Stability, validity, and sensitivity to input parameters of the slip-dissolution model for stress-corrosion cracking

The slip-dissolution model of stress-corrosion crack growth is reviewed and developed from several points of view: the differences between ‘discontinuous’ (Vermilyea) and ‘continuous’ (Ford, Andresen, Shoji) versions of the model; stability and possible multiplicity of predicted crack velocities for given mechanical parameters, and the sensitivity of the predicted crack velocity to small variations in the electrochemical and mechanical parameters. We find that for relatively steep anodic current decays on the bare metal surface at the crack tip (for example, i ∼ t−m with m = 0.8), the output of the continuous type of model is extremely sensitive to the strain hardening exponent and to the location or cutoff distance in the logarithmic plastic strain distribution at which the crack-tip strain rate is calculated. Difficulties also appear because this distance is likely to be a function of other parameters such as yield stress. The handling of loading rate (K˙) effects in Shoji’s treatment appears unrealistic, leading to a much too weak dependence of the crack-tip strain rate on K˙. However, irrespective of how K˙ is introduced, dual crack velocity solutions are found for negative K˙; the stability of these is discussed.