Diffraction techniques can powerfully and nondestructively probe materials while maintaining high resolution in both space and time. Unfortunately, these characterizations have been limited and sometimes even erroneous due to the difficulty of decoding the desired material information from features of the diffractograms.
In this talk, I will discuss our recent efforts using atomistic simulation and machine learning techniques to improve the decoding process of microstructure statistics from diffractograms.
First, I will focus on how a simulation-based approach can reveal inconsistencies in X-ray width
methods used to characterized nanomaterials. By assessing 1D X-ray diffractograms generated from atomistic simulations of a nanocrystalline, Ni nanowire subjected to anneal, creep, and fatigue, I will discuss challenges and solutions to guide the selection and interpretation of width methods. Second, I will discuss how to compute machine-identified features that fully summarize a diffractogram and how to employ machine learning to reliably connect these features to an expanded set of structural statistics. I will show that, when based on machine-identified features rather than human-identified features, a machine-learning model can not only predict 1-point statistics (i.e. density) but also a 2-point statistic (i.e. spatial distribution) of the defect population. Building on these recent results, I will end this talk with a perspective and opportunities to advance the state of the art for decoding diffractograms.
Sandia National Laboratories is a multi-mission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC., a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under Contract No.~DE-NA0003525.
