February 27, 2017
Journal Article

Long-time Atomistic Dynamics Through a New Self-Adaptive Accelerated Molecular Dynamics Method

Abstract

A self-adaptive accelerated molecular dynamics method is developed to model infrequent atomic- scale events, especially those events that occur on a rugged free-energy surface. Key in the new development is the use of the total displacement of the system at a given temperature to construct a boost-potential, which is slowly increased to accelerate the dynamics. The temperature is slowly increased to accelerate the dynamics. By allowing the system to evolve from one steady-state con guration to another by overcoming the transition state, this self-evolving approach makes it possible to explore the coupled motion of species that migrate on vastly different time scales. The migrations of single vacancy (V) and small He-V clusters, and the growth of nano-sized He-V clusters in Fe for times in the order of seconds are studied by this new method. An interstitial- assisted mechanism is rst explored for the migration of a helium-rich He-V cluster, while a new two-component Ostwald ripening mechanism is suggested for He-V cluster growth.

Revised: April 19, 2017 | Published: February 27, 2017

Citation

Gao N., L. Yang, F. Gao, R.J. Kurtz, D. West, and S. Zhang. 2017. Long-time Atomistic Dynamics Through a New Self-Adaptive Accelerated Molecular Dynamics Method. Journal of Physics: Condensed Matter 29, no. 14:Article No. 145201. PNNL-SA-112547. doi:10.1088/1361-648X/aa574b