I.2 - Models and data

Abstract

The assessment of the thermodynamic properties of individual phases, such as their Gibbs energy as a function of temperature, composition, and possibly pressure is the basis for the successful establishment of a thermodynamic databank. Gibbs energy data have to be made available for the phases with a wide variety of properties. This chapter provides an overview of the models used in the assessment of the Gibbs energies of such phases. Once the Gibbs energies are known, all other thermodynamic properties can be derived. The information about the thermodynamic properties of phases is formed into a physical picture or model, such as oscillating atoms on lattice sites with a restricted number of degrees of freedom to describe the maximum value of the heat capacity. To be able to quantify the measured properties, the next step is to put the physical model into a mathematical form, which can subsequently be used for making interpolations or even predictions. Before accepting the prediction given by a model, it is important to test the model by comparing a number of the predictions with experimental information already available or obtained by new experiments. Even after a successful passing of such a test, it must be realized that the physical model behind the mathematical description might not give a correct picture of the real world. The mathematical description may thus result in incorrect predictions outside the range in which it has been tested.