In developing the next generation of Calphad databases, new models are used in which each term contributing to the Gibbs energy has a physical meaning. To continue the development, finite temperature density-functional-theory (DFT) results are used in the present work to discuss and suggest the most applicable and physically based model for Calphad assessments of solid phases above the melting point (the breakpoint for modeling the solid phase in previous assessments). These results are applied to investigate the properties of a solid in the superheated temperature region and to replace the melting temperature as the breakpoint with a more physically based temperature, i.e., where the superheated solid collapses into the liquid. The advantages and limitations of such an approach are presented in terms of a new assessment for unary aluminum.

Aiming for better extrapolations and predictabilities of thermodynamic properties of materials, new thermodynamic models are implemented in the third‐generation Calphad databases. In these models, each term contributing to the Gibbs energy has an explicit physical meaning. Furthermore, descriptions of thermodynamic properties of materials are valid from 0 K up to high temperatures far above the melting point. As a starting point for the development of large self‐consistent third‐generation database, the new models in the present work are applied to the unary manganese system. Taking into account both the calculated first principles results and experimental data, thermodynamic model parameters are evaluated. Thermodynamic properties predicted using this description, agree very well with available data. The calculated properties vary smoothly in the whole temperature range, which is another important improvement compared to the second‐generation databases.

The main focus in developing the third generation of CALPHAD databases is to model thermodynamic properties of materials by using models which are more physically based and valid down to 0 K. First‐principles calculations are helpful to choose and validate those models. Reliable calculation results, for example, at very low temperatures or on metastable systems reveal physical facts which might be inaccessible by experiments. Following our earlier work for modeling thermodynamic properties of pure elements (i.e., Fe and Mn) in third‐generation CALPHAD databases, the ε (hcp) phase was modeled as a metastable phase in the present work. Although hcp phase is just observed in these two elements under ultra‐high pressure, in the binary Fe–Mn this phase is metastable at ambient temperatures and pressures. Therefore, it should be properly modeled in unaries for later optimization of binary systems. Based on density functional theory (DFT) calculations, the magnetic ground state and the magnetic properties of ε‐Fe, ε‐Mn, and their binary solution phase were calculated. It was found that ε‐Fe is anti‐ferromagnetic (type II) while ε‐Mn has a paramagnetic ground state. Accordingly, magnetic contributions to thermodynamic properties were accurately modeled. Moreover, by means of the extrapolation of experimental data for the thermodynamic properties of binary systems and high‐pressure data for unaries, the metastable hcp phases at ambient pressure were modeled for the third‐generation CALPHAD database, consistently with other stable phases in the elements Fe and Mn.

An updated thermodynamic description of pure Co was obtained by applying new models for the third generation of Calphad databases. In these models, different contributions to the heat capacity, especially the vibrational part, were treated separately, each with a clear physical meaning. More importantly, the phase stabilities of the various allotropes are now physically well defined. Thus, the derived thermodynamic properties vary more reasonably and smoothly from 0 K and up. Calculated thermodynamic properties were compared with experimental data and good agreement was obtained.

Thermodynamic descriptions in databases for applications in computational thermodynamics require representation of the Gibbs energy of stable as well as metastable phases of the pure elements as a basis to model multi-component systems. In the Calphad methodology these representations are usually based on physical models.Reasonable behavior of the thermodynamic properties of phases extrapolated far outside their stable ranges is necessary in order to avoid that they become stable just because these properties extrapolate badly. This paper proposes a method to prevent crystalline solid phases in multicomponent systems to become stable again when extrapolated to temperatures far above their melting temperature.