Both the Laves phases and the m-phases play an important role in the development of new ferritic steels for high-temperature applications. In the present work an improved description of both phases is used in the assessment of the systems Fe-Mo, Fe-W and Fe-Ta in the present work. The thermodynamic parameters are assessed by the CALPHAD method using results of ab-initio calculations.
Previously the Laves phase was treated as a line compound (A)2(B). However, experimental investigations on various systems containing the Laves phase C14 (Fe-Ta, Fe-Nb, Fe-Ti, Fe-W, Fe-Mo…) showed that the Laves phase exhibits a solubility range which cannot be neglected. Instead of a stoichiometric compound, the new description (A,B)2(A,B) allows for full miscibility of A and B elements on the sublattices of the phase.
The m-phase was previuously described with a three sublattice model (A)7(B)2(A,B)4. This gives a maximum solubility of B in the the m-phase of 46.15 at% B. The existence domain of the stable m-phase in real systems ranges from approximately 40 (Fe-W, Fe-Mo) to 80 at% B (Mo-Si). This was the reason of using the new model proposed by Joubert, (A,B)1(A,B)2(A,B)6(B)4 which allows the m-phase to be stable up to 100 at% B.
Total energies of formation of the Laves phase structure with various occupations on the sublattices (A2A, A2B, B2A, B2B) are important input values into the above-mentioned model. These values were calculated by means of the pseudopotential VASP code using the PAW-PBE (Projector Augmented Wave – Perdew-Burke-Ernzerhof) pseudopotential. The full relaxation of Laves and SER (standard element reference) structures