Automated Grain Boundary (GB) Segmentation and Microstructural Analysis in 347H Stainless Steel Using Computer Vision

Shoieb Chowdhury, PhD Candidate, Department of Mechanical Engineering

Friday, January 27, 2023
1:30 p.m.

Hopeman 224

Austenitic 347H stainless steel offers superior mechanical properties and corrosion resistance required for extreme operating conditions such as high temperature. The change in microstructure due to composition and process variations is expected to impact their properties. Identifying microstructural features such as grain boundaries thus becomes an important task in the process-microstructure-properties loop. Applying convolutional neural network (CNN) based deep-learning models is a powerful technique to detect features from material micrographs in an automated manner. In contrast to microstructural classification, supervised CNN models for segmentation tasks require pixel-wise annotation labels. However, manual labeling of the images for segmentation task poses a major bottleneck for generating training data and labels in a reliable and reproducible way within a reasonable timeframe. Specially, for faster material discovery by changing alloy compositions, the microstructural characterization needs to be expedited. In this study, we attempt to overcome such limitations by utilizing multi-modal microscopy to generate labels directly instead of manual labeling. We combine scanning electron microscopy (SEM) images of 347H stainless steel as training data and electron backscatter diffraction (EBSD) micrographs as pixel-wise labels for grain boundary detection as a semantic segmentation task. The viability of our proposed method is evaluated by experimenting with a set of deep CNN architectures. We demonstrate that despite producing instrumentation drift during data collection between two modes of microscopy, this method performs comparably in similar segmentation tasks that used manual labeling. Additionally, we find that naïve pixel-wise segmentation results in small gaps and missing boundaries in predicted grain boundary map. By incorporating topological information during model training, the connectivity of grain boundary network and segmentation performance is improved. Finally, our approach is validated by accurate computation on downstream tasks of predicting the underlying grain morphology distributions which are ultimate quantities of interest for microstructural characterization.