A published analysis of crack growth rate under ohmic control in a slow strain rate test is modified slightly to take account of a crack growth mechanism involving periodic cleavage or brittle intergranular fracture nucleated by a dealloyed layer. Assuming that a critical layer thickness must form to trigger each cleavage event, and that a surface-dealloyed layer is available to start the process off at zero strain, we obtain the following result: at low strain rate (s), the dependence of crack velocity (v) on s is close to v ~ s0.5, while at high strain rate the slope increases smoothly until it becomes v ~ s. At the highest strain rates we are not really dealing with stress corrosion cracking, since all that is happening is a single brittle event at the start of the test, followed by plastic strain to failure. The results of this calculation are consistent with data published by Serebrinsky and Galvele, who interpreted the behavior using the surface mobility model, even at strain rates higher than 1,000%/s.

You do not currently have access to this content.