Application of Master Curve Data for Reactor Vessel Steels

Abstract

The Master Curve method has been developed to determine fracture toughness of a specific material in the brittle-to-ductile transition range. This method is technically more descriptive of actual material behavior and accounts for the statistical nature of fracture toughness properties as an alternative to the current ASME Code reference toughness curves. The Master Curve method uses a single temperature, To, as an index of the Master Curve fracture toughness transition temperature. This method has been successfully applied to numerous fracture toughness data sets of pressure vessel steels contained in the Master Curve database, including the beltline materials for the Kewaunee reactor pressure vessel. The database currently contains over 5,500 toughness data records for vessel weld, plate and forging materials, and it is currently being updated to include more recent fracture toughness data. Application of Master Curve fracture toughness data to reactor pressure vessel (RPV) integrity evaluations requires some assumptions relative to the degree of constraint in the fracture toughness test specimens versus the actual assumed RPV flaw. An excessive degree of conservatism can be introduced if the constraint levels are substantially different. In performing a Master Curve evaluation, the analysis may be restricted by the type of fracture toughness data available. Any excess conservatism should be appropriately considered when the overall safety margin is applied. For example, the precracked Charpy three-point bend specimen actually has some advantages over the compact tension specimen when the application involves a shallow surface flaw in a RPV wall. This paper analyzes some key fracture toughness results from several weld data sets containing both unirradiated and irradiated data to evaluate constraint effects in fracture toughness and pre-cracked Charpy specimens. The evaluated To values were compared to determine if there is any difference in bias from specimen geometry between the unirradiated and irradiated data.