A study of the structure of dissimilar submerged arc welds

Abstract

Butt welds have been produced using the submerged arc welding (SAW) process to join a carbon steel plate of normal shipbuilding grade to an austenitic stainless steel 316LN plate. Variables used in this work were the position of the consumable wire in the weld preparation and the ferrite number of the consumable wire. Abnormally high hardnesses were measured in some regions of the welds. These were related to the central position of the consumable wire in the weld preparation. Undesirably low ferrite numbers were related to the ferrite number of the consumable wire and also to the central position of the consumable wire in the weld preparation. The position of the consumable wire in the weld preparation controlled the relative dilution from the parent plates, and when the dilution from the 316LN steel was increased by off-setting the wire to that side of the weld preparation, the high hardness regions were no longer found. Similar optical microstructures were found to have significantly different hardnesses, which were related to the dilution effects, which were also related to the wire position within the weld preparation. Martensite was observed by transmission electron microscopy (TEM) in some of the regions of high hardness, and in other areas, the presence of extensive precipitation of M23C6 was seen. The presence of M23C6 was due to the effects of high heat input or reheating of areas by subsequent passes. The determination of the ferrite number in dissimilar welds was problematic. Fertiscope determinations appeared to be influenced by magnetic effects from adjacent areas of the carbon steel. Image analysis was found to be less reliable than the Feritscope for the determination of the ferrite numbers in dissimilar welds.