Analysis of material distribution in dissimilar friction stir welded joints of Al 1050 and copper

The objective of this work is to analyse the material distribution in dissimilar friction stir welded joints of Al 1050 and Copper. During dissimilar friction stir welding, mixing of materials leads to the emergence of interfaces between two dissimilar materials. These interfaces provide venues for the formation of intermetallic compounds, which significantly affect the joint properties. Therefore it is of interest to understand the material movement and track the interfaces which are generated during friction stir welding. A three-dimensional computational fluid dynamics model is developed for the dissimilar friction stir welding. The volume of fluid method is used to track the material movement during the process. Friction stir welds of Al 1050 and Copper are prepared, with different tool rotational speeds and tool offsets towards Al 1050 side. Optical microscopy of the weld crosssections is performed to observe the final status of material movement and material dispersion in the stir zone. Despite the absence of direct contact between copper and FS tool initially, copper is pulled into the stir region. Numerical results suggest that the temperature, stress and strain rates in the vicinity of FS tool are sufficient to pull copper into the stir zone. Relatively lower viscosities in shoulder driven region promote material mixing and uniform distribution of both materials. While pockets rich in copper or Al 1050 are observed in pin driven region. For the same strain rate and temperature, copper exhibits higher viscosity than Al 1050. This impedes the movement of copper and prevents it from moving all around the tool. This allows Al 1050 to displace copper at advancing side, which leads to shifting of initial copper-Al 1050 interface towards advancing side. Small offset towards Al 1050 and high tool rotation speed promotes pulling in of the copper and vice versa. This leads to larger interface shifts at smaller offsets and higher tool rotations speeds and vice versa. These trends are well captured by the numerical model.