NCSU Libraries Repository (North Carolina State University Libraries) · 2003 · 10 citations · 2 references
Open access
EngineeringMaterials EngineeringNuclear EngineeringMechanical Behavior Of MaterialsNuclear Safety AnalysisReactor PhysicsDamage MechanicsFailure AnalysisFuel IntegrityReactor Core DesignSevere Accident AnalysisChannelsCore ModelsCreep Strain LocalizationLevel 1 Probabilistic Safety AssessmentLevel 2 Probabilistic Safety AssessmentReactor Core Damage ProgressionResidual StrainStrain Localization
AECL uses the MAAP CANDU code to calculate the progression of a severe core damage accident in a CANDU reactor to support Level 2 Probabilistic Safety Assessment and Severe Accident Management activities. Experimental data are required to ensure that the core damage models used in MAAP CANDU code are adequate. In SMiRT 16, details of single channel experiments were presented to elucidate the mechanisms of core debris formation. This paper presents the progress made in severe core damage experiments since then using single channels in an inert atmosphere and results of the model development work to support the experiments. The core disassembly experiments are conducted with one-fifth scale channels made of Zr-2.5wt%Nb containing twelve simulated fuel bundles in an inert atmosphere. The reference fuel channel geometry consists of a pressure tube/calandria tube composite, with the pressure tube ballooned into circumferential contact with the calandria tube. Experimental results from single channel tests showed the development of time-dependent sag when the reference channel temperature exceeded 850 degC. The test results also showed significant strain localization in the gap at the bundle junctions along the bottom side of the channel, thus suggesting creep to be the main deformation mechanism for debris formation. An ABAQUS finite element model using two-dimensional beam elements with circular cross-section was developed to explain the experimental findings. A comparison of the calculated central sag (at mid-span), the axial displacement at the free end of the channel and the post-test sag profile showed good agreement with the experiments, when strain localization was included in the model, suggesting such a simple modelling approach would be adequate to explain the test findings. The results of the tests are important not only in the context of the validation of the analytical tools and models adopted by AECL for the severe accident analysis of CANDU reactors but also for the overall confirmation of the slow nature of core disassembly progression in CANDU under severe core damage accident conditions. (author)
2