Abstract
Wire and arc additive manufacture (WAAM) is a process in which an arc is used to melt and deposit beads of feedstock wire layer by layer onto a substrate and then subsequent layers to create near-net shape components. WAAM has the potential to reduce costs and material waste, as well as shorten lead-time, compared to conventional manufacturing methods. WAAM is not currently commercially viable, however, and one of the largest barriers is lack of reproducibility. This work investigates the influence of wire batch on reproducibility of the WAAM of aluminium. Particular attention is given to the external appearance and porosity of aluminium alloy 2319 (AA2319) components. Components were manufactured using 1.2 mm diameter AA2319 welding wire from different wire batches and manufacturers, and different process parameters. The surface deposit that resulted from the process on some components, which was a visible indicator of variability, was characterised using a qualitative coverage scale and five complementary techniques: scanning electron microscopy, X-ray photoelectron spectroscopy, X-ray diffraction, plasma focussed ion beam and Raman spectroscopy. Porosity of the components was also measured. Batch-to-batch variation of wire was analysed, in particular, the surface finish, diameter and composition. Surface deposit composition and coverage did not depend on wire batch. The surface deposit contained predominantly metal and metal oxide nanospheres, microspatter and carbon-rich microparticles. For AA2319 components, the metal was aluminium and for AA5183, the metals were magnesium and aluminium. There was a relationship between surface deposit coverage and wire feed speed (WFS): the higher the WFS, the greater percentage of surface deposit coverage. The results support the suggestion that surface deposit formation is caused by condensation of vapour and microspatter from feedstock wire. There was no statistically meaningful relationship between surface deposit coverage and porosity and surface deposit formation does not appear to influence commercial viability of WAAM. Wire batch was the dominant influence on porosity. Porosity ranged from 0.02 % to 6.0 % depending on wire batch used. There was no correlation between process parameters and porosity for the range of parameters used. There is evidence to suggest that internal voids and diameter of the wire affected porosity; wire with internal voids resulted in high porosity and wire with smaller diameter resulted in high porosity. These features can affect hydrogen content to the weld pool, arc stability, weld pool conduction and cooling rates, all of which influence porosity formation. The commercial viability of WAAM can be improved through control of feedstock material. This work showed that porosity could be reproducibly reduced using wire manufactured with a new manufacturing process that included diamond dies and a new cleaning process. Wire manufactured from the same ingot of material as other batches using the new process resulted in much lower porosity compared to the old process: 0.2 % compared to 1.2 %. Results suggest that changes to the wire manufacturing process may have produced wire with no internal voids and a larger diameter.