In the initial set of experiments, the boundaries had a residue of sub-micron-size alumina particles from metallographic polishing. Later, specimens were prepared without this residue [10], and still later by growth from the melt, also without particles [11]. The boundaries in this set of specimens were found to be difficult to crack, even along the fast-diffusion direction. Two conclusions were drawn from this. The first is that dynamic embrittlement may well follow the pattern of other types of grain-boundary failure [12-14] in which boundaries with Σ-value less than about 29 are particularly resistant to failure. The second conclusion is that the presence of particles probably enhanced the surface concentration of tin by decohering ahead of the crack and allowing tin to segregate to the new surfaces thereby created. They would thus have served a role analogous to that of the small sulfides in the steel. The influence of grain-boundary particles is an area that deserves further study. EMBRITTLEMENT FROM THE ENVIRONMENT Motivated by reports of oxygen-induced intergranular cracking in Cu-0.25%Be [15] and in nickel-base alloys, particularly alloy 718 [16], experiments were carried out to see if these types of cracking were analogous to the phenomena described above. When the steel and the Cu-Sn alloy were loaded at fixed displacement as notched bars in vacuum at elevated temperatures and the Cu-Be and alloy 718 were loaded similarly in oxygen, essentially the same behavior was found in all cases. The crack velocity could be calculated from the loadrelaxation curve and the relation between specimen compliance and crack length. An example of the effect of oxygen pressure is given in Fig. 3. 10-8 10-7 10-6 10-5 10-4 30 40 50 60 70 da/dt (m/sec) K (MPa-m 1/2 ) 50 Pa O 2 105 Pa air 0.1 Pa O 2 Fig. 3 Crack velocity vs. stress intensity curves derived from the load-relaxation curves and the compliance-calibration curve for alloy 718 at 650˚C in various pressures of oxygen [17]. Examination [17] of the fracture surfaces of the alloy 718 showed the typical decohesion appearance when the oxygen pressure was high enough. However, at very low pressures the growth rate dropped by several orders of magnitude, and the fracture appearance changed to a series of striations marked by strings of oxide, indicating that the cracking occurred in jumps at time intervals long enough for significant oxidation at the crack tip. The cracking at
RkJQdWJsaXNoZXIy MjM0NDE=