Initial Observations from XCT Imaging of AGR-5/6/7 TRISO Particles

The Advanced Gas Reactor Fuel Development and Qualification (AGR) Program has entered the post-irradiation examination (PIE) phase of its final irradiation: the AGR-5/6/7 experiment. Compacts consisting of tri-structural isotropic (TRISO) particle fuel embedded in a graphitic matrix were irradiated at a range of conditions spanning timeaveraged volume-average temperatures of 756°C to 1313°C and 5.66% to 15.26% fissions per initial heavy metal atom with the objective of supporting fuel qualification for advanced gas reactors by expanding the irradiated dataset for reference-design TRISO particles beyond normal operating conditions. PIE of whole irradiated compacts and individual particles deconsolidated from irradiated compacts is critical to fulfilling this objective. Some of the AGR-5/6/7 compacts irradiated in the Advanced Test Reactor (ATR) at Idaho National Laboratory (INL) are shipped to Oak Ridge National Laboratory (ORNL) for PIE. These compacts are generally deconsolidated in nitric acid to free particles from the graphitic matrix, then individual particles are scanned for fission product inventory using a high-purity gamma detector to identify the rare, failed particles which leaked a significant inventory of some fission products. X-ray computed tomography (XCT) is a key characterization method applied in PIE of TRISO particles at ORNL. In XCT, 2D radiographs of the sample are acquired at regularly spaced angles of rotation, then reconstructed into a single 3D image in which image intensity is defined by the attenuation of x-rays by the material in each voxel. XCT has been applied at ORNL to irradiated particles from the beginning of the AGR program to qualitatively identify defects in particles with missing fission product inventory; however, modern developments in image processing for large datasets have enabled the accurate quantitative segmentation of XCT images of TRISO particles, greatly expanding the scope of analysis which may be performed on this data. For example, this capability has been successfully applied to previously imaged AGR particles to quantify the degree of buffer densification as a function of irradiation conditions. This presentation will focus on recent results from XCT imaging of irradiated AGR-5/6/7 particles, focusing in three main areas. First, results will be presented for failed particles identified by gamma scanning with an emphasis on identification of failure mechanisms and comparison to electron microscopy images taken of the failure points. Second, results will be presented for randomly selected irradiated particles which are representative of the general population of the fuel, focusing on features and behaviors of interest which, although unusual, did not result in observed particle failure or release of fission products. Finally, initial results will be presented on the use of XCT with automated image segmentation to measure the recession of the silicon carbide (SiC) layer of as-fabricated and irradiated AGR-5/6/7 particles during oxidation experiments based on automated correlation of pre- and post-oxidation imaging. These diverse results demonstrate the power of non-destructive 3D imaging using XCT for nuclear fuels experiments, particularly when combined with quantitative image segmentation.