Spotlight: How to... find success in failure

Materials World magazine
,
1 Apr 2018

In a typical year, Axiom Engineering Associates might undertake 50 metallurgical failure investigations. Steve Woodward takes a tour of its black museum of failure. 

Advertising failure is not generally considered to be good business practice. However, doing just that has seen Axiom grow from three employees in 2003 to 53 in 2018. Our black museum of failure lurks ominously in the corner of the conference room and has long been the focal point of tea break chats. Our clients enjoy nothing more than poring over these mysterious objects, in the hope of learning from the misfortunes of others, and hoping that it doesn’t happen to them. The images in this article offer a glimpse into the fascinating world of failure and the pitfalls that might blight those unaware of how to avoid them.            

Exhibit A

Plant operators often labour under the misapprehension that their boilers are designed to heat water. However, any metallurgist will tell you that, actually the water is there to cool the tubes. If cooling efficiency diminishes for any reason, problems can arise very rapidly, as seen in Exhibit A where high heat flux has resulted in steam blanketing, wick boiling, internal oxidation and decarburisation due to high temperature hydrogen attacks. Note the decarburised fringe with outwardly radiating ‘tendrils’.        

Exhibit B

Under-base corrosion of a large storage tank was remedied by internal over-plating. Image B is a polished and etched macro section of the base-to-shell junction showing a substantial weld deposit at the edge of the new plate. Unfortunately, 1950s vintage carbon steel plates are prone to harbouring elongated manganese sulphide inclusions, which form planes of inherent weakness. No flaws were revealed at the time of the repair but residual welding contraction stress was sufficient to eventually generate shell plate delamination cracks where after cyclic loading initiated a fatigue crack in the root of the external shell-to-base weld. The crack ran around 60% of the circumference, and when discovered several years later it was within 2mm of a full breach. This is one of the reasons why Aboveground Storage Tanks Inspectors, for example, recommend floor plate repairs should pass under the shell.   

Exhibit C

When you have eliminated the impossible, whatever remains, no matter how improbable, must be the truth. This was certainly true in this case, and although (due to budget constraints) we were not able to prove unequivocally that the nitrates in birdlime were responsible for stress corrosion cracking (SCC), there was little else it could have been. The defects in question occurred on a lofty carbon steel steam elbow that had remained un-lagged during construction and where it also served as an avian perch. Nine years later, following rain ingress, the lagging remained wet for long enough to initiate external nitrate SCC.     

Exhibit D

Fuels that feed waste-to-energy boilers harbour aggressive constituents that can wreak havoc on a furnace tube’s surface, causing rapid metal loss by various deleterious processes. Owners may achieve a 10-year tube life, but with those less fortunate having to overlay or replace tubes after only around five years. Looking to the interior, boiler feed water may contain dissolved copper, which plates-out in metallic form on bore. Copper melts at 1,085°C − a temperature that is easily achieved when welding on the outer surface of a substantially thinned tube. Liquid metal embrittlement of iron by copper is a widely known phenomenon, what’s perhaps more unusual is why we don’t see it more often. The meandering (1.5mm long) gold coloured line on this etched sectional photomicrograph is intergranular copper in a weld overlay and was situated immediately adjacent to one of many wall penetrating cracks. The lesson − don’t let your tubes thin to such an extent that you dare not weld them while water filled.    

Exhibit E

No metallurgical black museum would be complete without at least one example of chloride stress corrosion cracking (CSCC). The first occurrence was not formally recorded but doubtless it followed very soon after the invention of austenitic stainless steel (SS) in 1906 – and probably the first time it was used to contain hot water. This photomicrograph is very recent, illustrating, again, that upgrading from carbon steel to 316 SS is not the universal panacea for corrosion issues – especially in the case of this vessel, which was exposed to hot brine. I have avoided stating the parameters required to initiate CSCC as they are still the subject of debate, but it requires stress, halides and heat, and in surprisingly modest quantities. Perhaps the most alarming aspect of CSCC is the rapidity with which cracks propagate, having been measured at speeds of metres per year.     

Exhibit F

Graphitic corrosion (GC) is a long-term deterioration mechanism peculiar to grey flake cast iron and in the UK may be referred to as graphitisation due to the texture of the corrosion product. Not to be confused with graphitisation, which describes the high temperature (>425°C) microstructural deterioration of carbon and low alloy steels. GC of cast iron occurs at ambient temperature and affects buried components exposed to groundwater that is highly conductive, acidic, soft, and possibly brackish or contains hydrogen sulfide. Most corrosion processes result in dimensional change but as evident here, GC is characterised by the uniformity of cross-section, and although obvious in a polished cross-section, it can be almost invisible to the casual observer in the field. This morphed material has little or no strength, hence GC often results in catastrophic failure, which is of particular concern to utility companies that own thousands of miles of buried water mains and sewers. The conventional view of GC describes it as an electrochemical leaching process where de-alloying of the iron leaves in its wake a mixture of graphite flakes and corrosion product. Recent studies have revealed that, although well documented, this mechanism is not yet fully understood.