PSI - Issue 13

J Beswick et al. / Procedia Structural Integrity 13 (2018) 63–68

64

Beswick et al. / Structural Integrity Procedia 00 (2018) 000 – 000

2

defect or crack, would contribute to failure through a ductile mechanism via nucleation, growth and coalescence of voids. At lower temperatures, and combined with irradiation damage, the second phase particles infer a probability of cleavage fracture. This is a brittle failure mode, initiated at a micro-crack formed by particle rupture. When the stress is sufficient to cause instability, the micro-crack propagates through the material causing catastrophic failure. Defect tolerance assessments usually take a single cleavage fracture toughness value when calculating a limiting defect size or load. This is considered conservative in many cases, since the fracture toughness will be determined by testing of specimens with deep cracks, i.e. under conditions referred to as high constraint. It is known that the reduction in hydrostatic stresses observed at shallower cracks contributes to higher apparent fracture toughness. Further, defects are most onerous in the areas around RPV welds, and where post-weld heat treatment is not possible, such as control rod drive mechanisms. In these areas residual stresses are known to remain in the structure. In addition, the heating from welding may introduce damage through plastic strains, see e.g. Brayshaw et al. (2016). Presented in this paper are results from an experimental programme and analyses aimed at assessing the effects of plastic strains on the cleavage fracture toughness of a ferritic RPV material, at three levels of constraint – very low, intermediate and high. The results are firstly presented in the J-Q space, where J is the J -integral and Q quantifies constraint, deriving two-parameter fracture toughness curves. Modified Beremin local approaches are then used to investigate whether the fracture toughness of the pre-strained material could be derived after their calibration with as-received material, relying solely on the change of elastic, yield and flow properties.

Nomenclature a, r

crack length and distance from the crack tip (mm) thickness and width of SENB specimens (mm) Young’s modulus (GPa) and Poisson’s ratio

B, W E,  J, J c

J-integral and fracture toughness, i.e. J-integral at failure (kJ/m 2 ) constraint-corrected (apparent) fracture toughness (kJ/m 2 ) m shape parameter of Weibull distribution Q, Q c constraint parameter and its value at critical J-integral  equivalent plastic strain  0 ,  1 proportionality stress and maximum principal stress (MPa)  w ,  u Weibull stress and scale parameter of Weibull distribution (MPa) parameters of plasticity correction functions to Weibull stress  k parameters of constraint correction procedure  p J mat

2. Experimental and analysis methods

SENB specimens were used for the fracture toughness test programme as they allow for varying levels of constraint to be introduced by changing the crack depth. This type of specimen has a standardised design, see ASTM 1820-1 (2003), with a , W =50 mm and B =25 mm, denoting crack depth, specimen width and thickness, respectively. Three crack depths were considered, a/W =0.4, 0.2, 0.05, referred to as the high-, intermediate-, and low-constraint. Standards were followed for specimen design, test procedure and post-test fracture toughness calculations, where possible. Deviation from standards was necessary for the low constraint geometries ( a/W =0.05), where fatigue pre cracks above the specified length were introduced, followed by material removal from the notched edge. Specimens were machined from plates of A533B RPV steel. 10 SENB specimens of each crack length were prepared from the as-received (AR) material. Dog-bone tensile specimens were used to introduce approximately uniform 5% plastic strain in the remaining material, referred to as the pre-strained (PS). 10 SENB specimens of each crack length were prepared from PS material. These were machined with gauge lengths along the dog-bone gauge length, so that the principal plastic strain was normal to the crack. To ensure a cleavage mode of failure, all tests were performed at -140°C. The specimens were loaded to failure through three-point bending and the crack-mouth opening displacement was measured with clip-gauges attached to either integral or retrofitted knife edges. The fracture toughness, J c , was calculated using the load-displacement traces according to 1820-1 (2003), including plasticity correction factors for the low constraint geometries.