【摘要】 在经典的Hall-Petch关系中，多晶金属的强度通常随晶粒尺寸的减小而提高。然而，当晶粒尺寸极小时（通常小于10～20 nm），强度却降低（发生软化），此时Hall-Petch关系失效。Hall-Petch效应的失效表明，纳米晶粒金属的强化不仅受尺寸的影响，还受晶界（GB）稳定性的影响。本研究使用LAMMPS构建了Cu-Ni纳米晶模型，采用混合蒙特卡洛/分子动力学（MC/MD）的手段对纳米晶Cu-Ni合金的晶界偏析的不同情况进行了分子模拟。研究发现，晶界偏析的存在会抑制晶粒软化，晶粒尺寸效应不再主导纳米材料的力学性能（逆Hall-Petch效应），溶质偏析也不再对纳米级别的合金有明显的强化作用。在综合分析了晶界能量和位错后，可得出结论：通过抑制固溶体的形成，促进晶界偏析，同时优化溶质原子的浓度，将获得理想的最大强度模型。当纳米晶粒金属的均匀偏析度为约0.9时，该材料获得最强的力学性能。通过分子动力学模拟，从位错结合GB能的角度对偏析进行了全方位的深入研究。更多还原

【Abstract】 In the classical Hall-Petch relationship, the strength of polycrystalline metals usually increases with decreasing grain size. However, the strength decreases(softening occurs) when the grain size is extremely small(usually less than 10 to 20 nm), at which point the Hall-Petch relation fails. The failure of the Hall-Petch effect suggests that the strengthening of nanocrystalline metals is not only affected by the size, but also by the stability of the grain boundary(GB). In this paper, a Cu-Ni nanocrystalline model was constructed using LAMMPS, and molecular simulations of different scenarios of grain boundary segregation of nanocrystalline Cu-Ni alloys were carried out by means of hybrid Monte Carlo/molecular dynamics(MC/MD). It is found that the presence of grain boundary segregation inhibits grain softening, the grain size effect no longer dominates the mechanical properties of nanomaterials(inverse Hall-Petch effect), and solute segregation no longer has a significant strengthening effect on the nanoscale alloys. After a comprehensive analysis of grain boundary energies and dislocations, it can be concluded that the ideal maximum strength model is obtained by inhibiting solid solution formation and promoting grain boundary polarization while optimizing the concentration of solute atoms. The strongest mechanical properties of the material are obtained when the uniform degree of segregation of the nanograin metal is about 0.9. The MD simulations provide the first comprehensive and in-depth study of polarization from the perspective of dislocation binding GB energy.更多还原