ICF13A

4 Table 2 Loading Conditions in NNpH Solutions (All were sparged with 5% CO2) Condition Peak Stress Stress ratio Frequency No. of cycles A 100% SMYS 0.8 1×10-4 Hz 1,345 B-1 100% SMYS 0.63 7.6×10-5 Hz 1,394 B-2 106% SMYS 0.63 7.6×10-5 Hz 1,394 After testing in the solution for the desired number of cycles, the tests were stopped and the samples were taken out. After cleaning using EDTA and ethanol in the ultrasonic cleaner, the samples were rinsed and dried with warm air and the sample surface observations were made using optical microscopy (OM) and scanning electron microscope (SEM). Thereafter, the sample surfaces were sectioned step by step to examine crack and pit depths, and crack and pit shapes. The sectioning was perpendicular to the cracked and pitted surfaces, and the sectioning progressed perpendicular to the loading direction. After each grinding and polishing step, measurements were made using OM and SEM to reveal the crack and pit dimensions. 3. RESULTS AND DISCUSSION 3.1. BLUNT CRACK INITIATION AND CRACKS DISTRIBUTION When pitted specimens were cyclically loaded in NNpH solutions, many blunt cracks were seen to initiate from the bottom of the pits (Figure 2). These pits acted as stress concentrators (micronotches) and were the principal sites for crack initiation. The pits would locally raise the stresses near the pits. Thus, plastic deformation would occur locally, and there would be plastic zones near the corrosion pits. The iron atoms in the plastically deformed zones would have enhanced energy because of the increased dislocation density and increased stress, and would behave as anodes in the solutions relative to the less deformed and lower stressed nearby regions. Thus, stress cells would form leading to localized corrosion. So a preferential electrochemical attack occurred in the regions where plastic deformation was localized. Thus, dissolution in the plastic zones around the corrosion pits was accelerated. This, in turn, would further enhance the stress concentrations so that each cycle would create more deformed material to be removed by further corrosion. In the end, blunt cracks were initiated. These cracks were very wide, especially at the crack mouth, as shown in Figure 2. Hence, they were designated as blunt cracks or Stage I cracks. At the same time, in contrast to the fatigue cracks in air, the crack advance direction was perpendicular to the applied load direction. After cyclic loading in NNpH environment, it was clearly seen that most of cracks were distributed between 50 and 450 μm in depth, as shown in Figure 3a, and the most probable crack depth was around 150 to 250 μm. Figure 3b shows that the median of crack depth was at 221 μm, and the lower and upper box values (quartiles) were 156 and 351 μm, respectively. The whisker was from 61 to 633 μm with a few outliers shown with asterisks. All the evidence once again reflected that the majority of crack depths were below approximately 0.5 to 0.6mm, consistent with the field observations.

RkJQdWJsaXNoZXIy MjM0NDE=