ABSTRACT
Recent tests by Cowling and Hancock [118] have shown clearly that, taken to extremes of ductility and configuration, the onset of crack growth does not occur at a unique critical value of COD even for a given thickness. No doubt the well known difference in triaxiality as between (say) a centre-cracked and deep-double-edged notched plate in tension for rigid plastic material is the root cause of the different fracture responses. Nevertheless, for many practical purposes and perhaps with some restrictions on size, configuration and extremes of hardening or nonhardening (as discussed in connection with design, in Chapter 5, and testing, in Chapter 3), the concept that COD measured under rather high constraint is a useful lower bound on fracture behaviour, remains very plausible.
