Before closing our discussion from last week, I will share some examples to illustrate the differences between normal ambient-temperature “corrosion-assisted fatigue” and “boiler-lingo corrosion fatigue.”
Figure 1 shows a fracture surface from a kitchen chair that had a rough life (active teenage boy – last name Aliya) in a house where they did not clean the furniture very often and when they did, apparently used harsh cleaners. The chair actually served a long life, but when it broke, revealed some rusty-looking areas on the fracture surface that were found by EDS microchemical analysis to have a strong concentration of chlorine (possible harsh cleaner).
Could this be corrosion-assisted fatigue? The scanning electron micrograph shows nothing unusual in the appearance of the crack (Figure 2). Unfortunately, this low-budget job did not result in a metallographic cross section.
Figure 3 shows a portion of a shaft from a chemical processing plant. It has multiple crack initiations, typical of many rotating, bending applications. Aside from some areas of reddish discoloration, it really does not look very rusty. Unfortunately, this project was done some time ago, and the scanning-electron micrographs have been misplaced.
However, a cross section – including a portion of the crack profile a short distance from the main crack – is shown in Figure 4. Fatigue cracks in general (never say never in failure analysis!) do not spontaneously branch out into multiple propagating tips in this way. Clearly, if the same thing happened at the primary crack, it was being assisted by some sort of corrosion process.
You will note that there is not necessarily much difference in the visual severity of the “rust” in the two room-temperature cracks. Prior to diagnosing corrosion fatigue or corrosion-assisted fatigue, be sure you use the right laboratory tests to know what is going on.
Figure 5 shows an actual boiler-tube section that was diagnosed with what I am henceforth calling “boiler-style corrosion fatigue.” The hole could not be found until a large portion of cross section was ground and polished multiple times. It does not take a very large hole to leak at high pressure. The cross section is shown in Figure 6. Note how the apparent width of the crack is very jagged, possibly a result of uneven oxidation caused by sporadic cracking of the oxide lining the crack.