Mechanisms which have been proposed for the propagation of stress-corrosion cracks are reviewed, and it is concluded that the simplest are the electrochemical mechanisms proposed by Logan and by Hoar and Hines. These attribute crack formation to yielding at the crack tip, which leads to concentration of attack at that point. Mechanisms that involve mechanical stages of crack propagation provide a satisfactory explanation of a number of cases in which there is evidence that, or where it may plausibly be argued that, paths that are mechanically weak exist in the material. These mechanisms are less satisfactory in cases where the cracks do not follow pre-existing paths in the material, as it is difficult to explain why localized brittle failure should occur. The conditions under which the two electrochemical mechanisms are applicable are discussed in terms of a simple crack model, and it is shown that most, if not all, of the facts concerning stress-corrosion cracks in austenitic Cr-Ni steels can be explained qualitatively in this way; in some cases reasonable quantitative agreement is obtained. Logan's mechanism, which assumes that yielding at the crack tip merely prevents formation of a passive film, cannot be applied in this case, but it may well operate in cases where more active metals are involved. The Hoar/Hines mechanism, according to which yielding reduces the activation polarization for anodic dissolution at the crack tip, is satisfactory, and it is concluded that this is the best available explanation. Other cases of stress-corrosion cracking are discussed.
Read full abstract