Stress corrosion test methods

In the hope that some of the pitfalls may in future be avoided, the paper discusses problems associated with the various methods of stress corrosion testing, without providing a catalogue of test methods. Thus, the use of time to failure as a parameter that measures susceptibility may lead to erroneous conclusions, whilst designers may be misled if they are not aware that KISCC values, as sometimes determined from pre-cracked specimen tests, are not as reproducible as those design parameters, such as yield stress or ultimate tensile stress, with which they are more familiar. The effects of specimen surface area, and its relation to solution volume, as well as surface finish, are among other quantities that may cause discrepancies between results from different laboratories. Similarly, in relation to environmental aspects of testing, the dangers in the use of 'standard' solutions are indicated and the necessity, in simulating service failures, of precisely reproducing the composition of the environment and the relevant electrode potential are shown, as well as the importance of other factors.