Macrosegregation in the Weld Pool in Metal Inert-Gas Welding of Aluminium

Abstract

• Arc welding of Al alloys is modelled in lap and bead-on-plate geometries • Detailed thermophysical properties are given for aluminium–silicon–magnesium alloys • The filler metal is concentrated near the bottom of the fusion zone • A more vertical electrode orientation reduces macrosegregation in lap welds • The model is validated against measurements of weld cross-section composition A computational model of metal–inert-gas arc welding, in which the arc plasma, the wire electrode and the workpiece are included in the computational domain, is used to investigate the macrosegregation of the filler metal (the wire electrode alloy) and the base metal (the workpiece alloy) in the weld pool. The aluminium–silicon alloy AA4043 is used for the wire electrode, and the aluminium–magnesium alloy AA5754 for the workpiece. Detailed thermophysical property data for aluminium–silicon–magnesium alloys are given. Modelling results are presented for lap geometry, including different electrode angles, and for bead-on-plate geometry. It is found that the filler metal is concentrated in the region near the boundary of the fusion zone and the base metal. This concentration occurs because of the convective flow of the filler metal towards this boundary, which is in turn driven by the momentum transferred to the weld pool by the droplets. This mechanism is consistent with one of several macrosegregation mechanisms proposed in the literature based on experimental observations. The macrosegregation in lap welds is predicted to be reduced when the electrode orientation is closer to vertical. Measurements of the concentration of silicon and magnesium in weld cross-sections for bead-on-plate and lap geometry were used to confirm the distribution of filler metal predicted by the model are consistent with experiment.