Abstract
The current research adopts a novel approach by integrating correlative microscopy and machine learning in order to study creep cavitation in an ex-service 9%Cr 1%Mo Grade 91 ferritic steel. This method allows for a detailed investigation of the early stages of the creep life, enabling identification of features most prone to damage such as precipitates and the ferritic crystal structure. The microscopy techniques encompass Scanning Electron Microscopy (SEM) imaging and Electron Back-scattered Diffraction (EBSD) imaging, providing insights into the two-dimensional distribution of cavitation. A methodology for acquiring and analysing serial sectioning data employing a Plasma Focused Ion Beam (PFIB) microscope is outlined, complemented by 3D reconstruction of backscattered electron (BSE) images. Subsequently, cavity and precipitate segmentation was performed with the use of the image recognition software, DragonFly and the results were combined with the 3D reconstruction of the material microstructure, elucidating the decoration of grain boundaries with precipitation, as well as the high correlation of precipitates and grain boundaries with the initiation of creep cavitation. Comparison between the 2D and 3D results is discussed.
