Candela R, Gelin S, Mousseau N, Veiga R G A, Domain C, Perez M, Becquart C S
Univ. Lille, CNRS, INRAE, Centrale Lille, UMR 8207-UMET-Unité Matériaux et Transformations, F-59000 Lille, France.
Laboratoire commun EDF-CNRS Etude et Modélisation des Microstructures pour le Vieillissement des, Matériaux (EM2VM), France.
J Phys Condens Matter. 2021 Feb 10;33(6):065704. doi: 10.1088/1361-648X/abc6c3.
We present a mixed-lattice atomistic kinetic Monte-Carlo algorithm (MLKMC) that integrates a rigid-lattice AKMC approach with the kinetic activation-relaxation technique (k-ART), an off-lattice/self-learning AKMC. This approach opens the door to study large and complex systems adapting the cost of identification and evaluation of transition states to the local environment. To demonstrate its capacity, MLKMC is applied to the problem of the formation of a C Cottrell atmosphere decorating a screw dislocation in α-Fe. For this system, transitions that occur near the dislocation core are searched by k-ART, while transitions occurring far from the dislocation are computed before the simulation starts using the rigid-lattice AKMC. This combination of the precision of k-ART and the speed of the rigid-lattice makes it possible to follow the onset of the C Cottrell atmosphere and to identify interesting mechanisms associated with its formation.
我们提出了一种混合晶格原子动力学蒙特卡罗算法（MLKMC），该算法将刚性晶格AKMC方法与动力学激活-弛豫技术（k-ART，一种非晶格/自学习AKMC）相结合。这种方法为研究大型复杂系统打开了大门，使过渡态的识别和评估成本能够适应局部环境。为了展示其能力，将MLKMC应用于α-Fe中螺旋位错周围C科垂耳气团形成的问题。对于该系统，通过k-ART搜索在位错核心附近发生的跃迁，而远离位错发生的跃迁则在模拟开始前使用刚性晶格AKMC进行计算。k-ART的精度与刚性晶格的速度相结合，使得跟踪C科垂耳气团的形成过程并识别与其形成相关的有趣机制成为可能。
