Font Size: a A A

Monte Carlo simulation of mixture phase behavior: Methods and applications

Posted on:2000-08-09Degree:Ph.DType:Dissertation
University:Cornell UniversityCandidate:Potoff, Jeffrey JosephFull Text:PDF
GTID:1460390014465716Subject:Engineering
Abstract/Summary:
This work focuses on implementation of new Monte Carlo methodologies for the study of liquid-vapor phase transitions of mixtures. Extensions are proposed to histogram-reweighting and mixed-field finite-size scaling methods that allow for the calculation of phase coexistence and critical points for binary mixtures. These extensions are combined with grand canonical Monte Carlo simulations to obtain liquid-vapor coexistence curves and critical points of the pure fluid and a binary mixture of Lennard-Jones particles. The results of this work have much smaller statistical uncertainties relative to comparable Gibbs ensemble simulations.; The study of Lennard-Jones systems continues with the calculation of surface tensions for a wide range of temperatures with a finite-size scaling methodology proposed by Binder [Phys. Rev. A 25, 1699 (1982)]. The surface tension calculated by the method of Binder yields behavior that is consistent with the expected critical behavior, while the surface tensions predicted for temperatures less than 0.7 T*c are substantially higher than values found in the literature.; Grand canonical histogram-reweighting Monte Carlo simulations are combined with mixed-field finite-size scaling techniques to locate the critical points of Stockmayer fluids for a wide range of reduced dipole moments. The phase behavior and critical phenomena of three Stockmayer/Lennard-Jones mixtures is studied. The results of this work show that errors exist in the near critical simulation data of previous Gibbs ensemble calculations.; Finally, we investigate the predictive ability of several recently developed intermolecular potential models that reproduce accurately the phase behavior of pure components. These potentials are of the united-atom type and utilize the exponential-6 functional form for repulsive and dispersion interactions, and fixed point charges for electrostatic interactions. The mixtures studied are n-pentane/methane, ethane/CO2, propane/CO 2, n-pentane/CO2, H2O/ethane, CH3OH/n-hexane and CH3OH/CO2. Our results suggest that relatively simple intermolecular potential models can be used to predict the phase behavior of broad classes of binary systems.
Keywords/Search Tags:Phase, Monte carlo, Mixtures
Related items