**About this Abstract** |

**Meeting** |
**2018 TMS Annual Meeting & Exhibition
** |

**Symposium
** |
**Hume-Rothery Award Symposium: Computational Thermodynamics and Its Implications to Kinetics, Properties, and Materials Design
** |

**Presentation Title** |
Computational Thermodynamics: Humans and Machines |

**Author(s)** |
Marius Stan |

**On-Site Speaker (Planned)** |
Marius Stan |

**Abstract Scope** |
Equilibrium and non-equilibrium thermodynamic properties of multi-component materials are important for understanding and predicting material evolution during synthesis, manufacturing and operation. A comprehensive evaluation of thermodynamics requires experimental, theoretical and computational methods that operate at length and time scales from nanometers to meters and from picoseconds to days. In many cases, it is necessary to couple computational methods – as demonstrated by the work of Prof. Zi-Kui Liu on coupling density functional theory and CALPHAD. Computers can expand the investigation domain beyond the abilities of the human brain. In this talk we discuss evaluation of thermodynamic data and phase equilibria using atomistic, meso-scale and continuum methods. Data analysis is enhanced by machine learning using a Bayesian method that simultaneously accounts for different types of data and its provenance and delivers uncertainty intervals. The presentation ends with a discussion of the changing role of computation in science and engineering. |

**Proceedings Inclusion?** |
Planned: Supplemental Proceedings volume |