A “Toolbox” for Forensic Engineers

Failure Due to Manufacturing Faults 151
solely with the foundry. Its insurers accepted liability and settled the claim
in full.
5.5 Failure of Threaded Iron Pipe Coupling
Another case concerns a different type of cast iron where something went
wrong in the foundry.
The casting in question was a small straight coupling, joining two lengths
of 1-in. (25-mm) iron pipe in an oil line to a central heating installation in
an office block. A persistent odor of oil in the boilerhouse had been reported,
so a service engineer was sent to investigate. He determined that the source
of the smell was a slight seep of oil from the threaded pipe coupling close to
the burner, so he tightened the union until the ooze appeared to have stopped.
He packed up his tools and departed. That night there was a serious fire that
caused a great deal of damage. The source was found to be an escape of oil
through a split at one end of the pipe coupling.
There were deep indentations on one flange cut by the grips of a large
stilson type wrench, the indentations being consistent with the direction of
tightening. The wall was split right through from the end with this flange to
halfway along the body of the fitting. A metallographic section cut transversely
across the split revealed a structure characteristic of a malleable cast
iron — of the variety called blackheart, which consists of graphite rosettes
in a ferritic matrix — except for a narrow band on either side of the split,
which exhibited a flake graphite structure. The casting should have solidified
with all the carbon present as iron carbide (cementite, Fe 3C) which is the
essential starting condition for the malleabilizing process. When the castings
are put through the malleabilizing heat treatment the purpose is to cause a
controlled breakdown of the cementite constituent into iron and graphite,
but the graphite formed in an entirely different way from the flakes in a
common gray cast iron. Figure 4.5D shows this form of microstructure. It
appeared as ragged clumps or rosettes surrounded by virtually pure iron. For
some reason this particular casting had solidified with a band of flake graphite
structure through the wall, which could not change its physical form during
the heat treatment that broke down the iron carbide. Thus after the coupling
had undergone the malleabilizing heat treatment, this remained as a narrow
band of brittle material extending right through the pipe wall.
The cause of the failure was obvious. The service engineer had tightened
the coupling to stop the ooze, by screwing the coupling further onto the
tapered thread. He had applied far too much torque by using unnecessarily
large stilson grips, which cracked the band of flake graphite. Shortly after he