Optimization of databases for thermo-kinetic precipitation simulations
Project manager: Erwin
The current work in this project focuses on the development, optimization and validation of the thermodynamic, diffusion and physical databases for A) Fe-based steels, B) Ni-based superalloys and C) refractory Mo-based alloys in the MatCalc software .
Thermodynamic databases have been developed, which are adapted for optimum applicability in thermo-kinetic precipitation simulations, in order to improve predictions of particle sizes and number densities, as well as mechanical properties associated with precipitation. Recently, besides several improvements of our most advanced in-house database, mc_fe (mc stands for MatCalc) version 1.001, the new thermodynamic databases mc_ni version 1.007 and mc_mo version 0.003 have been created by editing data from the literature and optimization of thermodynamic model parameters.
The thermodynamic databases contain the optimized Gibbs energy functions of stoichiometric phases as a function of temperature, and for solid solution phases as a function of temperature and composition . Elements and phases of particular interest regarding the precipitation hardening in metals are implemented in the thermodynamic databases using the CALPHAD (Calculation of phase diagrams) approach. The optimization of new model parameters is based on the accurate assessment of thermodynamic and phase diagram data.
A) Steel database mc_fe
Adaptions of mc_fe concern refinements of element contents of austenite as a function of temperature for a various steel grades, assessment of carbide stabilities, as well as thermodynamic modeling of ordered phases.
B) Ni-base mc_ni
Based on our well established mc (MatCalc) steel database, the fcc (face-centered cubic)-structured nickel matrix phase has been completed by including parameters of Ni-containing subsystems. The focus of modeling has been put on the precipitate phases γ´(cubic fcc-type ordered Ni3X, major X=Ti, Al), δ (orthorhombic Ni3X, major X=Nb), and γ´´ (tetragonally distorted Ni3X, major X=Nb). The thermodynamic model parameters of these phases have been partly adopted from existing thermodynamic assessments, and partly optimized based on experimental equilibrium or close-to-equilibrium phase fractions and phase compositions from the literature, for alloys inside the sub-database Ni-Al-Nb-Ta-Ti-Cr-Mo-Co-W. The number of interaction parameters in addition to end-member Gibbs energies has been kept as small as possible. This means that the potential of future optimizations by defining further interaction parameters is held open.
High consistency of predicted equilibrium phase fractions and compositions with experimental data has been tested in a wide range of superalloy compositions.
C) Mo-base mc_mo
The ternary Mo-Hf-C system has been assessed in CDLESOP. Binary descriptions have been adopted from the open literature, and the self-diffusion data of Mo, as well as tracer diffusivities of C and Hf in Mo have been assessed from experimental data. New model parameters of most important carbide phases, Mo2C and HfC have been optimized with experimental phase diagram data. The resulting MatCalc databases, thermodynamic mc_mo.tdb and diffusion database mc_mo.ddb are prepared to be used in thermo-kinetic precipitation simulations for Hf- and C-doped Molybdenum-base alloys.
Calculated isothermal thermodynamic equilibrium sections of the Mo-Hf-C system are compared with experimental phase diagrams in the following figure.
Figure: Calculated isothermal sections of the Mo-Hf-C system (a) at 1400°C and (b) at 2000°C with mc_mo (left side)
and experimental phase diagrams from the open literature (right side). Red lines limit 3-phase fields,
greem conodes connect compositions of phases in 2-phase equilibria. Black curves border single-phase regions.
D) Development of mobility databases
The diffusion databases contain self diffusion and tracer diffusion data for all the elements that are included in the thermodynamic databases.
E) Development of physical databases
The physical databases contain optimized polynomials that describe the densities of phases as a function of temperature. The optimization of new model parameters is based on the accurate assessment of experimental data of thermal expansions, molar volumes and densities. Thus calculations of volume changes during phase transformations are possible, and phase fractions can be given as volume fractions, which can be directly compared with results from metallographic investigations. Also, lattice mismatch can be calculated in kinetic simulations.