Inverse Hall–Petch effect in nanocrystalline ice predicted by machine-learned coarse-grained molecular simulations

Abstract

We study the grain-size effect on mechanical behaviors of polycrystalline ice at nanoscale through large-scale molecular dynamics simulations, enabled by an accurate and efficient machine-learned coarse-grained water model [H. Chan, M. J. Cherukara, B. Narayanan, T. D. Loeffler, C. Benmore, S. K. Gray and S. K. Sankaranarayanan. Machine learning coarse grained models for water. Nature Communications 10(1), p. 379, 2019]. Polycrystalline ice systems with different grain sizes in the range of [Formula: see text][Formula: see text]nm are systematically investigated through uniaxial tensile tests. Simulation results reveal that Young’s modulus and yield strength increase as the grain size increases, leading to an inverse Hall–Petch effect in nanocrystalline ice. Such an observation is significantly different from the conventional Hall–Petch effect in polycrystalline ice observed by experiments at millimeter scale. The deformation behavior suggests that grain boundaries (GBs), rather than grain interiors, play the active role in accommodating external loading in nanocrystalline ice. Further void analysis of nanocrystalline ice during deformation reveals that nanocrack initiates and propagates along GBs and eventually leads to failure of systems with larger grain sizes ([Formula: see text][Formula: see text]nm), yielding a sudden drop in the stress–strain curve. While for systems with smaller grain sizes ([Formula: see text][Formula: see text]nm), only extensive plastic deformation is presented without significant void growth. Our simulation results demonstrate that the GB sliding is the governing mechanism for the inverse Hall–Petch effect observed in nanocrystalline ice.