English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Austenitic 347H stainless steel offers superior mechanical properties andcorrosion resistance required for extreme operating conditions such as hightemperature. The change in microstructure due to composition and processvariations is expected to impact material properties. Identifyingmicrostructural features such as grain boundaries thus becomes an importanttask in the process-microstructure-properties loop. Applying convolutionalneural network (CNN) based deep-learning models is a powerful technique todetect features from material micrographs in an automated manner. Manuallabeling of the images for the segmentation task poses a major bottleneck forgenerating training data and labels in a reliable and reproducible way within areasonable timeframe. In this study, we attempt to overcome such limitations byutilizing multi-modal microscopy to generate labels directly instead of manuallabeling. We combine scanning electron microscopy (SEM) images of 347Hstainless steel as training data and electron backscatter diffraction (EBSD)micrographs as pixel-wise labels for grain boundary detection as a semanticsegmentation task. We demonstrate that despite producing instrumentation driftduring data collection between two modes of microscopy, this method performscomparably to similar segmentation tasks that used manual labeling.Additionally, we find that na\"ive pixel-wise segmentation results in smallgaps and missing boundaries in the predicted grain boundary map. Byincorporating topological information during model training, the connectivityof the grain boundary network and segmentation performance is improved.Finally, our approach is validated by accurate computation on downstream tasksof predicting the underlying grain morphology distributions which are theultimate quantities of interest for microstructural characterization.

Automated Grain Boundary (GB) Segmentation and Microstructural Analysis in 347H Stainless Steel Using Deep Learning and Multimodal Microscopy