Lead-Induced Stress Corrosion Cracking of Alloy 690TT in Caustic Environments

UoM administered thesis: Phd

  • Authors:
  • Giulia Mazzei

Abstract

Lead-induced stress corrosion cracking (PbSCC) is a degradation mechanism that can take place in the secondary side of pressurized water reactor (PWR) and affect the integrity of the steam generator tubes which are manufactured out of thermally treated Alloy 690 (A690TT). Pb is present as in impurity in the secondary circuit and may build up in the steam generator, where caustic crevices might also form. In this regards, the resistance of A690TT to PbSCC deteriorates greatly in caustic solutions, likely because Pb can impair the passivity of the anodic film. It has been proposed that PbSCC of Ni-Cr-Fe alloys could be explained by a combination of film-rupture/slip-dissolution mechanism and de-alloying, although, due to lack of experimental evidence for A690TT, it is necessary to corroborate this hypothesis. Therefore, the aim of this project was to advance the mechanistic understanding of PbSCC, by investigating the early stage of cracking initiation and cracking. A range of complementary electron microscopy, combined with advanced techniques (scanning electron microscopy, focused ion beam, transmission Kikuchi diffraction microscopy, transmission electron microscopy and scanning transmission electron microscopy coupled with energy-dispersive X-ray spectroscopy) used to correlate the role of the microstructure with the materials performance. Stress corrosion cracking tests were conducted using C-rings and capsules specimens cut from tubes of A690TT, and exposed to a solution representative of a secondary side crevice with a range of PbO present ( ~10 pHhigh T, 0, 10, 500 ppm to 10,000 ppm of PbO), and Na3PO4 (0, 1700 and 3000 ppm); this latter was added to investigate its possible inhibiting effects on SCC. The effect of surface finish on PbSCC was also evaluated by testing surfaces that were either as received (with a surface deformed layer) or polished. In particular, crack initiation and propagation were investigated my means of short (150 h) and long exposure (500 h) to Pb-caustic solutions, and the effect of stress on PbSCC was estimated from the crack propagation depth on the C-ring specimens. It was found that although PbSCC in A690TT was transgranular in nature, crack nucleation was not always localized at transgranular sites (such as crystallographic slots and attacked slip bands), but also initiated from grain boundaries. Intergranular initiation was more common on polished surfaces where the as received surface deformed layer and associated ultrafine grained layer had been removed, thus exposing grain boundaries decorated with carbides to undergo localized attack. In this regards, the near-surface deformed region had the beneficial effect of homogenizing stress distribution, and reducing the number of sites of crack nucleation, although it did not affect the rate of propagation due to the limited thickness of this cold worked layer (< 5 μm). Furthermore, a positive correlation between applied stress and crack depth was identified. Metallic Ni was discovered on the surfaces of the specimens, in the form of Ni-enrichment along slot walls (polished surface) and Ni “islands� surrounded by a Cr-Fe oxide (as received surface). Also, layers of metallic Ni formed along the flanks of well-developed cracks, where it maintained the same crystallographic orientation as the cracked grain, therefore suggesting that a de-alloying mechanism was active within advancing PbSCC cracks, with some indications of crystallographic propagation along {100} and {110} planes. All these experimental evidences indicate that cracking was associated with a film-rupture/slip-dissolution mechanism that was aided by de-alloying. Finally, the use of phosphate, as Na3PO4, was evaluated as possible inhibitor for PbSCC of Alloy 690TT, although it was found that its effect was strongly dependent on the surface finish. In fact, whilst the AR surface displayed a reduction in cracking, this was not the case for polished surfaces, where localised attacked occurred. Interestingly, phosphorous was not detected at near-tip regions, suggesting that the phosphate ions did not participate to the PbSCC mechanism at the cracking front. This result therefore is consistent with the theory that phosphate sequesters Pb ions from the SCC medium. Additionally, the composition of the oxide at crack terminations was similar for environments with low Pb (due to phosphate addition) and high Pb, with an approximate value of 1 – 3 wt.%, indicating the Pb transport at crack tips was dominated by a migration process rather than diffusion from the bulk environment.

Details

Original languageEnglish
Awarding Institution
Supervisors/Advisors
Award date1 Aug 2021