Last summer, I was asked to help determine why a relatively new heat exchange that was part of an industrial boiler system had started springing leaks left and right. Since my client knows that I am more of a generalist-type failure analyst rather than a boiler-tube expert, I was asked if I thought I could do the work. I knew that it would take longer than my usual work, such as a fatigue crack in a conveyor system or a threaded fastener fracture during installation into an automotive assembly. But I told them that I thought my knowledge of their company would make it easier for them to take away useful advice from the technical information. They decided to give me a try.
Well, it was quite an experience. I have worked with high-temperature metallurgy a bit and corrosion problems more than a bit. One of the defined parts of my task was to review the other three reports that had been prepared since the problem first appeared. I found it extremely interesting that three reports performed by different individuals apparently reached three totally different conclusions as to the cause.
In one of the reports, the chemist had clearly misidentified some of the peaks on the EDS microchemical analysis, claiming that there was a huge buildup of calcium when, in fact, the peak represented carbon. This, of course, predisposed me to weigh the rest of their data carefully before making any judgment about the balance of their conclusions. The other two reports appeared to at least have correct basic data, but the conclusions just did not make any sense. The cause of the failure in one was said to be “corrosion fatigue.” I thought I knew what corrosion fatigue was, and this surely wasn’t it!
A few years ago, when teaching a fracture-analysis course, one of the students asked about corrosion fatigue. After consulting the ASM Metals Handbook and multiple discussions with colleagues, it became apparent that this was a harder question than it seemed at first. It turns out that a technically correct diagnosis of corrosion fatigue for an automotive or other ambient-temperature application is something that can be made only in a laboratory! This is because corrosion fatigue in these cases is defined as a crack that grows by a certain quantitatively defined amount FASTER than the same crack would grow in the same LOADING conditions without the corrosive substance. Furthermore, the aggressiveness of the corrosive substance must also be carefully quantified before assigning corrosion fatigue as a “mechanism.”
In an automotive application in real life, this is very unlikely to happen outside of a laboratory or other specifically controlled test experience. The exact corrosive substances and exact loading conditions are rarely that well known in actual service. So in a situation where a fatigue crack is present and the fracture surface is somewhat corroded, one might correctly speculate that the crack propagated by corrosion-ASSISTED fatigue, not corrosion fatigue. In this type of corrosion fatigue OR corrosion-assisted fatigue, somehow the corrosive environment is weakening the atomic bonds at the tip of the growing crack and making it easier for the crack to grow faster than it otherwise would.
In boiler failure-analysis “lingo,” corrosion fatigue means that corrosion has attacked the base material in question, leaving sharp-bottomed pits. Usually, the formation of sharp- (rather than rounded-) bottomed pits is a result of a water-treatment problem. The sharp-bottomed shape of the pits themselves MAY make it easier for fatigue cracks to propagate from the sharp corners. If cracks do initiate in this way, they may soon find themselves at the opposite side of the tube wall, effectively creating a leak path. But even this scenario is not what boiler failure analysts mean by corrosion fatigue.
What they mean is that the OXIDE (“corrosion product”) FILLING the sharp-bottom pits cracks as a result of mechanical or thermal (or thermomechanical) stress. Before cracking, the oxide actually protected the material at the bottom of the pit from further rapid attack. But now, in the cracked condition, the chemically aggressive water can get right to the freshly exposed metal at the pit bottom. New oxide can now form deeper in the wall and can then crack at the next stress cycle.
Well, that was interesting! Corrosion fatigue is a damage mechanism in boiler tubes that works by damaging the “rust!" In other more generalized applications, corrosion fatigue attacks the actual base metal. No wonder the poor plant-operations people could not make sense of what was going on. Each little island of specialists makes up their own vocabulary without even telling people that they are using a word that has a totally different meaning that may be much more commonly used than the one they are invoking, and possibly “by their lunch-table neighbors” working a different fracture in their own plant.
Next time, we will take a look at photos of six different failures.