Some of the earliest problems of weld metal
corrosion related to ships in arctic waters, where
the action of ice abraded the paint to expose bare
steel and damaged the anodes, thus rendering the
cathodic protection system ineffective. In these
cases, it was observed that enhanced corrosion
of the weld metal was due to electrochemical
potential differences between the weld metal
and the base metal, such that the weld metal is
anodic in the galvanic couple. Further detailed
studies were undertaken in the late 1980s to
assess more modern steels and welding consumables
in arctic waters off Canada (Ref 7). Both
HAZ and weld metal attack were observed, and
the general conclusions were that for steels
between 235 and 515 MPa (34 and 75 ksi) minimum
yield strength, high manganese content
(1.4%) in the parent steel resulted in enhanced
preferential HAZ attack, but this could be
reduced via increased heat input during welding.
Generally, the rate of weld metal attack was
dependent on the nickel and copper contents of
the welding consumable and was less influenced
by parent steel composition, although a steel with
copper, nickel, and chromium additions led to a
more noble parent steel, hence accelerating weld
metal attack. It was noted that parent steel with
low silicon content led to increased weld metal
corrosion, supporting the earlier findings that silicon
< 0.2% can be detrimental, but the opposite
was observed for silicon in the weld metal.