Characterization of the Grain Size of the Silicon Carbide Layer in AGR TRISO Fuels

The silicon carbide (SiC) layer is crucial for the structural integrity and fission product retention in tristructural isotropic (TRISO) fuel. The Advanced Gas Reactor (AGR) Fuel Development and Qualification program for over 20 years has developed many QA/QC methods for characterizing TRISO fuel. The previous procedure for analyzing SiC layer microstructure utilized scanning electron microscopy (SEM) with a pass/fail visual standard. This historic method was based on past performance data showing coarse columnar grains in the SiC are not ideal. More recent results from the AGR program indicate that TRISO particles with excessively fine grain SiC may experience enhanced grain boundary diffusion of fission products at elevated temperatures under simulated accident conditions. The visual standard approach is ill-equipped to handle a lower limit on SiC grain size specification and is also inherently unable to provide a quantitative measure of individual particle grain size or the statistical distribution of a batch. These limitations have been addressed by the development of new methods for SiC layer grain size quantification using backscattered electron (BSE) imaging and automated grain boundary detection algorithms. The study presented herein uses the SiC analysis technique published in recent AGR reports to provide a statistical sampling of TRISO particles that were fabricated under each AGR irradiation campaign to determine the mean grain size of the SiC layer. Grain boundary detection was applied to backscattered electron (BSE) images taken on different instruments to provide a measure of the impact of instrument resolution on grain size characterization of the SiC layer in TRISO particles. This technique was successful in providing a quantitative value for the grain size and distribution within each fabrication batch. Additionally, this procedure provides a rapid analysis of batches of particles which may enable industry implementation. The data from each campaign supports the development of a high-quality database composed of microstructural data on the SiC layer from analysis of historic TRISO fuel and those developed during the AGR program. The ability to link microstructure to the irradiation performance will aid in accelerating fuel qualification. Ongoing post irradiation examination (PIE) of AGR TRISO fuel may be able to link the data from the experiments to determine the ideal range of SiC microstructure for potential fuel specifications.