| Abstract | High-integrity die castings are often welded into complex structures. Compared to conventional die castings, high-vacuum can significantly reduce weld porosity after gas-metal arc or laser welding. However, achieving robust welding remains an industrial challenge due to the large number of variables involved. Surface characteristics, in addition to core material, are known to play a significant role in weldability. This paper investigates the impacts of die lubricant type, location on the part, cleaning method, and welding process. Critical factors contributing to robust joining capability were identified using a combination of advanced characterization and machine learning models. A 3-mm thick flat die insert with edge features was designed to generate detrimental filling patterns within the available envelope. Specimens were cast with the Aural™-2 alloy and welded in the F temper. Before welding, specimens underwent non-destructive testing by immersed ultrasound, with millimeter-scale traceability. Surface characteristics were also investigated after surface treatment using infrared spectroscopy to quantify oxides and remaining traces of die lubricants. Specimens were then welded in a bead-on-plate configuration with GMAW or autogenous laser processes. Given the potential shot-to-shot variations, five specimens were tested for each case. Post-weld quality was evaluated with 2D X-ray inspection and segmented along each weld. Finally, since inputs and outputs were traceable along the specimens and weld lines, quality was correlated with casting and welding parameters using visual interpretation methods and statistical analyses. This work highlights the impact of surface characteristics on post-weld quality to identify favourable process windows for robust joining procedures. |
|---|
