Testing and Modelling for Prediction of Hydrogen Embrittlement

A critical evaluation of testing and modelling for prediction of hydrogen embrittlement has been undertaken. Exposure time, temperature excursions and mechanical test method are shown to be factors which affect the reliable experimental determination of threshold stress/stress intensity factor for cracking and crack growth rates. In relation to test time, the key issue is the distance of the site of cracking from the primary source of hydrogen atoms, which will be influenced by exposure conditions, test specimen configuration and activity of the alloy. In some systems very long tests may be inevitable to ensure steady-state hydrogen uptake has been attained. Excursions from high to low temperature can be important in service and have been shown to be of significance in laboratory testing, but further studies are required. Mechanical test methods based on dynamic straining will provide a more conservative estimate of the threshold stress/stress intensity factor for cracking. It is concluded that guidelines for hydrogen embrittlement testing are needed to complement existing test standards.Mechanistic modelling of hydrogen embrittlement threshold and crack growth kinetics is fundamentally challenging. Analytical models have been developed but with questionable assumptions and a limited range of application. The primary obstacle is the need to predict the time-dependent distribution of hydrogen atoms at the crack tip as a function of test and operational variables. Progress has been made but existing models have limitations. Nevertheless, if further progress is to be made it is in the development of more robust models of crack-tip hydrogen coupled with a criterion for failure.