Classical molecular dynamics simulations with global neural network machine learning potential are used to study early stage oxidation and dissolution behaviors of bcc Fe surfaces contacting with stagnant oxygen disso...Classical molecular dynamics simulations with global neural network machine learning potential are used to study early stage oxidation and dissolution behaviors of bcc Fe surfaces contacting with stagnant oxygen dissolved liquid leadbismuth eutectic(LBE-O).Both static and dynamic simulation results indicate that the early stage oxidation and dissolution behaviors of bcc Fe show strong orientation dependence under the liquid LBE environments,which may explain the experimental observations of uneven interface between iron-based materials and liquid LBE.Our investigations show that it is the delicate balance between the oxide growth and metal dissolution that leads to the observed corrosion anisotropy for bcc Fe contacting with liquid LBE-O.展开更多
基金the National Natural Science Foundation of China(Grant No.U1832206).
文摘Classical molecular dynamics simulations with global neural network machine learning potential are used to study early stage oxidation and dissolution behaviors of bcc Fe surfaces contacting with stagnant oxygen dissolved liquid leadbismuth eutectic(LBE-O).Both static and dynamic simulation results indicate that the early stage oxidation and dissolution behaviors of bcc Fe show strong orientation dependence under the liquid LBE environments,which may explain the experimental observations of uneven interface between iron-based materials and liquid LBE.Our investigations show that it is the delicate balance between the oxide growth and metal dissolution that leads to the observed corrosion anisotropy for bcc Fe contacting with liquid LBE-O.
