Place: Bratislava Slovakia Annual Joint Working Group Meeting, Bratislava, Slovakia
Thermodynamic and kinetic descriptions created by the CALPHAD method are based in experimental data characterizing equilibrium states. The method can also be used for describing metastable states by extrapolation of the obtained equilibrium parametric functions to metastable regions. These extrapolations, however, can present some model artifacts and should be done with care. Many microstructure simulations done nowadays using phase field methods are coupled to the thermodynamic descriptions provided by the CALPHAD method (CALPHAD Gibbs energies databases). There are several quantities that can be exported from these databases, like phase diagram slopes, chemical potentials, thermodynamic factors, enthalpies of fusion, solid and liquid fractions, microsegregation during solidification, diffusivities, thermal expansion, etc. In many cases that quantities are calculated in metastable situations very far from the stable ones. This requires descriptions which are also evaluated in the metastable regions and which are, as much as possible, free of model artifacts. One way to obtain robust metastable description is to use information provide by first- principles methods wish can supply the configurational energetics and the topology of phase stability as function of compositon. The so called bcc, fcc and hcp phase diagrams showing order-disorder based on the same lattice, obtained, for example, from the Cluster Variation Method, can be incorporated into the CALPHAD description bringing confidence for use of the database in metastable regions. Stoichiometric description of a phase give a very poor contribution to phase-field simulations. The method mention above helps in the modelling of phases that experimentally appear as nearly stoichiometric as it provides information for modeling the phase's gibbs energy as function of composition which can bring a more significant contribution to the simulations. Another aspect that should be taken into account in coupling CALPHAD and phase-field methods is that even in the cases that calculations of equilibrium between phases are not required in the simulation and only single phase quantities are need, they cannot always be obtained directly from the analytical expressions. For example, for slopes of the Gibbs energy of phase for which sublattices are taken into account, a minimization is required as there are internal degrees of freedom. A list of tests to be done when evaluating a database for use in simulations will be provided in this presentation.