Simulating the scatter of a polycrystalline in synchrotron diffraction using crystal plasticity simulation

在过去的十年里，应用高能同步辐射衍射来表征工业多晶金属材料的行为已经引起了人们的极大关注。关键的是，模拟苛刻环境的现场实验，在这种环境下，核工作等安全关键行业的材料，一直是支撑其本构定律的重要工具。然而，由于X射线深度穿透的限制，这类实验通常只检查一小部分材料。在这个项目中，将使用微结构信息的晶体塑性模型来开发真正具有代表性的多晶材料。该模型将利用之前和未来的实验进行校准。传统的有限元框架中使用的晶体塑性模型运行缓慢。这个项目将利用频谱方法来进行所需的数值计算，这将比传统方法快几个数量级。光谱方法的这一优势将被用来对微观结构变化对大规模材料响应边界的影响进行统计分析。虽然这项工作的主要目标是量化材料在BARK的行为变化，但也将调查单个颗粒的响应作为其环境的函数。我们将利用机器学习在模拟结果上的应用来建立晶体塑性的代理模型，以加快分析过程。

Application of high energy synchrotron diffraction to characterise the behaviour of industrial polycrystalline metallic materials has gained significant attention in the past decade. Critically, the in-situ experiments simulating the demanding environments under which the materials in safety critical industries such as nuclear work, have been instrumental in underpinning their constitutive laws. However, such experiments often examine only a small volume of materials due to the restrictions associated with X-ray depth penetration.In this project a true representative of polycrystalline material will be developed using a microstructurally informed crystal plasticity model. The model will be calibrated using previous and future experiments. The crystal plasticity model traditionally used within a finite element framework are slow to run. This project will take advantage of spectral method instead to perform numerical calculations required which will be orders of magnitude faster than the traditional methods. This advantage of spectral methods will be exploited to carry out statistical analysis on the effects of microstructure variability on the bounds of large-scale material response.While the main objective of the work is to quantify the material behaviour variability at balk, the response of single grains as a function of their environment will be also investigated. We will take advantage of Machine Learning applied on the results of the simulating to build surrogate models of the crystal plasticity to accelerate the analysis process.