1990 Plenary Lecture: Strain Rate Effects in Stress Corrosion Cracking

Abstract

Slow strain rate testing (SSRT) was initially developed as a rapid, ad hoc laboratory method for assessing the propensity for metals and environments to promote stress corrosion cracking. It is now clear, however, that there are good theoretical reasons why strain rate, as opposed to stress per se, will often be the controlling parameter in determining whether or not cracks are nucleated and, if so, are propagated. The synergistic effects of the time dependences of corrosion-related reactions and microplastic strain provide the basis for mechanistic understanding of stress corrosion cracking in high-pressure pipelines and other structures. However, while this may be readily comprehended in the context of laboratory slow strain tests, its extension to service situations may be less apparent. Nevertheless, laboratory work involving realistic stressing conditions, including low-frequency cyclic loading, shows that strain or creep rates give good correlation with thresholds for cracking and with crack growth kinetics. Moreover, the time-dependent reduction of strain rates due to work hardening of steels can be related to the time dependences of crack nucleation and growth rates. When these latter quantities are combined in an appropriate way, including the coalescence of cracks, then it is possible to make predictions of the remaining life of cracked structures, a matter of growing importance in many industries, and all underpinned by the concept of strain rate.