|About this Abstract
||2018 TMS Annual Meeting & Exhibition
||Computational Thermodynamics and Kinetics
||Efficient and Accurate Computation of Melting Temperatures and Enthalpies and Entropies of Fusion from Ab Initio
||Blazej Grabowski, Li-Fang Zhu, Jörg Neugebauer
|On-Site Speaker (Planned)
Accurate knowledge of melting properties is of high relevance to materials design. Such information is in principle available from phase diagrams, but standard thermodynamic parameterizations rely on input data which is often experimentally unavailable (metastable phases). First principles calculations are an attractive alternative, but typically require large computational resources. We present here a new approach, the two optimized references thermodynamic integration using Langevin dynamics (TOR-TILD) method, that allows one to efficiently yet accurately compute melting properties from ab initio. In particular, our method enables a highly efficient computation of the full Gibbs energy of the liquid. We demonstrate the efficiency of the method for Cu and analyze the accuracy of standard density functional theory (LDA/GGA) in predicting melting properties, specifically the melting temperature, enthalpy and entropy of fusion.
||Planned: Supplemental Proceedings volume