Molecular Dynamics and Langevin Simulations of One-Dimensional Interfaces.
Hiroshi Koibuchi
Abstract
Open-access reader
Hiroshi Koibuchi
Abstract
Open-access reader
By making use of the techniques of molecular dynamics simulation and Langevin simulation, we study one-dimensional interfaces which separate two distinct fluids. In particular, we address the potential capabilities of the two techniques for simulating the canonical ensemble averages of one-dimensional interfaces. In the molecular dynamics simulations, a microcanonical ensemble average is simulated by introducing the canonical momentums of the molecules which make up the interface. In the Langevin simulations, we solve a discretized Langevin equation of the molecules, which are considered to be Brownian particles, by a Euler algorithm. We find that the two techniques and the Monte Carlo give the same results for internal energy and specific heat, and that there are systematic discrepancies between the molecular dynamics and the MC for the average square size and fractal dimensions. Our conclusion is that the Langevin technique can successfully simulate the canonical ensemble average of one-dimensional interfaces ; however, the molecular dynamics in this study cannot. We also examine a hybrid technique of molecular dynamics and Langevin simulations, and find that it yields the same results that the MC and Langevin simulations yield.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
By making use of the techniques of molecular dynamics simulation and Langevin simulation, we study one-dimensional interfaces which separate two distinct fluids. In particular, we address the potential capabilities of the two techniques for simulating the canonical ensemble averages of one-dimensional interfaces. In the molecular dynamics simulations, a microcanonical ensemble average is simulated by introducing the canonical momentums of the molecules which make up the interface. In the Langevin simulations, we solve a discretized Langevin equation of the molecules, which are considered to be Brownian particles, by a Euler algorithm. We find that the two techniques and the Monte Carlo give the same results for internal energy and specific heat, and that there are systematic discrepancies between the molecular dynamics and the MC for the average square size and fractal dimensions. Our conclusion is that the Langevin technique can successfully simulate the canonical ensemble average of one-dimensional interfaces ; however, the molecular dynamics in this study cannot. We also examine a hybrid technique of molecular dynamics and Langevin simulations, and find that it yields the same results that the MC and Langevin simulations yield.
Key concepts: Brownian dynamics, Langevin dynamics, Statistical physics, Molecular dynamics, Canonical ensemble, Langevin equation, Microcanonical ensemble, Monte Carlo method