Effect of Submerged Arc Welding Parameters on the Microstructure of SA516 and A709 Steel Welds

Abstract

The effects of submerged arc welding (SAW) current and speed on the microstructures of SA516 grade 70 and A709 grade 50 steel welds were studied in this research. Steel plates 17 mm-thick were submerged arc welded using different welding currents (from 700 to 850 A) and welding speeds (from 5.3 to 15.3 mm/s). The effect of heat input on the weld metal chemistry, morphologies and chemistry of inclusions and nucleation of acicular ferrite (AF), grain boundary ferrite (GBF) and Widmanstatten ferrite (WF) were evaluated. Optical microscopy (OM), scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDS) microanalysis and transmission electron microscopy (TEM) were used to examine the microstructures of the developed weld joints. PAX-it image analysis software program was utilized for quantitative analysis of the microstructures.

The results showed that it is difficult to ascribe changes in the microstructure that occurred in the heat affected zone (HAZ) and the weld metal regions to a single welding process parameter. Inclusion analysis revealed two types of inclusions formed in the weld metals for both steels. They are spherical and faceted inclusions. It was also observed that acicular ferrite nucleated only on the spherical inclusions. EDS analysis showed that the two inclusions have different chemical compositions.

The results further showed that the total oxygen content of the weld metals of both steels generally increased with welding current, but decreased with increasing welding speed. The prior austenite grain width decreased with increasing welding speed, but increased with increasing welding current (increased heat input). For both SA516 and A709 steel welds, the proportion of acicular ferrite (AF) in the weld metals increased initially, while those of grain boundary ferrite (GBF) and Widmanstatten ferrite (WF) decreased with increasing welding current when welding current was increased from 700 A to 800 A. With further increase in the welding current above 800 A, less acicular ferrite was produced as both GBF and WF proportions increased. However, welding speed did not affect appreciably the amounts of ferrite products in the weld metals.

Non-linear regression models were developed using welding current and welding speed to predict the ferrites (AF, GBF, WF) that formed in the weld metals of the two steels. The adequacy of the models was checked by using the F-statistics.