Duplex and superduplex stainless steels may be welded to carbon steels using one of a range of consumable types. When selecting the most appropriate consumable, the need to have weld metal strength overmatching with respect to the lower strength of the two parent steels should be considered.
The highest weld metal strength probably will be achieved with the use of a superduplex consumable. Alternatively, where such high strength is not required, a 22%Cr duplex stainless steel filler, a nickel-based filler or a 309 type austenitic stainless steel may be used. It should be noted that dilution of these highly alloyed fillers by carbon steel may have a number of effects:
- It will tend to reduce the weld metal strength to below that normally expected for the chosen consumable type.
- It will promote martensite formation in the stainless steel weld deposits, with associated increase of weld metal hardness and sensitivity to hydrogen cracking.
- For duplex and superduplex filler metals, it will tend to suppress ferrite formation.
Hence, dilution should be kept to a minimum. Hard-narrow martensite layers will tend to form at the interface to the carbon steel, regardless of filler choice and dilution level. If a nickel-based filler type is considered, it should be noted that niobium-containing consumables are not favoured for welding on duplex and superduplex steels due to the tendency for nitrogen diffusion out of the heat affected zone (HAZ) and into the weld metal and the associated formation of niobium nitrides in the weld metal. This can cause local reduction of mechanical and corrosion properties in the HAZ and in the weld metal adjacent to the fusion boundary.
It may also be noted that if bend tests are required for weld procedure qualification of dissimilar joints of this type, they should be performed with due consideration of the objective of such tests. Joints may fracture along the fusion line at fairly low bend angles, as a consequence of the sharp transition in material strength across the fusion boundary. The presence of a welding defect may not be required to initiate failure. Hence, such fractures should not be interpreted as bend test 'failures', as the test is designed to open and reveal any welding flaws, rather than to check the cross-weld mechanical properties. If these properties are of concern, cross-weld tensile testing should be performed. If fractures persist during bend testing in the absence of pre-existing defects, a suitably reduced bend angle or radius should be adopted.