Statistical Modeling of Fracture Toughness Data

Abstract

The fracture toughness of the zirconium alloy (Zr-2.5Nb) is an important parameter in determining the flaw tolerance for operation of pressure tubes in reactor. Fracture toughness data have been generated by performing rising pressure burst tests on sections of pressure tubes removed from operating reactors. The test data were used to generate a lower-bound fracture toughness curve, which is used in defining the operational limits of pressure tubes. The thesis presents a comprehensive statistical analysis of burst test data and develops a multivariate statistical model to relate toughness with material chemistry, mechanical properties, and operational history. The proposed model can be useful in predicting fracture toughness of specific in-service pressure tubes, thereby minimizing conservatism associated with a generic lower bound approach.