Ocean of Data: Integrating First-Principles Calculations and CALPHAD Modeling with Machine Learning
- Pennsylvania State Univ., University Park, PA (United States). Dept. of Materials Science and Engineering
Thermodynamics is a science concerning the state of a system, whether it is stable, metastable or unstable, when interacting with the surroundings. Computational thermodynamics enables quantitative calculations of thermodynamic properties as a function of both external conditions and internal configurations, in terms of first and second derivatives of energy with respect to either potentials or molar quantities. Thermodynamic modeling based on the CALPHAD method enables the thermodynamics beyond stable states and is the foundation of Materials Genome and materials design. In last several decades, first-principles calculations based on density functional theory have provided invaluable thermochemical data to improve the robustness of CALPHAD modeling. Today with ever increasing computing power and large amount of data repositories online, it calls for a new paradigm for CALPHAD modeling approach incorporating machine learning to create a sustainable ecosystem for the ocean of data and for emergent behaviors.
- Research Organization:
- Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States). National Energy Research Scientific Computing Center (NERSC); Univ. of California, Oakland, CA (United States)
- Sponsoring Organization:
- USDOE Office of Science (SC)
- DOE Contract Number:
- AC02-05CH11231
- OSTI ID:
- 1543497
- Journal Information:
- Journal of Phase Equilibria and Diffusion, Vol. 39, Issue 5; ISSN 1547-7037
- Publisher:
- ASM International
- Country of Publication:
- United States
- Language:
- English
Similar Records
CALPHAD in the Cloud
Machine learning the metastable phase diagram of covalently bonded carbon