|About this Abstract
||2018 TMS Annual Meeting & Exhibition
||Fatigue in Materials: Fundamentals, Multiscale Modeling and Prevention
||Identification of Fatigue Weak Links in Aluminum Alloys Using a Data-driven Approach
||Brian Wisner, Krzysztof Mazur, Antonios Kontsos
|On-Site Speaker (Planned)
The fatigue damage process in Aluminum alloys is complex and depends on several microstructural parameters, which explains in part the highly variable observations reported on both damage incubation/initiation as well as damage growth. In an effort to investigate the effect of local changes induced by cyclic loading, Aluminum microstructures are observed at the grain-scale and are quantitatively linked with deformation measurements made by Digital Image Correlation (DIC) and Acoustic Emission (AE) obtained from inside the Scanning Electron Microscope. Since such local monitoring approaches produce relatively large datasets, the identification of fatigue weak links for this particular type of alloy is attempted in this work by a data-processing framework that involves techniques for data reduction and removal of noise. Furthermore, by cross-evaluation of classification methods combined with outlier analysis and in situ observed microstructural changes, fatigue weak links are identified and progressive damage estimates are provided.
||Planned: Supplemental Proceedings volume