Surface Tension Evaluation Via Thermodynamic Analysis of Statistical Data From Molecular Dynamics Simulations

Volume 4 ◽  
2004 ◽  
Author(s):  
Aaron P. Wemhoff ◽  
Van P. Carey

Surface tension determination of liquid-vapor interfaces of polyatomic fluids using traditional methods has shown to be difficult due to the requirement of evaluating complex intermolecular potentials. However, analytical techniques have recently been developed that determine surface tension solely by means of the characteristics of the interfacial region between the bulk liquid and vapor regions. A post-simulation application of the excess free energy density integration (EFEDI) method was used for analysis of the resultant density profile of molecular dynamics (MD) simulations of argon using a simple Lennard-Jones potential and diatomic nitrogen using a two-center Lennard-Jones potential. MD simulations were also run for an approximation of nitrogen using the simple Lennard-Jones potential. In each MD simulation, a liquid film was initialized between vapor regions and NVE-type simulations were run to equilibrium. The simulation domain was divided into bins across the interfacial region for fluid density collection, and the resultant interfacial region density profile was used for surface tension evaluation. Application of the EFEDI method to these MD simulation results exhibited good approximations to surface tension as a function of temperature for both a simple and complex potential.

2013 ◽  
Vol 845 ◽  
pp. 345-349
Author(s):  
I Ling Chang ◽  
Ming Liang Liao ◽  
Chi Hsiang Chuang

This paper used molecular dynamics (MD) simulations to investigate influences of cone arrangements (including the cone orientation, arrangement pattern and cone spacing) on hydrogen adsorption of open-tip carbon nanocone (CNC) arrays at temperatures of 100 and 300 K. To consider curvature effects for the cone structure of the CNCs, the curvature-modified Lennard-Jones potential parameters were adopted to describe the interactions between the hydrogen and carbon atoms. It was found that the cone orientation (aligned, opposite, and alternate) does not have obvious influences on hydrogen adsorption of the CNC arrays. The arrangement pattern (square and triangular), however, had significant influences on the hydrogen adsorption. The square-patterned CNC array was noticed to have higher storage weight percentage than the triangular-patterned one. Regarding to the influences of cone spacing, the storage weight percentage grew with the increase of the cone spacing and arrived at a stable value as the cone spacing reached a certain critical value. The influences cone arrangements could be ascribed to repulsive effects, which are evident as cone spacing become narrow.


Author(s):  
Van P. Carey

Nanoscale aspects of interfacial phenomena can be critically import in convective vaporization and condensation in nanochannels or microchannels. Molecular dynamics (MD) simulations have been extensively used to model and explore the physics of interfacial phenomena at the molecular level. Efforts to improve MD simulations have often focused on development of more physically realistic interaction potentials used to model intermolecular force interactions, or on development of more efficient computing strategies. An important, and often overlooked aspect of MD simulations is the role that theoretical models from statistical thermodynamics can play in MD simulations. This paper argues that use of alternate statistical thermodynamics models, and unconventional strategies for using them, can be effective ways of enhancing MD simulations. The advantages of these types of approaches are explored in the context of three recent MD simulation studies of interfacial region thermophysics that have made use of statistical thermodynamics theory in novel ways. Examples considered include studies of the interfacial region between bulk liquid and vapor phases, thin liquid films on solid surfaces, and stability free thin liquid films. These examples illustrate ways that MD simulations can be combined with other models to enhance computational efficiency or extract more information from the MD simulation results. Successful strategies for implementing these types of scheme are examined, and their general applicability is assessed.


2011 ◽  
Vol 8 (1) ◽  
pp. 182-188
Author(s):  
D.F. Marin

The paper presents results on performance and efficiency of GPU utilization in a simulation of molecular dynamics processes. The simulation was done with the usage of Lennard-Jones potential and leapfrog computational scheme.


Sign in / Sign up

Export Citation Format

Share Document