No abstract

Test blocks of a manganese-nickel-molybdenum (Mn-Ni-Mo) low-alloy steel, machined from a nozzle cut-out of a 230-mm-thick forging shell, were provided by France in 1978 to be investigated in the second phase of an IAEA-coordinated research program on the behavior of advanced pressure vessel steels under neutron irradiation. The results obtained by different laboratories exhibit a great deal of scatter in Charpy-V transition temperatures, both in the initial condition and after neutron irradiation. Therefore, the shifts in Charpy-V transition curves induced by neutron irradiation at similar fluences of about 3 × 1023 n ∙ m-2 (E ⩾ 1 MeV) cover a wide temperature range from 0 up to 110°C. A more detailed investigation of these results was undertaken, and additional Charpy tests were conducted to clarify the considerable uncertainty about the sensitivity of this forging to radiation embrittlement. The differences in results were found to be due to the various orientations and locations of the Charpy-V specimens extracted by the different laboratories, in some cases to insufficient test data in the transition region of the Charpy curve, and to the detrimental effect on Charpy-V properties of segregated zones (ghost lines) found in specific locations of large forgings.

Tests were performed on a 2¼ Cr–1 Mo steel to measure the fracture toughness at initiation, KIc and at arrest, KIa . The results were compared with those obtained on another pressure vessel steel (A508) of similar strength. Two techniques were used to measure KIa : (i) isothermal compact crack arrest (CCA) tests, and (ii) specially designed thermal shock experiments using an externally notched ring. These specimens were cooled to −196 °C and then heated by induction in the centre of the ring to produce very steep thermal gradients. This caused crack initiation from the notch. The crack propagates very rapidly (∼500 m s−1 ) and stopped when it reached the warmer region of the specimen. The specimens were analysed using an elastic–plastic finite element method to determine KIa values. These tests reveal a greater temperature shift (∼100 °C) between KIc and KIa in 2¼ Cr–1 Mo steel than in A508 steel. Detailed metallographical examinations of the micromechanisms of crack propagation and arrest in the 2¼ Cr–1 Mo steel showed that this involves the nucleation of a three‐dimensional network of cleavage microcracks which change their direction at bainitic packet boundaries. The remaining uncracked ligaments between the cleavage microcracks break by ductile rupture mechanism

At the beginning of 1986, thirty-two 900-MWe and four 1300-MWe nuclear power stations were in service in France. A regulatory surveillance program is in progress concerning the base metal, weld metal, and heat affected zones of every nuclear pressure vessel; a reference metal sample is also present in each reactor. This surveillance program was implemented largely in accordance with U.S. rules. The specifications for the core shell forgings in MnNiMo low alloy steel and the associated welds require low contents in embrittling elements such as copper and phosphorus. The results presented here concern materials containing less than 0.08 wt% Cu and 0.011 wt% P, except for the welds of the first six reactor vessels, which have 0.13 wt% Cu and 0.019 wt% P maximum. The main aim is to obtain the RTNDT shift from Charpy-V tests, in relation to the fluence measured from fissile and activation dosimeters. Tensile, compact, and three-point bend specimens are also included in the capsules. Results for the first capsules removed after up to four years of operation, corresponding to fluences (E > 1 MeV) less than 1.5 × 1019 n · cm−2 (2.9 × 10−2 dpa), are presented and evaluated. The mean and upper values of the RTNDT shift are about 20 and 40°C, respectively, and there is no significant difference between parent metal and weld. A new formula to predict embrittlement has been derived from test reactor experiments, because of the strict limits on the residual element contents in French materials when compared with materials for which other formulae have been established. The surveillance program results are discussed and compared with the predictions of these formulae.

No abstract