observation methods. With conventional microscopes, such as optical microscopes (OM), transmission electron microscopes (TEM), and scanning electron microscopes (SEM), however, successive, quantitative three-dimensional observations of the crack nucleation portion in the specimen surface could not be conducted. In most of these studies, the crack-initiation mechanisms were discussed qualitatively. Since the surface morphology of materials can be observed with atomic-scale resolution, the scanning atomic force microscopy (AFM) is a powerful technique to study mechanisms of fatigue and fracture of solid materials. Nakai and his co-workers studied fatigue slipbands, fatigue crack-initiation, and the growth behavior of micro-cracks in a structural steel [3], and α-brass [4, 5, 6]. In our previous study for α-brass, slip-band formation and fatigue crack-initiation processes were observed by means of AFM [4, 5]. Surface of a fatigued specimen was observed at the maximum stress, unloading state, and the minimum stress at the same number of cycles. In the initial stage of the fatigue process, slip-bands, which were formed only under tension stress or compression stress, were observed. These kinds of slip-bands, however, disappeared shortly. Under tension stress, cracks could be detected easily with the AFM image just after their initiation. Before crack-initiation, the height and the width of extrusions and the depth and the width of intrusions gradually increased with the number of cycles. When slip-bands developed cracks, one of these sizes was changed drastically, where sizes to be changed depended on the slip-band angle relative to the stressaxis and the shape of slip-bands. The shape of slip-bands tips was also observed, and they were compared with the results from the continuum distributed dislocation theory. The slip-bands had steep slope when they were blocked by grain boundaries, and the slip-bands descended by gradual slopes to plain surfaces when they terminated within grains [6]. In the present paper, slip-band formation and fatigue crack-initiation processes in α-brass were observed by means of AFM, and the conditions for the crack-initiation was discussed. EXPERIMENTAL PROCEDURE The material for the present study was 70-30 brass (α-brass). The chemical composition of the material (in mass %) was as follows: 69.92 Cu, 30.07 Zn, 0.0071 Fe, and 0.0026 Pb. After the specimens were made by the electric-discharge machining, they were heat treated at 320◦C for 180 s (Material A), or at 850◦Cfor 3600 s (Material B). The grain sizes of Material A and B were 20 µmand 1,100 µm, respectively. Before fatigue tests, surface of the specimens were electro-chemically polished. The specimen has a minimum cross-section of width 8 mm, and a thickness of 3 mm. It has a weak stress concentration with the elastic stress concentration factor of 1.03 under plane bending [4]. Since most of fatigue cracks were not initiated from the shallow notch root, the stress concentration factor was not considered in the calculation of the stress amplitude. The fatigue tests were carried out in a computer-controlled electro-dynamic vibrator operated at a frequency of 30 Hz under fully reversed cyclic plane bending moment (R=−1), or pulsating bending moment (R=0). To conduct a quantitative analysis of the development of fatigue slip-bands, the scanning atomic force microscopy (AFM) was employed for the present study. The scanning area for the observations was 30 µm30 µm. Since it was very difficult to identify in advance where fatigue cracks would be initiated, replicas of the specimen surface were taken at the predetermined number of fatigue cycles. The replica films were coated by gold (Au) before observation. The replications were conducted at the maximum tensile stress. Although the height of the surface in the replica film was reversed from the specimen surface, the height of the replica film in the AFM images was reversed by an image processing technique. EXPERIMENTAL RESULTS Transgranular Cracking
RkJQdWJsaXNoZXIy MjM0NDE=