Reactor Pressure Vessels (RPV) are manufactured from medium strength low allow ferritic steel specifically selected for its high toughness and weldability. The normal operating temperature of RPV steels is sufficiently high to ensure that the material remains ductile throughout its service life with an extremely low probability of cleavage under normal and off-normal loading conditions. Understanding and having the ability to predict ductile fracture behaviour is consequently important.
The ductile fracture mechanism is characterised by the nucleation, growth and coalescence of voids at initiating particles within the volume of high triaxial stress and plastic strain ahead of a crack-tip or stress concentrator. The fracture properties of the steels are conventionally determined using standard pre-cracked compact test specimens. Mechanistically based models of fracture can be calibrated against those data.
This paper describes the use of 3D laboratory X-ray tomography to characterise the void distribution associated with the ductile fracture in test specimens and use the data to calibrate the Gurson-Tvergaard-Needleman ductile fracture model.
The tomography successfully captures voids ≥ 6um in diameter and has been used to define the average distribution of void volume fraction as a function of distance below the fracture surface. The tomography results also allow an estimate of the critical and final void volume fractions to be made as well as capture secondary void peaks well below the fracture surface.
This distribution of voids was used to calibrate the Gurson-Tvergaard-Needleman model in order to correlate experimental observations with the finite element models. The models have been able to replicate the observed trends of the void volume fraction distributions away from the fracture surface including the secondary peaks observed by tomography and to reproduce similar J-R curve behaviour as that observed in the test specimens.