This course covers advanced topics in statistical physics, mostly out of equilibrium. The relevant formalism (statistics of sums of random variables, stochastic processes,…) is first introduced. Then the course covers the standard topics of Langevin and Fokker-Planck equations, as well as linear response close to equilibrium and fluctuation-dissipation relations. Far from equilibrium extensions like fluctuation relations are also presented. In the last part, the course discusses examples of N-body distributions in out-of-equilibrium models, as well as the more versatile approach using an effective description in terms of a one-body distribution. Some examples of applications are taken from the recent literature.
Lecture 1: Sums of random variables
Statistics of sums of random variables, central limit theorem and its generalization to broadly distributed random variables. Examples of correlated or non-identically distributed random variables. Large deviations and their relevance in statistical physics.
Application of sums of random variables to equilibrium disordered systems, example of the Random Energy Model.
Lecture 2: Stochastic processes
Markov jump processes, master equation, detailed balance. Random walk on a lattice. Diffusive limit and diffusion equation. Continuous time random walks and anomalous diffusion.
Statistics of the first return time of a random walk.
Relaxation to equilibrium, increase of entropy, expansion over the eigenvectors of the Markov operator. Non-stationary dynamics at large time and aging phenomenon.
Lecture 3: Langevin and Fokker-Planck equations
Langevin equation, definition and physical motivation. Detailed balance.
Fokker-Planck equation. Derivation from a biased random walk on a lattice. Kramers-Moyal expansion. Stationary solution without flux (equilibrium) and with flux. Physical examples.
Relaxation to equilibrium and Ornstein-Uhlenbeck process.
Langevin equation with multiplicative noise, Ito and Stratonovitch discretization rules. Corresponding Fokker-Planck equation. Wong-Zakai theorem. Stochastic on-off intermittency as an example of multiplicative noise. Stochastic calculus and Ito's lemma.
Lecture 4: Response to an external field
Linear response close to equilibrium. Response function and Kubo formula. Time-correlation function and fluctuation-dissipation relation. Microreversibility and symmetry of response functions. Kramers-Krönig relation.
Statistics of trajectories and fluctuation relation far from equilibrium. Comparison of a trajectory with the time-reversed one. Statistics over trajectories. Fluctuation relation for discrete-time stochastic processes. Jarzinsky and Crooks relations.
Lecture 5: Examples of N-body distributions in stationary nonequilibrium states
One-dimensional lattice particle models, Zero-Range Process, Asymmetric Simple Exclusion Process. Matrix Product Ansatz solution.
Tapping dynamics for granular matter: Edwards postulate for the N-body distribution.
Approximate N-body distribution for interacting active particles with random self-propulsion forces.
Lecture 6: Effective description in terms of one-body problem
Dynamics with persistent interactions, approximation in terms of a non-linear Fokker-Planck equation with a self-consistent force field.
Collisional dynamics in the dilute regime, kinetic theory for the one-body distribution.
Principle of the derivation of continuum equations for the slow modes.
examen écrit 3h