THERMOPHYSICAL PROPERTY PREDICTIONS FROM THE SOFTWARE PACKAGE EOSMIX

A two-fluid corresponding-states model for predicting the thermodynami c properties of mixtures has been developed, which uses accurate equat ions of state for the reference substances of the components. A new mi xing parameter has been introduced, which takes account of the differe nces in molecular sizes in binary fluid mixtures. The three mixing par ameters for methane-ethane and methane-propane have been optimized usi ng different types of experimental data. Comparisons of predictions wi th data are given for experimental densities and speed of sound.