Convolutional neural networks for expediting the determination of minimum volume requirements for studies of microstructurally small cracks, Part I: Model implementation and predictions
Convolutional neural networks for expediting the determination of minimum volume requirements for studies of microstructurally small cracks, Part I: Model implementation and predictions
复制标题
DOI：
10.1016/j.commatsci.2022.111290
复制
发表时间：
2022-05
影响因子：
3.3
通讯作者：
Karen J. DeMille;A. Spear
中科院分区：
文献类型：
--
作者：
Karen J. DeMille;A. Spear
卷积神经网络（CNN）的实施，以加快微结构小裂纹（RVE MSC）的代表性体积元素的测定。根据定义，RVE MSC是微结构小裂纹（MSC）周围所需的最小微结构体积，以实现裂纹前缘参数相对于体积尺寸的收敛。在以前的研究中，RVE MSC确定使用计算昂贵的有限元（FE）框架，涉及许多微观结构实例的模拟。为了提高确定RVE MSC的计算效率，本文利用CNN来减少确定RVE MSC所需的FE模拟的数量。使用来自先前基于FE的RVE MSC研究的数据，训练CNN以预测RVE MSC，ip值，其量化裂纹前沿参数收敛相对于在单个裂纹前沿点p处评估的微观结构实例i的体积大小，给定局部微观结构和几何信息。预测的RVE MSC，ip值随后用于估计RVE MSC值。进行研究以确定训练数据的最佳量，评估基于CNN的RVE MSC估计性能，并通过启用在体积要求方面被认为至关重要的微结构的向下选择来展示CNN作为微结构实例化筛选工具的用途。比较了个体和集合CNN预测。虽然没有发现CNN足够准确以取代所有FE模拟，但发现CNN作为快速筛选工具是有效的，用于提高基于FE的RVE MSC确定框架的效率和加快未来的RVE MSC研究。
Convolutional neural networks (CNNs) are implemented to expedite the determination of representative volume elements for microstructurally small cracks (RVE MSC). By definition, RVE MSC is the minimum volume of microstructure required around a microstructurally small crack (MSC) to achieve convergence of crack-front parameters with respect to volume size. In a previous study, RVE MSC was determined using a computationally expensive finite-element (FE) framework involving the simulation of many microstructural instantiations. With the aim of increasing the computational efficiency of determining RVE MSC, CNNs are leveraged herein to reduce the number of FE simulations required to determine RVE MSC. Using data from the previous FE-based RVE MSC study, CNNs are trained to predict RVE MSC, ip values, which quantify crack-front parameter convergence with respect to volume size for microstructural instantiation i evaluated at individual crack-front points p, given local microstructural and geometrical information. Predicted RVE MSC, ip values are subsequently used to estimate RVE MSC values. Studies are carried out to determine the optimal amount of training data, assess CNN-based RVE MSC estimation performance, and demonstrate the use of CNNs as microstructural-instantiation screening tools by enabling downselection of microstructures that are considered critical in terms of volume requirements. Individual and ensemble CNN predictions are compared. While CNNs are not found to be accurate enough to replace all FE simulations, CNNs are found to be effective as a rapid screening tool for improving the efficiency of the FE-based RVE MSC determination framework and for expediting future RVE MSC studies.