Improved computational method to generate properly equilibrated atomistic microstructures

Thumbnail Image
Journal Title
Journal ISSN
Volume Title

Atomistic simulations play an important role in unravelling the fundamental behavior of nanocrystalline (NC) metals/alloys. To ensure the validity of the simulated results, the initial NC structures must be representative of a real material to the extent possible. Using proper equilibration techniques, it must also be ensured that these NC structures reach a state of metastable equilibrium before probing their response. To this effect, the influence of simulated thermal equilibration of atomistic NC Ni structures on the resulting mechanical behavior is discussed in this work. It is shown that the well-equilibrated NC structures become stiffer in terms of both elastic response and yielding behavior and accumulate less residual strain upon unloading, thus, signifying the importance of proper equilibration. However, it is found that the regular equilibration method of thermal relaxations at 300 K, typically employed in atomistic modeling studies, takes significantly longer to drive the NC structures towards a metastable equilibrium state. Finally, an improved two-step equilibration method is presented that drastically expedites the equilibration process while resulting in the structural and mechanical properties comparable with the regular equilibration method performed for significantly larger simulation times. The major modification in the improved method involves: Subjecting only the grain boundary and the surrounding atoms to thermal relaxations at relatively higher temperature. (C) 2021 The Authors. Published by Elsevier B.V.

Thermal equilibration, Atomistic simulations, Nanocrystalline structure, NANOPHASE NI, GRAIN-GROWTH, DEFORMATION, SIMULATIONS, BEHAVIOR, Multidisciplinary Sciences
Gupta, A., Rajaram, S. S., Thompson, G. B., & Tucker, G. J. (2021). Improved computational method to generate properly equilibrated atomistic microstructures. In MethodsX (Vol. 8, p. 101217). Elsevier BV.