Fast-Forward Langevin Dynamics with Momentum Flips

Abstract

Stochastic thermostats based on the Langevin equation, in which a system is coupled to an external heat bath, are popular methods for temperature control in molecular dynamics simulations due to their ergodicity and their ease of implementation. Traditionally, these thermostats suffer from sluggish behaviour in the limit of high friction, unlike thermostats of the Nosé-Hoover family whose performance degrades more gently in the strong coupling regime. We propose a simple and easy-to-implement modification to the integration scheme of the Langevin algorithm that addresses the fundamental source of the overdamped behaviour of high-friction Langevin dynamics: if the action of the thermostat causes the momentum of a particle to change direction, it is flipped back. This fast-forward Langevin equation preserves the momentum distribution, and so guarantees the correct equilibrium sampling. It mimics the quadratic behavior of Nosé-Hoover thermostats, and displays similarly good performance in the strong coupling limit. We test the efficiency of this scheme by applying it to a 1-dimensional harmonic oscillator, as well as to water and Lennard-Jones polymers. The sampling efficiency of the fast-forward Langevin equation thermostat, measured by the correlation time of relevant system variables, is at least as good as the traditional Langevin thermostat, and in the overdamped regime the fast-forward thermostat performs much better, improving the efficiency by an order of magnitude at the highest frictions we considered.

Accepted for publication by J. Chem. Phys.; 7 pages, 6 figures

References in corpus (8)

• Canonical sampling through velocity-rescaling
• Competing quantum effects in the dynamics of a flexible water model
• Accurate sampling using Langevin dynamics
• Efficient stochastic thermostatting of path integral molecular dynamics
• Colored-noise thermostats à la carte
• Accelerating the convergence of path integral dynamics with a generalized Langevin equation
• Stochastic thermostats: comparison of local and global schemes
• Fine Tuning Classical and Quantum Molecular Dynamics using a Generalized Langevin Equation

Cited by in corpus (6)

• i-PI 2.0: A Universal Force Engine for Advanced Molecular Simulations
• i-PI 3.0: a flexible and efficient framework for advanced atomistic simulations
• Nuclear Quantum Effects in liquid water at near classical computational cost using the adaptive Quantum Thermal Bath
• Hydrogen Bonding and Nuclear Quantum Effects in Clays
• Predicting the unobserved: a statistical mechanics framework for non-equilibrium material response with quantified uncertainty
• Faster Molecular Dynamics with Neural Network Potentials via Distilled Multiple Time-Stepping and Non-Conservative Forces