Isobaric versus canonical ensemble formalism for monte carlo studies of liquids

The Metropolis Monte Carlo method is one of two main approaches to computer simulation of liquid properties. It has virtually always been employed within the canonical ensemble formalism. By including density as one of the variables for the random walk, the Metropolis method becomes applicable to the analysis of a closed, isothermal, isobaric system (isobaric ensemble conditions). The analysis is directed toward equilibrium properties of classical models of dense polyatomic liquids such as water. Density, compressibility, constant pressure heat capacity, enthalpy and coefficient of thermal expansion are obtained directly in terms of mean values and variances of a two-dimensional distribution, _g(U, V) of random walk steps. The method appears to be well suited to the study of liquids in the vicinity of the triple point.