Unraveling Mg < c plus a > slip using neural network potential

PHILOSOPHICAL MAGAZINE(2022)

引用 6|浏览13
暂无评分
摘要
Magnesium (Mg) activates < c + a > dislocation slip on the second order pyramidal slip plane. This slip mode is very complex compared to other modes including several metastable structures. Due to the complexity and very similar energies of the different structures, reliably modelling this slip mode is challenging. The problem is exacerbated when considering alloying, in which a combination of 1st order and 2nd order pyramidal slip is usually observed. Motivated by the need for a high fidelity potential for Mg alloys, we have developed first a highly accuracy potential for pure Mg. The new potential shows better agreement with density functional theory and experimental calculations than previous interatomic potentials for Mg. With the help of this new potential, we demonstrate that the basal dissociated < c + a > core is not sessile, as previously thought, and that constant stress molecular dynamics demonstrate clear preference for the 2nd order pyramidal system over the 1st order system.
更多
查看译文
关键词
Machine learning, neural network, interatomic potential, magnesium, slip plane
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要