In the early 1970s, I spent a few years working as an engineer performing device failure analysis at an aerospace company in Maryland. The labs were fully instrumented for complete device evaluation by electrical, optical, structural, mechanical, and chemical means. There was keen interest in determining exactly why every failure occurred. Most of the devices that were analyzed were semiconductor-based integrated circuits and transistors used in electronics. Every once in a while, however, a job of a different sort would show up for analysis.

I was given the task of determining why an electrical-resistance heating rod had failed. Rod dimensions were about 4 ft long with a 1-in diameter. After determining the location of the defect, and using a brick saw to cross-section the rod, I was able to determine where the rod had an electrical short. This was likely due to a defect in the rod, which, under electrical and thermal stress, allowed metal to bridge the gap between parallel, oppositely running current lines of the heating element. That is, the lowered heater resistance allowed more current to flow than the design allowed. The standard arrangement for preventing thermal runaway and protecting the rod from overheating by shorted condition was to use a circuit breaker. Unfortunately, this rod had no protection and, as a result, was able to heat up way beyond design limits. Herein lies the real story. The night watchman who discovered the faulted condition was curious about a dripping noise that he heard in one of the large bays that were part of his nightly route.

As a bit of background, the labs used gaseous dry nitrogen for things like backfilling a chamber to clear out humid air or oxygen. The source of the nitrogen gas started with a tank that fed liquid nitrogen into a pipe manifold, which then fed out nitrogen gas. As the liquid heated up, nitrogen gas was generated. To generate gas faster, the manifold was placed next to heating rods. To make better thermal contact between the manifold and the heating rods, they were both inserted into opposite ends of an aluminum cylinder that had been filled with aluminum dust. The cylinder was about
6 ft long and 1 ft in diameter and was bolted to the bay ceiling.

The dripping sound that the night watchman heard that night was molten aluminum falling from the bay ceiling onto the floor. This resulted from the overheated heating rod, which had no electrical protection, generating so much heat that it melted the aluminum dust and the container wall. Aluminum melts at 660 °C or 1,221 °F. Although it would have cooled somewhat from falling through the air, it was still much hotter than boiling water.