Answers are grounded in this paper and cite their sources.
We apply the so-called ``synthetic'' nonequilibrium molecular-dynamics method to the calculation of the self-diffusion constant of a Lennard-Jones fluid at a number density of 0.85/σ³ and a temperature of 1.08{ε}/kB (where {ε} and {σ} are the energy and length parameters of the potential and kB is the Boltzmann constant). By comparing with the Green-Kubo calculation for the same state of the system and for the same number of particles, N, we find the latter calculation to yield more precise values of the self-diffusion constant for a given number of molecular-dynamics time steps. Even at small values of the diffusion current, a nontrivial time is needed for the nonequilibrium calculation to reach the steady state. For larger values of the driving force, the steady-state flow appears to become unstable and evidence of a secondary flow pattern is presented. The presence of these instabilities acts as a limit to the range of the driving force for which the steady-state method can be applied. With increasing N the range of stable values of the diffusion current density decreases. For the Green-Kubo calculations, the N dependence of the self-diffusion constant is found to be anomalous for N=108, with the 1/N dependence only exhibited for at least 500 particles. The nonequilibrium results, while approximately independent of N for 108 and 500 particles, are found to have a similar anomalous N dependence when we extend the calculations to 1372 particles, thereby bringing the Green-Kubo and nonequilibrium results into agreement in the large-system limit.
No takes yet. Share an insight, caveat, or question.
Jerome J. Erpenbeck (1987) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: