Optimization and calculation of thermodynamic properties and phase diagrams of multicomponent oxide systems

A thermodynamic database for oxide systems has been successfully developed. As a result, all thermodynamic as well as phase diagram data of an oxide system are represented and stored by means of a small set of coefficients. Consequently, multi-component multiphase equilibria involving oxides with other phases such as gas, metals etc. can be predicted by a Gibbs energy minimization program using the present database.;To build a database, simple models such as the Raoultian etc. have been employed for solid oxide solutions. For liquid oxide solutions, as discussed in Chapter 1 (where some practical solution models are reviewed), the modified quasi-chemical model has been chosen to build the present thermodynamic oxide database.;According to the modified quasi-chemical model, thermodynamic properties are non-linear functions of the model coefficients. Hence, non-linear least squares techniques have been developed and a binary optimization program has been written accordingly. All the evaluated liquid and solid thermodynamic properties and the phase diagram are internally consistent with each other. This thermodynamic internal consistency has been demonstrated in Chapter 3 when evaluating a series of rare earth oxide-Al;The modified quasi-chemical model has been successfully extrapolated in ternary and multi-component systems. In Chapter 4, two improvements have been made to the ternary optimization program developed previously.;In Section 4.2 and Chapter 5, successful examples of predicting multi-component and multiphase equilibria are provided.;The components considered in the present study are SiO...