SearcharxivSearch

arXiv subjects

M. H. Ernst

Publications and source records attributed to M. H. Ernst.

At least 19 recordsLinked to original sources

Quasi-elastic solutions to the nonlinear Boltzmann equation for dissipative gases

The solutions of the one-dimensional homogeneous nonlinear Boltzmann equation are studied in the QE-limit (Quasi-Elastic; infinitesimal dissipation) by a combination of analytical and numerical techniques. Their behavior at large velocities differs qualitatively from that for higher dimensional systems. In our generic model, a dissipative fluid is maintained in a non-equilibrium steady state by a stochastic or deterministic driving force. The velocity distribution for stochastic driving is regular and for infinitesimal dissipation, has a stretched exponential tail, with an unusual stretching exponent $b_{QE} = 2b$, twice as large as the standard one for the corresponding $d$-dimensional system at finite dissipation. For deterministic driving the behavior is more subtle and displays singularities, such as multi-peaked velocity distribution functions. We classify the corresponding velocity distributions according to the nature and scaling behavior of such singularities.

cond-mat.stat-mech

Boltzmann equation for dissipative gases in homogeneous states with nonlinear friction

Combining analytical and numerical methods, we study within the framework of the homogeneous non-linear Boltzmann equation, a broad class of models relevant for the dynamics of dissipative fluids, including granular gases. We use the new method presented in a previous paper [J. Stat. Phys. 124, 549 (2006)] and extend our results to a different heating mechanism, namely a deterministic non-linear friction force. We derive analytically the high energy tail of the velocity distribution and compare the theoretical predictions with high precision numerical simulations. Stretched exponential forms are obtained when the non-equilibrium steady state is stable. We derive sub-leading corrections and emphasize their relevance. In marginal stability cases, power-law behaviors arise, with exponents obtained as the roots of transcendental equations. We also consider some simple BGK (Bhatnagar, Gross, Krook) models, driven by similar heating devices, to test the robustness of our predictions.

cond-mat.stat-mech

The nonlinear fragmentation equation

We study the kinetics of nonlinear irreversible fragmentation. Here fragmentation is induced by interactions/collisions between pairs of particles, and modelled by general classes of interaction kernels, and for several types of breakage models. We construct initial value and scaling solutions of the fragmentation equations, and apply the "non-vanishing mass flux "criterion for the occurrence of shattering transitions. These properties enable us to determine the phase diagram for the occurrence of shattering states and of scaling states in the phase space of model parameters.

cond-mat.stat-mech

The Boltzmann equation for driven systems of inelastic soft spheres

We study a generic class of inelastic soft sphere models with a binary collision rate $g^ν$ that depends on the relative velocity $g$. This includes previously studied inelastic hard spheres ($ν=1$) and inelastic Maxwell molecules ($ν=0$). We develop a new asymptotic method for analyzing large deviations from Gaussian behavior for the velocity distribution function $f(c)$. The framework is that of the spatially uniform nonlinear Boltzmann equation and special emphasis is put on the situation where the system is driven by white noise. Depending on the value of exponent $ν$, three different situations are reported. For $ν<-2$, the non-equilibrium steady state is a repelling fixed point of the dynamics. For $ν>-2$, it becomes an attractive fixed point, with velocity distributions $f(c)$ having stretched exponential behavior at large $c$. The corresponding dominant behavior of $f(c)$ is computed together with sub-leading corrections. In the marginally stable case $ν=-2$, the high energy tail of $f(c)$ is of power law type and the associated exponents are calculated. Our analytical predictions are confronted with Monte Carlo simulations, with a remarkably good agreement.

cond-mat.stat-mech

The rich behavior of the Boltzmann equation for dissipative gases

Within the framework of the homogeneous non-linear Boltzmann equation, we present a new analytic method, without the intrinsic limitations of existing methods, for obtaining asymptotic solutions. This method permits extension of existing results for Maxwell molecules and hard spheres to large classes of particle interactions, from very hard spheres to softer than Maxwell molecules, as well as to more general forcing mechanisms, beyond free cooling and white noise driving. By combining this method with numerical solutions, obtained from the Direct Simulation Monte Carlo (DSMC) method, we study a broad class of models relevant for the dynamics of dissipative fluids, including granular gases. We establish a criterion connecting the stability of the non-equilibrium steady state to an exponentially bound form for the velocity distribution $F$, which varies depending on the forcing mechanism. Power laws arise in marginal stability cases, of which several new cases are reported. Our results provide a minimal framework for interpreting large classes of experiments on driven granular gases.

cond-mat.stat-mech

Generalized Green-Kubo formulas for fluids with impulsive, dissipative, stochastic and conservative interactions

We present a generalization of the Green-Kubo expressions for thermal transport coefficients $μ$ in complex fluids of the generic form, $μ= μ_\infty +\int^\infty_0 dt V^{-1} < J_ε\exp(t {\cal L}) J >_0$, i.e. a sum of an instantaneous transport coefficient $μ_\infty$, and a time integral over a time correlation function in a state of thermal equilibrium between a current $J$ and a transformed current $J_ε$. The streaming operator $\exp(t{\cal L})$ generates the trajectory of a dynamical variable $J(t) =\exp(t{\cal L}) J$ when used inside the thermal average $<...>_0$. These formulas are valid for conservative, impulsive (hard spheres), stochastic and dissipative forces (Langevin fluids), provided the system approaches a thermal equilibrium state. In general $μ_\infty \neq 0$ and $J_ε\neq J$, except for the case of conservative forces, where the equality signs apply. The most important application in the present paper is the hard sphere fluid.

cond-mat.stat-mech

New Green-Kubo formulas for transport coefficients in hard sphere-, Langevin fluids and the likes

We present generalized Green-Kubo expressions for thermal transport coefficients $μ$ in non-conservative fluid-type systems, of the generic form, $μ$ $= μ_\infty$ $+\int^\infty_0 dt V^{-1} \av{I_ε\exp(t {\cal L}) I}_0$ where $\exp(t{\cal L})$ is a pseudo-streaming operator. It consists of a sum of an instantaneous transport coefficient $μ_\infty$, and a time integral over a time correlation function in a state of thermal equilibrium between a current $I$ and its conjugate current $I_ε$. This formula with $μ_\infty \neq 0$ and $I_ε\neq I$ covers vastly different systems, such as strongly repulsive elastic interactions in hard sphere fluids, weakly interacting Langevin fluids with dissipative and stochastic interactions satisfying detailed balance conditions, and "the likes", defined in the text. For conservative systems the results reduce to the standard formulas.

cond-mat.stat-mech

Power law tails of time correlations in a mesoscopic fluid model

In a quenched mesoscopic fluid, modelling transport processes at high densities, we perform computer simulations of the single particle energy autocorrelation function C_e(t), which is essentially a return probability. This is done to test the predictions for power law tails, obtained from mode coupling theory. We study both off and on-lattice systems in one- and two-dimensions. The predicted long time tail ~ t^{-d/2} is in excellent agreement with the results of computer simulations. We also account for finite size effects, such that smaller systems are fully covered by the present theory as well.

cond-mat.stat-mech

Universal power law tails of time correlation functions

The universal power law tails of single particle and multi-particle time correlation functions are derived from a unifying point of view, solely using the hydrodynamic modes of the system. The theory applies to general correlation functions, and to systems more general than classical fluids. Moreover it is argued that the collisional transfer part of the stress-stress correlation function in dense classical fluids has the same long time tail $\sim t^{-1-d/2}$ as the velocity autocorrelation function in Lorentz gases.

cond-mat.stat-mech

Anomalous velocity distributions in inelastic Maxwell gases

This review is a kinetic theory study investigating the effects of inelasticity on the structure of the non-equilibrium states, in particular on the behavior of the velocity distribution in the high energy tails. Starting point is the nonlinear Boltzmann equation for spatially homogeneous systems, which supposedly describes the behavior of the velocity distribution function in dissipative systems as long as the system remains in the homogeneous cooling state, i.e. on relatively short time scales before the clustering and similar instabilities start to create spatial inhomogeneities. This is done for the two most common models for dissipative systems, i.e. inelastic hard spheres and inelastic Maxwell particles. In systems of Maxwell particles the collision frequency is independent of the relative velocity of the colliding particles, and in hard sphere systems it is linear. We then demonstrate the existence of scaling solutions for the velocity distribution function, $F(v,t) \sim v_0(t)^{-d} f((v/v_0(t))$, where $v_0$ is the r.m.s. velocity. The scaling form $f(c)$ shows overpopulation in the high energy tails. In the case of freely cooling systems the tails are of algebraic form, $ f(c)\sim c^{-d-a}$, where the exponent $a$ may or may not depend on the degree of inelasticity, and in the case of forced systems the tails are of stretched Gaussian type $f(v)\sim\exp[-β(v/v_0)^b]$ with $b <2$.

cond-mat.stat-mech

Exact steady state solution of the Boltzmann equation: A driven 1-D inelastic Maxwell gas

The exact nonequilibrium steady state solution of the nonlinear Boltzmann equation for a driven inelastic Maxwell model was obtained by Ben-Naim and Krapivsky [Phys. Rev. E 61, R5 (2000)] in the form of an infinite product for the Fourier transform of the distribution function $f(c)$. In this paper we have inverted the Fourier transform to express $f(c)$ in the form of an infinite series of exponentially decaying terms. The dominant high energy tail is exponential, $f(c)\simeq A_0\exp(-a|c|)$, where $a\equiv 2/\sqrt{1-α^2}$ and the amplitude $A_0$ is given in terms of a converging sum. This is explicitly shown in the totally inelastic limit ($α\to 0$) and in the quasi-elastic limit ($α\to 1$). In the latter case, the distribution is dominated by a Maxwellian for a very wide range of velocities, but a crossover from a Maxwellian to an exponential high energy tail exists for velocities $|c-c_0|\sim 1/\sqrt{q}$ around a crossover velocity $c_0\simeq \ln q^{-1}/\sqrt{q}$, where $q\equiv (1-α)/2\ll 1$. In this crossover region the distribution function is extremely small, $\ln f(c_0)\simeq q^{-1}\ln q$.

cond-mat.stat-mech

Asymptotic solutions of the nonlinear Boltzmann equation for dissipative systems

Analytic solutions $F(v,t)$ of the nonlinear Boltzmann equation in $d$-dimensions are studied for a new class of dissipative models, called inelastic repulsive scatterers, interacting through pseudo-power law repulsions, characterized by a strength parameter $ν$, and embedding inelastic hard spheres ($ν=1$) and inelastic Maxwell models ($ν=0$). The systems are either freely cooling without energy input or driven by thermostats, e.g. white noise, and approach stable nonequilibrium steady states, or marginally stable homogeneous cooling states, where the data, $v^d_0(t) F(v,t)$ plotted versus $c=v/v_0(t)$, collapse on a scaling or similarity solution $f(c)$, where $v_0(t)$ is the r.m.s. velocity. The dissipative interactions generate overpopulated high energy tails, described generically by stretched Gaussians, $f(c) \sim \exp[-βc^b]$ with $0 < b < 2$, where $b=ν$ with $ν>0$ in free cooling, and $b=1+{1/2} ν$ with $ν\geq 0$ when driven by white noise. Power law tails, $f(c) \sim 1/c^{a+d}$, are only found in marginal cases, where the exponent $a$ is the root of a transcendental equation. The stability threshold depend on the type of thermostat, and is for the case of free cooling located at $ν=0$. Moreover we analyze an inelastic BGK-type kinetic equation with an energy dependent collision frequency coupled to a thermostat, that captures all qualitative properties of the velocity distribution function in Maxwell models, as predicted by the full nonlinear Boltzmann equation, but fails for harder interactions with $ν>0$.

cond-mat.stat-mech

Scaling Solutions of Inelastic Boltzmann Equations with Over-populated High Energy Tails

This paper deals with solutions of the nonlinear Boltzmann equation for spatially uniform freely cooling inelastic Maxwell models for large times and for large velocities, and the nonuniform convergence to these limits. We demonstrate how the velocity distribution approaches in the scaling limit to a similarity solution with a power law tail for general classes of initial conditions and derive a transcendental equation from which the exponents in the tails can be calculated. Moreover on the basis of the available analytic and numerical results for inelastic hard spheres and inelastic Maxwell models we formulate a conjecture on the approach of the velocity distribution function to a scaling form.

cond-mat.stat-mech

Randomly Driven Granular Fluids: collisional statistics and short scale structure

We present a molecular dynamics and kinetic theory study of granular material, modeled by inelastic hard disks, fluidized by a random driving force. The focus is on collisional averages and short distance correlations in the non-equilibrium steady state, in order to analyze in a quantitative manner the breakdown of molecular chaos, i.e. factorization of the two-particle distribution function, $f^{(2)}(x_1,x_2) \simeq χf^(1)(x_1) f^{(1)}(x_2)$ in a product of single particle ones, where $x_i = \{{\bf r}_i, {\bf v}_i \}$ with $i=1,2$ and $χ$ represents the position correlation. We have found that molecular chaos is only violated in a small region of the two-particle phase space $\{x_1,x_2\}$, where there is a predominance of grazing collisions. The size of this singular region grows with increasing inelasticity. The existence of particle- and noise-induced recollisions magnifies the departure from mean field behavior. The implications of this breakdown in several physical quantities are explored.

cond-mat.stat-mech

Towards a Landau-Ginzburg-type Theory for Granular Fluids

In this paper we show how, under certain restrictions, the hydrodynamic equations for the freely evolving granular fluid fit within the framework of the time dependent Landau-Ginzburg (LG) models for critical and unstable fluids (e.g. spinodal decomposition). The granular fluid, which is usually modeled as a fluid of inelastic hard spheres (IHS), exhibits two instabilities: the spontaneous formation of vortices and of high density clusters. We suppress the clustering instability by imposing constraints on the system sizes, in order to illustrate how LG-equations can be derived for the order parameter, being the rate of deformation or shear rate tensor, which controls the formation of vortex patterns. From the shape of the energy functional we obtain the stationary patterns in the flow field. Quantitative predictions of this theory for the stationary states agree well with molecular dynamics simulations of a fluid of inelastic hard disks.

cond-mat.stat-mech

Cahn-Hilliard Theory for Unstable Granular Flows

A Cahn-Hilliard-type theory for hydrodynamic fluctuations is proposed that gives a quantitative description of the slowly evolving spatial correlations and structures in density and flow fields in the early stages of evolution of freely cooling granular fluids. Two mechanisms for pattern selection and structure formation are identified: unstable modes leading to density clustering (compare spinodal decomposition), and selective noise reduction (compare peneplanation in structural geology) leading to vortex structures. As time increases, the structure factor for the density field develops a maximum, which shifts to smaller wave numbers. This corresponds to an approximately diffusively growing length scale for density clusters. The spatial velocity correlations exhibit algebraic decay $\sim r^{-d}$ on intermediate length scales. The theoretical predictions for spatial correlation functions and structure factors agree well with molecular dynamics simulations of a system of inelastic hard disks.

cond-mat.stat-mech

Randomly Driven Granular Fluids: large scale structure

The nonequilibrium steady state of a granular fluid, driven by a random external force, is demonstrated to exhibit long range correlations, which behave as $\sim 1/r$ in three and $\sim \ln(L/r)$ in two dimensions. We calculate the corresponding structure factors over the whole range of wave numbers, and find good agreement with two-dimensional molecular dynamics simulations. It is also shown by means of a mode coupling calculation, how the mean field values for the steady state temperature and collision frequency, as obtained from the Enskog-Boltzmann equation, are renormalized by long wavelength hydrodynamic fluctuations.

cond-mat.stat-mech