Transient Nonequilibrium Molecular Dynamic Simulations of Thermal Conductivity: 1. Simple Fluids

Thermal conductivity has been previously obtained from molecular dynamics (MD) simulations using either equilibrium (EMD) simulations (from Green--Kubo equations) or from steady-state nonequilibrium (NEMD) simulations. In the case of NEMD, either boundary-driven steady states are simulated or constrained equations of motion are used to obtain steady-state heat transfer rates. Like their experimental counterparts, these nonequilibrium steady-state methods are time consuming and may have convection problems. Here we report a new transient method developed to provide accurate thermal conductivity predictions from MD simulations. In the proposed MD method, molecules that lie within a specified volume are instantaneously heated. The temperature decay of the system of molecules inside the heated volume is compared to the solution of the transient energy equation, and the thermal diffusivity is regressed. Since the density of the fluid is set in the simulation, only the isochoric heat capacity is needed in order to obtain the thermal conductivity. In this study the isochoric heat capacity is determined from energy fluctuations within the simulated fluid. The method is valid in the liquid, vapor, and critical regions. Simulated values for the thermal conductivity of a Lennard-Jones (LJ) fluid were obtained using this new method over a temperature range of 90 to 900 K and a density range of 1–35 kmol · m-3. These values compare favorably with experimental values for argon. The new method has a precision of ±10%. Compared to other methods, the algorithm is quick, easy to code, and applicable to small systems, making the simulations very efficient.

[1]  C. Hoheisel.  Memory functions and the calculation of dynamical properties of atomic liquids , 1990 .

[2]  D. Heyes.  Transport coefficients of simple fluids with steeply repulsive potentials , 1994 .

[3]  Massobrio,et al.  Nonlinear thermal response of a Lennard-Jones fluid near the triple point. , 1986, Physical review. A, General physics.

[4]  D. Heyes.  Thermal conductivity and bulk viscosity of simple fluids. A molecular-dynamics study , 1984 .

[5]  W. M. Haynes,et al.  The Viscosity and Thermal Conductivity Coefficients for Dense Gaseous and Liquid Argon, Krypton, Xenon, Nitrogen, and Oxygen , 1974 .

[6]  John Erjavec,et al.  Modern Statistics for Engineering and Quality Improvement , 2000 .

[7]  Denis J. Evans,et al.  Homogeneous NEMD algorithm for thermal conductivity—Application of non-canonical linear response theory , 1982 .

[8]  K. Tankeshwar,et al.  Transport coefficients of classical dense fluids: a simple approach , 1995 .

[9]  J. Sengers,et al.  The thermal conductivity of argon at elevated densities , 1963 .

[10]  F. Müller-Plathe.  A simple nonequilibrium molecular dynamics method for calculating the thermal conductivity , 1997 .

[11]  D. Bedrov,et al.  Thermal conductivity of molecular fluids from molecular dynamics simulations: Application of a new imposed-flux method , 2000 .

[12]  R. Reid,et al.  The Properties of Gases and Liquids , 1977 .

[13]  M J Gillan,et al.  The calculation of thermal conductivities by perturbed molecular dynamics simulation , 1983 .

[14]  G. Ciccotti,et al.  Thermal conductivity of the Lennard‐Jones liquid by molecular dynamics calculations , 1987 .

[15]  B. L. Neindre.  Contribution a l'etude experimentale de la conductivite thermique de quelques fluides a haute temperature et a haute pression , 1972 .

[16]  T. Ikeshoji,et al.  Non-equilibrium molecular dynamics calculation of heat conduction in liquid and through liquid-gas interface , 1994 .

[17]  M. Luckas,et al.  Shear viscosity and thermal conductivity of the Lennard-Jones liquid computed using molecular dynamics and predicted by a memory function model for a large number of states , 1988 .