Shallow Flaws Under Biaxial Loading Conditions—Part I: The Effect of Specimen Size on Fracture Toughness Values Obtained From Large-Scale Cruciform Specimens1

Abstract

A technology to determine shallow-flaw fracture toughness of reactor pressure vessel (RPV) steels is being developed. This technology is for application to the safety assessment of RPVs containing postulated shallow-surface flaws. It has been shown that relaxation of crack-tip constraint causes shallow-flaw fracture toughness of RPV material to have a higher mean value than that for deep flaws in the lower transition temperature region. Cruciform beam specimens developed at Oak Ridge National Laboratory (ORNL) introduce far-field, out-of-plane biaxial stress components in the test section that approximates the nonlinear stresses resulting from pressurized-thermal-shock (PTS) loading of an RPV. The biaxial stress component has been shown to increase stress triaxiality (constraint) at the crack tip, and thereby reduce the shallow-flaw fracture toughness enhancement. The cruciform specimen permits controlled application of biaxial loading ratios, resulting in controlled variation of crack-tip constraint. An extensive matrix of intermediate-scale cruciform specimens with a uniform depth surface flaw was previously tested and demonstrated a continued decrease in shallow-flaw fracture toughness with increasing biaxial loading. This paper describes the test results for a series of large-scale cruciform specimens with a uniform depth surface flaw. These specimens were all of the same size with the same depth flaw and were tested at the same temperature and biaxial load ratio (1:1). The configuration is the same as the previous set of intermediate-scale tests, but has been scaled upward in size by 150 percent. These tests demonstrated the effect of biaxial loading and specimen size on shallow-flaw fracture toughness in the lower transition temperature region for RPV materials. For specimens tested under full biaxial (1:1) loading at test temperatures in the range of 23°F (−5°C) to 34°F (1°C), toughness was reduced by approximately 15 percent for a 150-percent increase in specimen size. This decrease was slightly greater than the predicted reduction for this increase in specimen size. The size corrections for 1/2T C(T) specimens did not predict the experimentally determined mean toughness values for larger size shallow-flaw specimens tested under biaxial (1:1) loading in the lower transition temperature region.