U. Marini Bettolo Marconi

Clustering and Non-Gaussian Behavior in Granular Matter

We investigate the properties of a model of granular matter consisting of N Brownian particles on a line, subject to inelastic mutual collisions. This model displays a genuine thermodynamic limit for the mean values of the energy, and the energy dissipation. When the typical relaxation time t associated with the Brownian process is small compared with the mean collision time tc the spatial density is nearly homogeneous and the velocity probability distribution is Gaussian. In the opposite limit t ? tc one has strong spatial clustering, with a fractal distribution of particles, and the velocity probability distribution strongly deviates from the Gaussian one. [S0031-9007(98)07496-1]

Influence of correlations on the velocity statistics of scalar granular gases

The free evolution of inelastic particles in one dimension is studied by means of Molecular Dynamics (MD), of an inelastic pseudo-Maxwell model and of a lattice model, with emphasis on the role of spatial correlations. We present a new exact solution of the 1d granular pseudo-Maxwell model for the scaling distribution of velocities and discuss how this model fails to describe correctly the homogeneous cooling stage of the 1d granular gas. Embedding the pseudo-Maxwell gas on a lattice (hence allowing for the onset of spatial correlations), we find a much better agreement with the MD simulations even in the inhomogeneous regime. This is seen by comparing the velocity distributions, the velocity profiles and the structure factors of the velocity field.

Kinetic approach to granular gases.

We address the problem of the so-called granular gases, i.e., gases of massive particles in rapid movement undergoing inelastic collisions. We introduce a class of models of driven granular gases for which the stationary state is the result of the balance between the dissipation and the random forces which inject energies. These models exhibit a genuine thermodynamic limit, i.e., at fixed density the mean values of kinetic energy and dissipated energy per particle are independent of the number N of particles, for large values of N. One has two regimes: when the typical relaxation time tau of the driving Brownian process is small compared with the mean collision time tau(c) the spatial density is nearly homogeneous and the velocity probability distribution is Gaussian. In the opposite limit tau>>tau(c) one has strong spatial clustering, with a fractal distribution of particles, and the velocity probability distribution strongly deviates from the Gaussian one. Simulations performed in one and two dimensions under the Stosszahlansatz Boltzmann approximation confirm the scenario. Furthermore, we analyze the instabilities bringing to the spatial and the velocity clusterization. Firstly, in the framework of a mean-field model, we explain how the existence of the inelasticity can lead to a spatial clusterization; on the other hand, we discuss, in the framework of a Langevin dynamics treating the collisions in a mean-field way, how a non-Gaussian distribution of velocity can arise. The comparison between the numerical and the analytical results exhibits an excellent agreement.

Lattice Boltzmann method for inhomogeneous fluids

We present a lattice-based numerical method to describe the non-equilibrium behavior of a simple fluid under non-uniform spatial conditions. The evolution equation for the one-particle phase-space distribution function is derived starting from a microscopic description of the system. It involves a series of approximations which are similar to those employed in theories of inhomogeneous fluids, such as the density functional theory. Among the merits of the present approach: the possibility to determine the equation of state of the model, the transport coefficients and to provide an efficient method of numerical solution under non-uniform conditions. The algorithm is tested in a particular non-equilibrium situation, namely the steady flow of a hard-sphere fluid across a narrow slit. Pronounced non-hydrodynamic oscillations in the density and velocity profiles are found.

Electro-osmotic flows under nanoconfinement: A self-consistent approach

We introduce a theoretical and numerical method to investigate the properties of electro-osmotic flows under conditions of extreme confinement. The present approach, aiming to provide a simple modeling of electrolyte solutions described as ternary mixtures, which comprises two ionic species and a third uncharged component, is an extension of our recent work on binary neutral mixtures. The approach, which combines elements of kinetic theory, density functional theory with Lattice-Boltzmann algorithms, is microscopic and self-consistent and does not require the use of constitutive equations to determine the fluxes. Numerical solutions are obtained by solving the resulting coupled equations for the one-particle phase-space distributions of the species by means of a Lattice-Boltzmann discretization procedure. Results are given for the microscopic density and velocity profiles and for the volumetric and charge flow.

Fluctuation-Induced Casimir Forces in Granular Fluids

We numerically investigate the behavior of driven noncohesive granular media and find that two fixed large intruder particles, immersed in a sea of small particles, experience, in addition to a short-range depletion force, a long-range repulsive force. The observed long-range interaction is fluctuation-induced and we propose a mechanism similar to the Casimir effect that generates it: The hydrodynamic fluctuations are geometrically confined between the intruders, producing an unbalanced renormalized pressure. An estimation based on computing the possible Fourier modes explains the repulsive force and is in qualitative agreement with the simulations.

Noise rectification and fluctuations of an asymmetric inelastic piston

We consider a massive inelastic piston, whose opposite faces have different coefficients of restitution, moving under the action of an infinitely dilute gas of hard disks maintained at a fixed temperature. The dynamics of the piston is Markovian and obeys a continuous Master Equation: however, the asymmetry of restitution coefficients induces a violation of detailed balance and a net drift of the piston, as in a Brownian ratchet. Numerical investigations of such non-equilibrium stationary state show that the velocity fluctuations of the piston are symmetric around the mean value only in the limit of large piston mass, while they are strongly asymmetric in the opposite limit. Only taking into account such an asymmetry, i.e. including a third parameter in addition to the mean and the variance of the velocity distribution, it is possible to obtain a satisfactory analytical prediction for the ratchet drift velocity.

Nonequilibrium fluctuations in a driven stochastic Lorentz gas

We study the stationary state of a one-dimensional kinetic model where a probe particle is driven by an external field E and collides, elastically or inelastically, with a bath of particles at temperature T. We focus on the stationary distribution of the velocity of the particle, and of two estimates of the total entropy production Δs(tot). One is the entropy production of the medium Δs(m), which is equal to the energy exchanged with the scatterers, divided by a parameter θ, coinciding with the particle temperature at E=0. The other is the work W done by the external field, again rescaled by θ. At small E, a good collapse of the two distributions is found: in this case, the two quantities also verify the fluctuation relation (FR), indicating that both are good approximations of Δs(tot). Differently, for large values of E, the fluctuations of W violate the FR, while Δs(m) still verifies it.

Cooling of a lattice granular fluid as an ordering process.

We present a microscopic model of granular medium to study the role of dynamical correlations and the onset of spatial order induced by the inelasticity of the interactions on the velocity field. In spite of its simplicity and intrinsic limitations, it features several aspects of the rich phenomenology observed in granular materials and allows to make contact with other topics of statistical mechanics such as diffusion processes, domain growth, aging phenomena. Interestingly, while local observables, being controlled by the largest wavelength fluctuations, seem to suggest a purely diffusive behavior, the formation of spatially extended structures and topological defects, such as vortices and shocks, reveals a more complex scenario. Finally, only for quasielastic systems, we observe a neat scale separation, which represents a fundamental hypothesis to develop a granular hydrodynamics.

Driven granular gases with gravity

We study fluidized granular gases in a stationary state determined by the balance between external driving and bulk dissipation. The two considered situations are inspired by recent experiments, where gravity plays a major role as a driving mechanism: in the first case, gravity acts only in one direction and the bottom wall is vibrated; in the second case, gravity acts in both directions and no vibrating walls are present. Simulations performed under the molecular chaos assumption show averaged profiles of density, velocity, and granular temperature that are in good agreement with the experiments. Moreover, we measure velocity distributions that show strong non-Gaussian behavior, as experiments pointed out, but also density correlations accounting for clustering, at odds with the experimental results. The hydrodynamics of the first model is discussed and an exact solution is found for the density and granular temperature as functions of the distance from the vibrating wall. The limitations of such a solution, in particular in a broad layer near the wall injecting energy, are discussed.