Sa. Chavez et al., COMPARISON OF STRESS-BASED AND STRAIN-BASED CREEP FAILURE CRITERIA FOR SEVERE ACCIDENT ANALYSIS, Nuclear Engineering and Design, 155(3), 1995, pp. 603-622
We conducted a parametic analysis of stress-based and strain-based cre
ep failure criteria to determine if there is a significant difference
between the two criteria for SA533B vessel steel under severe accident
conditions. Parametric variables include debris composition, system p
ressure, and creep strain histories derived from different testing pro
grams and mathematically fit, with and without tertiary creep. Results
indicate significant differences between the two criteria. Stress gra
dient plays an important role in determining which criterion will pred
ict failure first. Creep failure was not very sensitive to different c
reep strain histories, except near the transition temperature of the v
essel steel (900 K to 1000 K). Statistical analyses of creep failure d
ata of four independent sources indicate that these data may be pooled
, with a spline point at 1000 K. We found the Manson-Haferd parameter
to have better failure predictive capability than the Larson-Miller pa
rameter for the data studied.