STP939: Recent Studies of Crack Initiation During Stress Corrosion Cracking of Zirconium Alloys

    Cox, B
    Atomic Energy of Canada Limited, Chalk River Nuclear Laboratories, Chalk River, Ontario

    Haddad, R
    Atomic Energy of Canada Limited, Chalk River Nuclear Laboratories, Chalk River, Ontario

    Pages: 17    Published: Jan 1987


    Abstract

    In order to study crack initiation, we had first to develop a technique for reproducibly obtaining large numbers of very small crack initiation regions. The previously used test methods of stressed split-rings or pressurized (or compressed) tubes gave low success rates. Specimens either showed through-wall cracks, or no cracks, with few incipient cracks being observed. This suggested that the propagation time was short compared with the incubation time. Attempts to overcome this by unloading specimens following the first acoustic emission signal still gave a low frequency of incipient cracks.

    The possibility that an “active path” mechanism was the cause of the difficulties in obtaining incipient cracks reproducibly was investigated by a stress-pulse technique. The results were apparently consistent with an “active path” developing ahead of the crack tip, and cracking during the stress-pulse. Two iodine tracer techniques, one using radioactive 131I, the other using natural 127I (located by Rutherford backscattering), showed that iodine did not diffuse ahead of the crack tip. When specimens were not statically loaded, but merely given a stress-pulse, the results showed that the appearance of an “active path” mechanism resulted from the length of time taken for the chemistry inside the apparatus to reach a condition capable of initiating cracking. When the system was allowed to come to chemical and thermal equilibrium before the specimen was loaded, cracking started almost immediately (<2 min). This virtual disappearance of the incubation time with adequate equilibration of the apparatus was equally true for experiments in I2 or Cs/Cd vapours. It was concluded that, if the chemistry of the system could be rapidly equilibrated, there would probably be effectively no incubation time and cracking would start immediately.

    Many small incubation sites were produced in these tests. No second-phase particles were seen at these sites in the SEM, and no local contamination of the initiation site was seen by energy dispersive X-ray analysis (EDAX). The initiation sites were individual grains (or small groups of grains) which fractured transgranularly to give a fractography otherwise identical with transgranular fracture surface produced during crack propagation in the bulk of the specimen. It was concluded that localized residual stresses, when added to the applied load, cause a few correctly oriented grains at the surface to crack immediately upon loading, given that the “correct chemistry” was present.

    The presence of organic-iodine compounds (in particular methyl iodide) as impurities in the vapor phase was found to act as a promoter of crack initiation, giving a large increase in the frequency of initiation sites and an increase in crack velocity. Methyl iodide also gave a characteristic fractography which was explained as a modification of the propagation mode of the crack front.

    Keywords:

    Zircaloy, Iodine, Cesium/Cadmium, Stress corrosion cracking, pellet-clad interactions


    Paper ID: STP28155S

    Committee/Subcommittee: B10.02

    DOI: 10.1520/STP28155S


    CrossRef ASTM International is a member of CrossRef.