ICAMS / Interdisciplinary Centre for Advanced Materials Simulation


Towards a true predictive model database for the thermo-physical properties of Co-based alloys

Date: 23.07.2013
Place: International workshop on advanced cobalt-base superalloys, Erlangen, Germany

Suzana Fries
Mauro Palumbo
Abed Breidi

Gibbs energies model databases for multicomponent systems, traditionally, stand on experimental observation. Parameters of the models are optimized providing the best thermodynamic description for individual phases as well as the equilibrium between them. This allows to calculate phase diagrams and reproduce experimentally determined thermodynamic properties related to the Gibbs energies and its derivatives like enthalpies, heat capacities, chemical potentials, etc. The method known as CALPHAD uses these parametric models to calculate not only the already known information but also interpolated and extrapolated quantities. That is usually called “prediction”.

It is very recent that the quantum mechanics based density functional theory (DFT) got to the status of being able to calculate, from first-principles, quantities like enthalpies of formation, with an accuracy similar to the one obtained experimentally. An enthalpy of formation obtained by DFT is a true prediction as it is based on the solution of Schroedinger equation describing a quantum system constituted by atoms without to assign definite values, but predicting them using a probability distribution.

In this talk we will show how the predictive power of DFT can be coupled to the CALPHAD method allowing fundamental knowledge and understanding to support the creation and properties optimization of Co-based alloys.

« back