A critical evaluation of mechanistically based modelling of environment-assisted cracking has been made. Models based on anodic reaction, namely slip-dissolution, anodic reaction-induced cleavage and surface mobility are evaluated from both a mechanistic and quantitative perspective. The film-rupture slip-dissolution mechanism is increasingly perceived to be relevant only to intergranular cracking though not as an exclusive mechanism for this process. The basic principle of film-induced cleavage has been demonstrated experimentally but the range of applicability of the mechanism remains contentious and requires further work. The concept of the surface mobility model is interesting but the quantitative crack growth model is based on unsatisfactory assumptions. In relation to models of environment-assisted cracking based on the effect of hydrogen atoms it is evident that for the non-hydride forming systems the hydrogen enhanced localised plasticity mechanism is gaining wider acceptance though studies of Fe-Si single crystals suggest that for some systems a cleavage mechanism of crack advance may prevail. Quantitative modelling of crack growth based on anodic reaction processes is reasonably well developed, though with a need to provide information about crack-tip electrochemistry and strain rates. These models are comparatively simple but modelling of cracking due to hydrogen atoms is more complex, for example, because of the diffusion and trapping of hydrogen atoms in the metal. Accordingly, models of crack growth in this case tend to be based on rate-limiting processes and in corrosion fatigue in particular there is a need to integrate the chemistry, mechanics, transport and materials interaction in an improved quantitative framework. Mechanistic modelling has an important role in understanding and predicting environment-assisted cracking but the application to life prediction may be limited to specific systems because of the complexities of multiple nucleation events, crack coalescence, crack-tip shielding, crack branching and complex stresses. The intelligent combination of empirical modelling supported by insight from mechanistic modelling provides the most effective approach to predicting environment-assisted cracking in metal-environment systems.
