Thermal embrittlement of some cast austenitic stainless steels (CASS) occurs at reactor operating temperatures can lead to a reduction in the fracture toughness and increase in strength. Some aged CASS materials have the potential to have exceedingly low toughness and also show high variability due to the nature of their microstructure or compositional variation within the casting. Because of their low aged toughness with the variability, flaw evaluations of CASS material need to be done with an understanding of the materials aged condition, especially since most US PWR nuclear plants have been given plant life extensions for 60-year operation, and consideration of further extension to 80 years is underway.
In this paper, a flaw evaluation procedure for CASS materials is presented using a new statistical model developed to predict the toughness of fully aged CASS using the material’s chemical composition. The new statistical model was developed based on the experimental toughness using standard 1T CT specimens (generally in the L-C orientation) at 288C to 320C and chemical compositions of the CF8m CASS materials. While the detail development of the model is beyond the scope of this paper, a brief validation of predicted toughness using chemical compositions is presented in this paper. Using the predicted toughness, a flaw evaluation procedure was developed using the Dimensionless-Plastic-Zone-Parameter (DPZP) analysis to determine when limit-load is applicable and also approximate the elastic-plastic correction factor (Z-factor) that needs to be applied to the limit-load solution to predict the failure stress for CASS pipe and fittings with a circumferential surface crack. Variability within a single casting was also determined from available test results which was included in the procedure to determine Z-factor. The procedure was then validated against several CF8m pipe test results which include various pipe diameters, crack sizes, ferrite contents, failure modes (i.e., limit load or EPFM), etc. The as-developed flaw evaluation procedure was also used to determine the Z-factors for four different pipe diameters for a database of 274 pipe/elbows in US PWR plants (whose chemical compositions were known) — essentially solving 1096 sample problems to understand what range of Z-factors might exists in US PWR plants (for CF8m CASS materials) considering all variations in pipe dimensions, ferrite contents, materials’ toughness, etc. Finally, the applicability of the CF8m-based statistical model for use with CF3 and CF8 CASS materials was also investigated by comparing the predictions with available test results.