Computational materials design by thermodynamic modeling
Restricted (Penn State Only)
- Author:
- Sun, Hui
- Graduate Program:
- Materials Science and Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- February 15, 2024
- Committee Members:
- John Mauro, Program Head/Chair
Zi-Kui Liu, Co-Chair & Dissertation Advisor
Jingjing Li, Outside Unit & Field Member
Hojong Kim, Major Field Member
Allison Beese, Co-Chair & Dissertation Advisor - Keywords:
- CALPHAD
FGM
Thermodynamic modeling
Cr-Fe-Mo-Nb-Ni
Materials Design - Abstract:
- Joining dissimilar materials can be achieved by, for example, casting, welding, and additive manufacturing (AM). However, the formation of undesired phases and cracks in these processes can diminish or even destroy mechanical and other properties of the processed parts. Thermodynamic modeling using the CALculation of Phase Diagrams (CALPHAD) approach has demonstrated its efficacy in enabling accurate prediction of the formation of phases as well as the associated cracks in the processed parts. However, the CALPHAD approach is limited by the lack of a reliable thermodynamic database and the absence of reliable software tools to develop and use thermodynamic databases. The topologically close-packed (TCP) phases are major brittle phases formed during manufacturing process of Fe- and Ni-based superalloys. However, there are no reliable thermodynamic databases for TCP phases capable of describing their atomic behavior at each Wyckoff position. The pervious databases adopted the simplified sublattice models to describe TCP phases based on the combined Wyckoff positions due to the complex crystal structures of the TCP phases and the lack of input data for modeling. Furthermore, the reliability of the multicomponent thermodynamic database significantly depends on the accuracy of the low-order systems. For example, any inaccuracies in the binaries will be propagated to the ternary systems. Therefore, to obtain accurate modeling of TCP phases in multicomponent system, it is necessary to employ more reliable sublattice models for all related binary and ternary systems. With the reliable thermodynamic database, the other key challenge is to have proper tools to apply the database for materials design. One key capability is to predict the formation of phases with desired properties under various processing conditions in both welding and AM processes. In this dissertation, a framework is established in terms of the development of thermodynamic databases and the software tools for materials design. A thermodynamic database for the Cr-Fe-Mo-Nb-Ni system is constructed with the sublattice models for TCP phases according to their Wyckoff positions and the model parameters evaluated from input data from experimental measurements in the literature, DFT-based first-principles calculations, and machine learning models. This database is used for accurate predictions of formation of phase and cracks in functionally graded materials (FGMs) across Fe- and Ni-based alloys, including Ni20Cr (wt. %) to V FGM and stainless steel 304L (SS304L) to Inconel 625 (IN625) FGM. Furthermore, an open-source software tool, names MaterialsMap, is developed to design composition pathways between dissimilar materials through high-throughput equilibrium calculations and Scheil-Gulliver simulations. It is shown that MaterialsMap not only maximizes the exploration of phase stability through high-throughput calculations, but also expands its utility by implementing other mechanical materials properties.