Prediction of vacancy formation energies at tungsten grain boundaries from local structure via machine learning method

Prediction of vacancy formation energies at tungsten grain boundaries from local structure via machine learning method
复制标题

通过机器学习方法从局部结构预测钨晶界的空位形成能

DOI:
10.1016/j.jnucmat.2021.153412
复制
发表时间:
2021-11
影响因子:
3.1
通讯作者:
Xuebang Wu
Xuebang Wu
中科院分区:
工程技术2区
文献类型:
--
作者:
Yuxuan Wang;Xiaolin Li;Xiangyan Li;Yuxiang Zhang;Yange Zhang;Yichun Xu;Yawei Lei;C.S. Liu;Xuebang Wu

文献摘要

参考文献

被引文献

相似文献

晶界对核材料的力学性能和抗辐照性能起着至关重要的作用。因此,理解和预测晶界附近的缺陷性质是非常必要的。在这里,我们提出了一个框架,通过机器学习(ML)技术来预测钨(W)中的空位形成能(E V f)。计算了W中46种[001]对称倾斜GB(STGB)附近4496个原子位的EVf值作为数据库,并选取了8个合适的变量来表征周围原子的构型和原子位的位置。通过径向基核函数支持向量机(RBF-SVM)的交叉验证(CV)和广义验证的良好预测结果证明了所选变量和RBF-SVM方法的适用性和有效性。此外,由于位错排列和原子组态的差异较大,采用分离CV方法需要将STGB分为高角度、低角度I和低角度II三种类型,其预测精度有较大的提高。由于该方法采用了空间尺寸特性、密度和位置等几何因素作为ML分析的描述符,因此对α-Fe等其他材料具有鲁棒性和通用性,有利于预测和理解界面附近空位的形成。
Grain boundary (GB) plays a crucial role in the mechanical properties and irradiation resistance of nuclear materials. It is thus essential to understand and predict the defect properties near GBs. Here, we present a framework for predicting vacancy formation energy (E V f) near GBs in tungsten (W) by machine learning (ML) technique. The E V f values of 4496 atomic sites near 46 types of [001] symmetry tilt GB (STGB) in W are calculated as database and eight appropriate variables are selected to characterizing the surrounding atomic configuration and location of atomic sites. Via the support vector machine with the radial basis kernel function (RBF-SVM), the good predicted results of cross validation (CV) and generalized verification prove the suitability and effectiveness of the selected variables and RBF-SVM method. Beside, due to their big differences in dislocation arrangement and atomic configuration, the STGBs need to be divided into three types, high angle, low angle-I and low angle-II STGBs, for adopting the Separate CV, and their predicted accuracies were found to have big improvements. Because the present method adopts geometrical factors, such as spatial size characteristic, density and location, as descriptors for the ML analysis, it is robust and general to other materials such as α-Fe, and beneficial to predict and understand the vacancy formation near interfaces.
DOI: 10.1038/nmat1842
发表时间: 2007-03-01
期刊: NATURE MATERIALS
影响因子: 41.2
作者:
Sickafus, Kurt E.;Grimes, Robin W.;Uberuaga, Blas P.
通讯作者: Uberuaga, Blas P.
通过虚拟筛选对界面结构和能量的预测。
DOI: 10.1126/sciadv.1600746
发表时间: 2016-11
期刊: Science advances
影响因子: 13.6
作者:
Kiyohara S;Oda H;Miyata T;Mizoguchi T
通讯作者: Mizoguchi T
DOI: 10.1088/0959-5309/52/1/315
发表时间: 1940
期刊: --
影响因子: --
作者:
W. Bragg
通讯作者: W. Bragg
DOI: 10.1088/0959-5309/52/1/304
发表时间: 1940-01-01
影响因子: --
作者:
Burgers, JM
通讯作者: Burgers, JM
表征辐照引起的点缺陷与 Fe、Mo 和 W 中几个倾斜对称晶界之间相互作用的主要物理参数
DOI: 10.1016/j.jnucmat.2013.09.046
发表时间: 2014
影响因子: 3.1
作者:
Xiangyan Li;Wei Liu;Yichun Xu;C.S. Liu;Q.F. Fang;B.C. Pan;Jun-Ling Chen;G.-N. Luo;Zhiguang Wang
通讯作者: Zhiguang Wang