Strain rate cracking

In Section 6.8, a detailed analysis of the most frequently used impact tests (i.e., Charpy and Izod impact tests) is used to characterize fracture toughness. Temperature, strain rate, crack tip curvature, specimen thickness, annealing, aging, irradiation, and environmental effects are discussed as test variables using the Iramework of fracture mechanics. 

Higher than design strain rates in service (> lO s ) can also cause disruption of the corrosion scale, leading to enhanced metal loss and corrosion-assisted cracking of the substrate "". Laboratory exposures show parabolic... 

Various workers have used equation 8.8, or some modified version thereof, to compare observed with calculated crack velocities as a function of strain rate, but Fig 8.8 shows results from tests on a ferritic steel exposed to a carbonate-bicarbonate solution. The calculated lines move nearer to the experimental data as the number of cracks in equation 8.9 is increased, while the numbers of cracks observed varied with the applied strain rate, being about 100 for 4pp 10 s , but larger at slower 4pp and smaller at higher 4pp. 

Fig. 8.8 Comparison of calculated and experimental crack velocities as a function of strain rate for a ferritic steel exposed to 1 n Na2C03 + 1 n NaHCOj at —650mV(SCE) and 75°C...

If crack propagation occurs by dissolution at an active crack tip, with the crack sides rendered inactive by filming, the maintenance of film-free conditions may be dependent not only upon the electrochemical conditions but also upon the rate at which metal is exposed at the crack tip by plastic strain. Thus, it may not be stress, per se, but the strain rate that it produces, that is important, as indicated in equation (8.8). Clearly, at sufficiently high strain rates a ductile fracture may be propagated faster than the electrochemical reactions can occur whereby a stress-corrosion crack is propagated, but as the strain rate is decreased so will stress-corrosion crack propagation be facilitated. However, further decreases in strain rate will eventually result in a situation where the rate at which new surface is created by straining does not exceed the rate at which the surface is rendered inactive and hence stress corrosion may effectively cease. 

The implications of a significant role for strain rate are wider than the obvious one that stress corrosion should only occur over a restricted range of strain rates. Thus, in constant load tests, since cracks will continue to propagate only if their rate of advancement is sufficient to maintain the crack-tip strain rate above the minimum rate for cracking, it is to be expected that cracks will sometimes stop propagating, particularly below the threshold stress. Such non-propagating cracks are indeed observed below the thres-hold . Moreover, in constant-load or constant-strain tests, the strain rate diminishes with time after loading, by creep exhaustion if the stress remains sensibly constant, and it is found that the stress-corrosion results are sensitive to the relative times at which the stress and electrochemical... 

Fig. 8.10 Effect of strain rate upon the cracking propensity of a Mg-Al alloy immersed in a...

The sometimes contradictory results from different workers in relation to the elements mentioned above extends to other elements . Some of these differences probably arise from variations in test methods, differences in the amounts of alloying additions made, variations in the amounts of other elements in the steel and the differing structural conditions of the latter. Moreover, the tests were mostly conducted at the free corrosion potential, and that can introduce further variability between apparently similar experiments. In an attempt to overcome some of these difficulties, slow strain-rate tests were conducted on some 45 annealed steels at various controlled potentials in three very different cracking environments since, if macroscopic... 

Fig. 8.15 Effects of potential upon the stress-corrosion cracking of various steels in CO3-HCO3 solution in slow strain rate tests (after Parkins et al )...

---