Langevin方程与数值模拟-NUSPhysics.DOC

Monte Carlo Simulation of Non-equilibrium Dynamics in Statistical Physics Bo Zheng Zhejiang University Contents I Introduction II Langevin Eq. and Monte Carlo dynamics III Phase ordering dynamics IV Critical dynamics far from equilibrium V Disordered systems VI Concluding remarks I Introduction Many-body Systems difficult to solve the Eqs. of motion Statistical Physics Equilibrium Ensemble theory Non-equilibrium ? At macroscopic level * Non-equilibrium state: observable does evolve as time t ( t → ∞ ) * Equilibrium state: observable does not evolve as time t Exceptional * Time-dependent perturbations: t → ∞ stationary state ～ equilibrium * External fields: t → ∞ transportation ～ dynamic behavior Non-equilibrium statistical physics * Dynamic behavior in or around equilibrium: long-time behavior transportation ‘stationary state’ * Dynamic behavior far from equilibrium: ‘short time’ behavior Equations of motion * Microscopic fundamental Eqs.: Newton, Heisenberg Eqs. * Effective equations: Liouville’s 、Boltzmann Eqs. * Phenomenological Eqs.: Langevin Eq., Monte Carlo dynamics Master Eq. …… Dynamics far from equilibrium * Relaxational * Phenomenological Langevin Eq., Monte Carlo dynamics Examples phase ordering damage spreading critical dynamics spin glass dynamics structural glass dynamics interface growth chemical reaction…… II Langevin Eq. and Monte Carlo dynamics 1、Langevin方程 问题：系统的作用量或Hamiltonian量为S 平衡态分布 假设系统时处于一初始状态 系统如何演化至平衡态？ 如果初始状态不是平衡态，这便是一个驰豫动力学过程。 如果初始状态是平衡态，这是平衡态的动力学涨落问题。 单自由度的Langevin方程和Fokker-Planck方程 Langevin方程 对固定 ( 例如：