It is speculated that the initial slip in the single-crystal nanowire (NW) is dominant to the final breaking position. To reveal the correlation between the initial slip and breaking position, in this paper, up to 300 independent [111] oriented single-crystal copper NWs under uniaxial tension at a temperature of 10 K have been performed using molecular dynamics (MD) simulations. Combining with Machine Learning, the position of the initial slip is determined by the density-based spatial clustering of applications with noise (DBSCAN) algorithm. The breaking position is calculated using the atom number ratio (ANR). The calculations illustrate that the final breaking positions follow two Gaussian distributions at two ends of the NW. Similarly, the distribution of the initial slip positions also has statistical peaks at the two ends and an additional distribution at the center of the NW within a large span. The results demonstrated that there is a clear correlation between the initial slips and the breaking positions on the basis of the distribution features. Further, four representative samples were studied systematically from the aspects of the microstructure and the stress–strain properties, showing that the final breaking position is determined at the middle and late stages of the plastic deformation. It infers that the apparent correlation between the initial slip positions and the final breaking positions of NWs exhibits statistically rather than causally. The conclusion derived from this study may contribute to promoting the fundamental research of metal nanomaterials and the optimization of the nanodevices.
Read full abstract