corrosion. The large surface-to-volume ratio of the specimen was chosen with a view to decreasing the hydrogen evolution time and increasing the sensitivity of the measurements.

Hydrogen evolved from the corroded specimen with controlled heating in an inert atmosphere and was measured by a gas chromatograph. The specimen was placed in a 10-mm-diameter quartz tube and was held in place by an inert porous bottom (quartz wool). The tube was inserted in a temperature-controlled (±1°C) vertical furnace and was heated at a controlled rate. A continuous, high-purity nitrogen flow was maintained through the tube at a rate of 20 mL/min and was then driven to a gas chromatograph equipped with a TCD detector. Calibration runs were performed using standard H2-N2 samples. Blind experiments were conducted with an empty tube heated to 600°C and no hydrogen was detected.

The rate of heating is an important parameter in the experiments. A very slow rate (typically 0.5°C/min), followed by extended constant-temperature ramps, was used initially to accurately detect the temperatures at which hydrogen first appears. This procedure was adopted to avoid overlap of desorption fields from different trapping states. However, quantitative estimate of total amount evolved at each stage is awkward with this setup; small quantities of hydrogen evolve for extended periods of time, and the integration of the concentration vs. time curve is not reliable. Thus, after detection of the temperatures at which hydrogen first appears, the bulk of the experiments was conducted at an optimum heating rate (5-6°C/min), which permitted separation of the peaks and reliable integration.

9.6.3 Results and Discussion Microstructural Characterization of Corrosion

Performed metallographic and stereoscopic corrosion analysis has shown that in all accelerated laboratory corrosion tests, corrosion developed gradually from pitting into intergranular attack. In contrast, outdoor exposure for durations up to 24 months did not lead to appreciable corrosion damage. Protection by anodizing and sealing decreased significantly the rate of corrosion attack. Characteristic results are shown for the EXCO test as well as the alternate immersion test for alloy 2024-T3. Pitting density (in pits/m2) and maximum depth of attack for the EXCO test are given in Fig. 9.5. Between 5 and 24 h, the pitting density increased slowly. Between 24 and 36 h, pitting density increased even more, and after 48 h exfoliation corrosion commenced. The depth of attack increased with exposure time (Fig 9.5) and reached a value of 0.35 mm after 96 h of exposure. Characteristic micrographs of sectioned specimens are shown in Fig. 9.6 a, and b. Intergranular attack (cracks) (Fig. 9.6a) developed in the area adjacent to the corrosion pits. For long exposure times (>72 h), the cracks run parallel to the specimen surface (parallel to rolling direction) resulting in macroscopic exfoliation of the specimen (Fig 9.6b). The derived maximum depth of attack of all alloys investigated in this study to exfoliation corrosion solution are summarized in Table 9.38. These values were used as reference to machine the corrosion-attacked material surface for performing tensile tests to investigate whether the corrosion-induced tensile property degradation is volumetric. For specimens exposed at the alternate immersion test pitting density (in pits m2) the maximum depth of attack is given in Fig. 9.7. The rate of increase in pitting density was lower than in the EXCO test, but it was roughly constant

0 0

Post a comment