Theoretical and experimental evidence has suggested that fast fracture in plated structures is controlled by structural behavior and material properties. Recent work in the United Kingdom Atomic Energy Authority involving tests to failure on 5-ft diameter, 14-ft long, cylindrical steel pressure vessels containing 6 to 24-in. long slits, confirms that for constant-load conditions (such as those existing in a pneumatic test) crack arrest does not occur; while under conditions which are nearer to the constant-strain case (such as those in a hydraulic test), the extent of crack propagation is controlled by fracture toughness and total energy available. In addition, this work indicates that a correlation of the form f03/2 = function of (ϕ fu , fu) exists between the critical parameters, crack length l, gross stress fg , Charpy energy absorption ϕ, yield stress fy, and ultimate tensile stress/fu. When considered in conjunction with the relationship between the main parameters and plastic zone size, this correlation suggests that, in addition, the critical condition is affected by the path along which instability is approached; for instance, by increase in crack size, increase in stress, or change in material properties.
Since energy availability controls fracture appearance through plastic zone size, strain rate, and crack speed, cleavage and shear fracture may exist side by side at the same temperature, and it is suggested that the interpretation of service and test failures of steel structures should be based on an analysis of factors such as load history, stress level, flaw size, and Charpy value, rather than on a fracture appearance temperature approach. The safety implications of these effects in relation to proof testing, annealing out of irradiation damage and material monitoring are discussed.