Because properties are highly dependent on the weld area microstructure, weld procedure qualification (WPQ) testing of ferritic-austenitic steels normally includes metallographic evaluation. This entails demonstration that the ferrite/austenite balance is within set limits, together with little or no intermetallic phase formation, although other microstructural changes such as nitride formation or secondary austenite are not generally recognised.

The imposition of microstructural evaluation leads to problems in qualifying a welding procedure, because quantitative limits vary between different specifications while acceptance criteria may be conservatively and unrealistically set. Moreover, quantitative metallography requires considerable skill, is subjective, and is prone to inherent statistical variation.

The paper discusses the philosophy of incorporating quantitative metallography into WPQ testing. Limits of measurement accuracy are indicated, and it is pointed out that service properties are not directly relateable to the metallographic criteria generally specified. It is concluded that acceptance of a given weld procedure should be based primarily on measurement of the mechanical and corrosion properties of practical concern, and that microstructural evaluation should not be mandatory. However, metallographic investigation remains of potential value should service entail characteristics not evaluated in WPQ testing, and to indicate the likely tolerance of the welding procedure to variation in conditions used.