UPDATED SUBLATTICE MODELS OF TOPOLOGICALLY CLOSE PACKED PHASES WITH A REVISED PHASE DESCRIPTION OF σ PHASE
Open Access
- Author:
- Feurer, Matthew
- Graduate Program:
- Materials Science and Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- October 17, 2019
- Committee Members:
- Zi-Kui Liu, Thesis Advisor/Co-Advisor
Ismaila Dabo, Committee Member
Allison Michelle Beese, Thesis Advisor/Co-Advisor
Susan B Sinnott, Program Head/Chair - Keywords:
- CALPHAD
First Principles
Topologically close packed phases
TCP phases
Sigma phase - Abstract:
- Frank Kasper Topologically Close Packed (TCP) phases are a wide class of intermetallic phases known to occur in a wide variety of metallic alloy systems, often occurring where elements have different radii and electronic properties. They are known as detrimental phases in alloys, occurring in many of the most technologically significant alloys. The formation of these phases often leading to poor mechanical properties and detriment to corrosion properties. There has been great progress since Kasper in understanding the structure and occurrence of TCP phases, however modeling of these phases has still been a difficult task using the CALculation of PHAse Diagram (CALPHAD) method. High solubility, magnetism, and complex structure all provide additional challenges in modeling these phases within the context of CALPHAD. Such challenges have led to the adoption of multiple models for describing a single phase, and has been an impediment to the development of higher order databases. In many situations it is the choice of the modeler to either retain compatibility or to increase physical accuracy. In the following work the criteria for sublattice models will be discussed and identified. Following this the sublattice models of common TCP phases will be reviewed with the previously identified criteria. Development of an updated sublattice model with increased physical accuracy will be demonstrated for the sigma phase. Discussion of how High throughput methods can be incorporated into the creation of CALPHAD models (density functional theory, optimization and machine learning methods) will also be included. Finally revised sublattice models will be presented with the intent for future modelers to coordinate their efforts between disparate and higher order systems.