Large Dataset Generation, Integration and Simulation in Materials Science (Preprint)

摘要

As new and rapid materials characterization tools become widely available, large datasets are been generated at a faster and faster pace, which is both exciting and challenging. The exciting aspects are that large amounts of composition, structure, and spectral information about materials are being gathered routinely. This promises to accelerate the traditional scientific pursuits in Materials Science of establishing processing-structure and structure-property relationships and hypothesis testing. Even more exciting is the promise to provide quantitative validations of the physics-based forward models of these input-response relationships by providing statistically significant datasets for this validation. The main challenges presented by the emergence of these techniques are (1) rapid acquisition of larger and larger datasets, (2) using this information to provide predictive capabilities in materials systems, and (3) analysis of the increasing volumes of data.