influenced by equilibrium correlations between test monomers.
Multicomponent quantum mixtures in one dimension can be characterized by their symmetry under particle exchange. For a strongly interacting Bose-Bose mixture, we show that the time evolution of the momentum distribution from an initially symmetry-mixed state is quasiconstant for a SU(2) symmetry conserving Hamiltonian, while it displays large oscillations in time for the symmetry-breaking case where inter- and intraspecies interactions are different. Using the property that the momentum distribution operator at strong interactions commutes with the class-sum operator, the latter acting as a symmetry witness, we show that the momentum distribution oscillations correspond to symmetry oscillations, with a mechanism analogous to neutrino flavor oscillations.
We will discuss open questions in particulate flows, and why present methods fall short in answering the basic questions. We'll then talk about ways forward, and I will present some of my recent results.
Tuesday, 17 December 2024
-
I will discuss the motion of a slow distinguishable particle (an impurity) in one-dimensional quantum liquids within a microscopic theory. The impurity experiences the friction force due to scattering off thermally excited quasiparticles. I will present detailed analysis of an arbitrarily strong impurity coupling constant in a wide range of temperatures and uncover new regimes of the impurity dynamics.
I will present a few recent exact results about macroscopic fluctuations (large deviations) in non-equilibrium states. These include fluctuations in the non-equilibrium stationary state of systems coupled with unequal reservoirs, in the non-stationary state evolving towards equilibrium, and in active matter. Most of these results are based on fluctuating hydrodynamic descriptions. I will conclude by presenting an approach for this macroscopic description, starting from microscopic dynamics.
-
A model for opinion formation is proposed where an individual's opinion is influenced by interactions with a group of agents. The model introduces a novel bias mechanism that favours one opinion. Several results are reported including the evidence of a critical slowing down as the bias vanishes.
I will review the work done with my collaborators on intruders in single-file systems. In such systems, particles are on a one-dimensional line and cannot pass one another. This confinement induces an anomalous dynamics for any given tagged particle (subdiffusion or sub-ballistic motion). Focusing on lattice models (SEP), I will explain how to uncover the underlying structure behind this anomalous dynamics using two methods that we developed. At equilibrium we were able to solve for non-stationary density-displacement corrélations. And for out-of-equilibrium problems at high density, we show that everything is encoded in the random walk of a single vacancy.
Consider a thermodynamic system that shows a phase transition between an ordered and a disordered phase. The question we ask is: in the parameter regime in which the system exhibits a disordered phase, can we induce order by manoeuvring the system (i) either by forcefully establishing order in a small subset of the total number of degrees of freedom,(ii) or, by shuffling the inherent properties of the individual system constituents among themselves ? Within the ambit of the Kuramoto model, a paradigmatic nonlinear dynamical many-body system, we discuss both analytical and experimental results on how schemes (i)
and (ii) lead to a rich dynamics and, most remarkably, establishing of macroscopic order even in parameter regimes in which the bare dynamics does not support any such ordering.
-
Amorphous materials are ubiquitously used in various mechanical appliances, including in diverse athermal forms. Understanding the response, from a microscopic perspective, to mechanical perturbations is thus fundamental to developing materials with specific functionalities. In this talk, I will summarize some recent results, obtained via extensive numerical simulations, on how these athermal materials respond to different kinds of perturbations, be it at the macro-scale or micro-scale, thus allowing us to gain insights into the physical processes at play.
We shall discuss properties of quantum scars in a Rydberg ladder with staggered detuning. We shall demonstrate that they are qualitatively different from their counterparts in the well-known Rydberg chain and lead to long-time ETH violating imbalance in absence of disorder.
-
Wednesday, 18 December 2024
Many-body localization is a fascinating phenomenon observed in strongly disordered interacting quantum systems. In this talk, I will describe some of our recent works focusing on quantum coherence and single particle excitations across the MBL transition. I will discuss exact relations between various norms of coherence and measure of localization for any generic quantum system and discuss it for a standard model of MBL. Interestingly, though coherence of the full system vanishes in the MBL phase, subsystem coherence increases as the disorder strength increases which can have strong application potential in superconducting qubit arrays and other quantum devices where controlling coherence is a big challenge. On a completely different note, I will discuss single-particle excitations across the MBL transition in systems with random and quasiperiodic potentials and demonstrate that they belong to two different universality classes.
We study a quantum spin chain where the dissipation is induced by the coupling of the density to local baths à la Caldeira and Leggett. In our perspective the bath acts as an annealed disorder with slow dynamics and can induce ordering in the system. At sufficiently strong coupling and zero temperature, it leads in fact to a phase transition between a Luttinger liquid phase and a spin density wave. The nature of the dissipative phase depends on the properties of both the system and the bath and in the incommensurate case it occurs in absence of the opening of a gap but it is due to fractional excitations. We also show, by computing the DC conductivity, that the system is insulating in the presence of a subohmic bath. We interpret this phenomenon as localization induced by the bath.
Release of synaptic vesicles carrying neurotransmitters (also called “quantal content”), form the basis of electrochemical signal transmissions across all synapses. For 70 years, it has been known experimentally that the statistical distribution of each such individual release is a Binomial. Yet the size of the reservoir from which these vesicles get released, fluctuates. Hence the question of the actual distribution of quantal content averaged over these fluctuations, remained open. The problem is difficult due to history dependence -- we make progress by focusing on the steady state. Our work reveals that for fixed frequency electrical input stimulation, the statistically averaged distribution is still a Binomial — for this case, we compare our theory to experimental data from MNTB-LSO synapses of juvenile mice. On the other hand, for random input stimulations the averaged distribution is generically non-Binomial. Often under physiological conditions presynaptic input signals are random. So the exact results in our paper will hopefully help in analyzing experimental distributions in such cases, and make estimates of the model parameters associated with the concerned neuron.
We show that heterogeneity in self-propulsion speed leads to the emergence of effective short-range repulsion among active particles coupled via strong attractive potentials. Taking the example of two harmonically coupled active Brownian particles, we analytically compute the stationary distribution of the distance between them in the strong coupling regime, i.e., where the coupling strength is much larger than the rotational diffusivity of the particles. The effective repulsion in this regime is manifest in the emergence of a minimum distance between the
particles, proportional to the difference in their self-propulsion speeds. Physically, this distance of the closest approach is associated to the orientations of the particles being parallel to each other. We show that the physical scenario remains qualitatively similar for any long-range coupling potential, which is attractive everywhere. Moreover, we show that, for a collection of N particles interacting via pairwise attractive potentials, a short-range repulsion emerges for each pair of particles with different self-propulsion speeds. Finally, we show that our results are robust and hold irrespective of the specific active dynamics of the particles.
-
I will cover examples of low-Reynolds-number turbulence in fluids with polymer additives, in binary-fluid mixtures, and in active fluids.
Consider two systems initially at different temperatures that are then quenched to the same final low temperature. The Mpemba effect is a counterintuitive phenomenon where the initially hotter system reaches equilibrium faster than the colder one. While initially observed in the freezing of water, the Mpemba effect is not limited to this scenario and can be explored in the relaxation dynamics of various systems, including those far from equilibrium, such as granular systems. In this presentation, I will provide a general overview of the Mpemba effect, and then focus on the effect in trapped colloidal particles, both active and inactive. Additionally, I will address the challenges in defining the Mpemba effect and explore potential underlying mechanisms.
In the hydrodynamic theory, the non-equilibrium dynamics of a many-body system is approximated, at large scales of space and time, by irreversible relaxation to local entropy maximisation. This results in a convective equation corrected by viscous or diffusive terms in a gradient expansion, such as the Navier-Stokes equations. Diffusive terms are evaluated using the Kubo formula, and possibly arising from an emergent noise due to discarded microscopic degrees of freedom. In one dimension of space, diffusive scaling is often broken as noise leads to super-diffusion. But in linearly degenerate hydrodynamics, such as that of integrable models, diffusive behaviors are observed, and it has long be thought that the standard diffusive picture remains valid. In this letter, we show that in such systems, the Navier-Stokes equation breaks down beyond linear response. We demonstrate that diffusive-order corrections do not take the form of a gradient expansion. Instead, they are completely determined by ballistic transport of initial-state fluctuations, and obtained from the non-local two-point correlations recently predicted by the ballistic macroscopic fluctuation theory (BMFT); the resulting hydrodynamic equations are reversible. To do so, we establish a regularised fluctuation theory, putting on a firm basis the recent idea that ballistic transport of initial-state fluctuations determines fluctuations and correlations beyond the Euler scale. This extends the idea of ``diffusion from convection'' previously developed to explain the Kubo formula in integrable systems, to generic non-equilibrium settings.
We study the exact fluctuating hydrodynamics of the scaled Light-Heavy model (sLH), in which two species of particles (light and heavy) interact with a fluctuating surface. This model is similar in definition to the unscaled Light-Heavy model (uLH), except it uses rates scaled with the system size. The consequence, it turns out, is a phase diagram that differs from that of the unscaled model. We derive the fluctuating hydrodynamics for this model using an action formalism involving the construction of path integrals for the probability of different states that give the complete macroscopic picture starting from the microscopic one. This is then used to obtain a form for the two-point static correlation between fluctuations in density fields in the homogeneous phase in the steady state. We find that these theoretical results match well with microscopic simulations away from the critical line.
In first part of the talk, we will present our study of melting in a two-dimensional system of classical particles with Gaussian-core interactions in disordered environments. The clean system validates the conventional two-step melting with a hexatic phase intervening between the solid and the liquid. This picture gets significantly modified in the presence of disorder. Impurities in a random distribution of pinning centers force a hexatic-like low-temperature phase to extend up to T=0, which transits into the liquid at a single melting temperature T_RP. In contrast, pinning centers located at randomly chosen sites of a perfect crystal of the clean system anchor a solid at low temperatures, which undergoes a direct transition to the liquid at T_CP. Thus, the two-step melting is lost in either case of disorder. Addressing dynamics across melting, we will demonstrate intriguing signatures of cooperative motion of particles in string-like paths found at low temperatures. Such motional footprints are standard to glasses and supercooled liquids, but we realize them in equilibrium dynamics, even in a pure system.
In second part, we will discuss the coherent many-body dynamics after a superconducting attraction is quenched from a an initial to final value of a conventional BCS superconductor. We will show how the asymptotic steady state features nonequilibrium “phases” with different properties depending on the parameter space of quench.
Thursday, 19 December 2024
Of special interest in aggregation processes is the occurrence of extremely large clusters, or condensates, which hold a finite fraction of the mass. We find that condensates coexist with a power law distribution in the Takayasu model of aggregation with input, even though the mass is not conserved. While approaching steady state, mini-condensates form on a growing length scale. There is a single mobile condensate in steady state, and its movement leads to a dynamic re-organization of the landscape on a macroscopic scale, along with an emergent power law which differs from the usual Takayasu power. In an open system, the exit of the condensate from the edges leads to intermittent fluctuations of the total mass in steady state, quantified through a divergence of the scaled kurtosis.
[1] A. Das and M. Barma, Indian J. Phys., Special issue on Nonequilibrium Statistical Physics (2024)
https://link.springer.com/article/10.1007/s12648-023-03030-1
[2] R. Negi, R. Pereira and M. Barma, arXiv:2407.09827 , to appear, Phys. Rev. E (2024)
he emergence of surprising collective behaviors in systems driven out of equilibrium by local energy injection at the particle level remains a central theme in the study of active matter. Recently, chaotic flows reminiscent of turbulence have garnered significant attention due to their appearance in diverse biological and physical active matter systems. In this talk, I will demonstrate how even the simplest model of active particles -— self-propelled point particles -— can exhibit mesoscale flows, characterized by streams and vortices, when very persistent active forces compete with crowding at high densities.
In the second part, I will introduce a minimal model of non-reciprocal interactions inspired by human crowds, which generates collective flows strikingly similar to those of the self-propelled particles. Interestingly, as the system approaches the equilibrium limit by reducing non-reciprocity, it undergoes an absorbing phase transition characterized by an infinite number of absorbing states and critical exponents consistent with the conserved directed percolation universality class.
Recent experiments have implemented resetting by means of a time-varying external trap whereby trap stiffnesses are changed from an initial to a final value in finite-time. Such setups have also been studied in the context of Landauer's erasure principle. We analyse the thermodynamic costs of such a setup in steady state.
Stochastic resetting has recently become a subject of immense interest. Most of the theoretical studies so far focused on instantaneous resetting at a constant rate which can be a major impediment to practical realisation or experimental verification in the field. This is because in the real world, taking a particle from one place to another requires finite time and also the resetting rate would be time dependent. In this talk I will discuss possible generalisations of the existing theory by incorporating time dependent rate in the instantaneous resetting problem and by considering non-instantaneous resetting. I will demonstrate how different features of a brownian particle, such as non-equilibrium stationary state, relaxation to it and search efficiency get affected by these generalisations.
We explore how self-organization in swarms of interacting self-propelled particles can be used to optimize their displacement in confined geometries. Using a discrete model with Vicsek-like interactions, we examine how channel geometry influence pattern formation and transport properties. Wall-induced particle accumulation leads to clogs and band formations that obstruct movement. This analysis enables us to develop global strategies for controlling particle alignment and optimizing displacement. We apply reinforcement learning techniques to devise policies that enhance transport efficiency.
We will discuss two spin-1 versions of the Derrida’s Random Energy model.
Friday, 20 December 2024
The far-field decay of the concentration of motile particles around a static inclusion is well studied in the case of “dry” active matter. We show here that the scenario is dramatically different when the viscous hydrodynamic interaction enters, especially if the object is polar in shape. Advection by fluid flow and diffusion enter on the same footing, a “marginality” that leads to a power-law decay exponent for the concentration varying continuously with a dimensionless measure of the force required to hold the inclusion in place, and a singular distinction between the axisymmetric and non-axisymmetric cases. We
hope our findings will inspire experimental studies on inclusions in swimmer suspensions. This work was done in collaboration with Thibaut Arnoulx de Pirey and Yariv Kafri.
-
Motivated by recent debates around the many-body localization (MBL) problem, and in particular its stability against systemwide resonances, we investigate long-distance spin-spin correlations across the phase diagram of the random-field XXZ model, with a particular focus on the strong disorder regime. Building on state-of-the-art shift-invert diagonalization techniques, we study the high-energy behavior of transverse and longitudinal correlation functions, computed at the largest possible distance, for a broad range of disorder and interaction strengths. Our results show that while transverse correlations display a fairly stable exponential decay over the entire XXZ phase diagram, longitudinal correlations exhibit markedly different behavior, revealing distinct physical regimes. More precisely, we identify an intermediate disorder region where standard observables show well-converged MBL behavior [J. Colbois et al., Phys. Rev. Lett. 133, 116502 (2024)] while the distributions of longitudinal correlations reveal unexpected fat-tails towards large values. These rare events strongly influence the average decay of longitudinal correlations, which we find to be algebraic in a broad region inside the supposed MBL phase, whereas the typical decay remains mostly exponential. At stronger disorder and weaker interactions, this intermediate regime is replaced by a more conventional exponential decay with short correlation lengths for both typical and average correlators, as expected for standard localization. Our findings shed light on the systemwide instabilities and raise important questions about the impact of such rare but large long-range correlations on the stability of the MBL phase. Finally, we discuss the possible fate of the intermediate region in the context of recent perspectives in the field.
-
I will present recent findings from the study of the Random Field Ising Model across various dimensions. The aim is to explore the existence of Universality and Dimensional Reduction within this model. Using results from large-scale numerical simulations, I will argue that
Dimensional Reduction holds true at D=5, and we observe the consequences of supersymmetry.
A fundamental aspect crucial for the survival of various animal species is their ability to successfully return home, whether it involves migration, foraging for food, or locating a breeding site. This innate behavior, known as Homing, is surprisingly ubiquitous, allowing
animals to navigate back from seemingly unfamiliar locations over considerable distances. In this talk, I will try to shed some light on this phenomena from the perspective of stochastic resetting. This connection helps us to uncover universal characteristics of
Homing paths using a self-propelled Robotic Forager.
We construct a d-dimensional Eddy Damped Quasi-Normal Markovian (EDQNM) Closure Model to study dynamo action in arbitrary dimensions. In particular, we find lower and upper critical dimensions for sustained dynamo action in this incompressible problem. Our model is adaptable for future studies incorporating helicity, compressible effects and a wide range of magnetic Reynolds and Prandtl numbers.
In this talk, I shall discuss our results characterizing the spatiotemporal chaos in classical spin systems-- both magnetically ordered as well as frustrated with the latter showing a classical spin-liquid down to very low temperatures.
The air we inhale brings along a variety of harmful aerosols, which if deposited on the wall of the airways can cause severe respiratory illness. The lung's primary defense against airborne particles is provided by a film of mucus which lines the airway walls. This film is continuously transported, upward and out of the lungs, by a carpet of wall-attached cilia; thus, harmful particles are evacuated provided they deposit on the mucus. In this talk, I will show that particles can manage to avoid the mucus and deposit on the exposed airway wall. The fate of particles depends on their size—which sets the strength of Brownian or inertial forces—as well as the coupled flow of air and mucus. The Rayleigh-Plateau instability of the annular inter-fluid interface plays a key role, by controlling the morphology of the mucus film, and ultimately leads to a counter-intuitive result: More mucus does not imply more trapping; instead, more particles—allergens, pathogens, and hopefully aerosolised drugs—are able to reach the wall.
We will exactly determine dynamical correlation functions for conformal field theories (CFTs) driven by conformal generators beyond the dilatation operator, in 1+1 as well as 3+1 dimensions. Under floquet dynamics the system falls into one of three universal phases which we try to explain geometrically.