550˚C in vacuum, as shown in Fig. 2. Fractography by SEM showed that each burst represented cracking of a number of grain boundaries. (ii) The local rate of crack advance is highly variable, depending on the grainboundary structure, which determines the rate of diffusion of the surface element into the boundary. This results in a highly irregular crack front and in slow-cracking ligaments left behind the main crack front. (iii) The tip of the growing crack is so sharp that no appearance of blunting could be found by examination of the fracture surface at the highest available resolution in the scanning electron microscope. To test the hypothesis that this is a generic form of brittle fracture, similar experiments were carried out on a Cu-8%Sn alloy (without the prior high-temperature treatment). Compacttension specimens loaded at 265˚C in vacuum exhibited behavior essentially similar to the steel [5], including intergranular decohesion and cracking in bursts. In this model material the tin played the role of sulfur in the steel. It is surface active [6], and it is a low-melting embrittling element in steel [7]. The conclusion is that the low creep ductility [8] and the hot shortness [9] exhibited by Cu-Sn alloys are both analogous to stress-relief cracking in steels. Fig. 2 Example of crack growth in bursts in the early stages of sulfur-induced dynamic embrittlement (stress-relief cracking) [1]. To test the idea that dynamic embrittlement requires diffusive penetration of the solid, which depends on the grain-boundary diffusion rate and thus grain-boundary structure, experiments were done on Cu-Sn bicrystals [10]. In specimens with Σ=5 (031)[100] symmetrical tilt boundaries made by diffusion bonding and loaded in vacuum at 265˚C, it was found that a crack could be grown easily along the tilt axis, but not transverse to the tilt axis. This corresponds to cracking along the fast-diffusion direction but not the slow-diffusion direction.
RkJQdWJsaXNoZXIy MjM0NDE=