To meet the need for multifunctional industrial application, such as mechanical properties, corrosion and wear resistances, and electrical and thermal conductivities, functionally graded materials (FGMs) were developed. FGMs are characterized by a gradual change in composition or structure or both, and thus in properties, over the part in a certain direction.

    Thanks to its layer-by-layer manufacturing strategy allowing a high control over spatial resolution, additive manufacturing (AM) turns out to be well suited for the development of FGMs. However, in FGMs the properties are mainly determined by the graded interface due to poor bounding and mixing between the two materials/structures leading to strong macrosegregation, pore and crack formation, as well as the precipitation of brittle phases which are detrimental to the properties.

    Although of key importance to understand the formation of the aforementioned undesirable structures, no real time experimental investigations were performed to study how the two elements mix at the interface during AM. Although, it can be mentioned that numerical simulations were developed to tackle this problem, they only reliable on ex situ analyses due to the lack of real-time experimental results.

    The present work tries to asses both experimental and numerical parts by focusing on the solute redistribution evolution during the printing of a Cu alloy powder onto a thin Al base plate during the laser powder bed fusion (LPBF) AM process. The experiments were performed on the ID19 beamline at the European Synchrotron Radiation Facility using a LPBF machine. Imaging experiments were carried out with a polychromatic beam having a peak energy at 19 keV, a frame rate of 80 kHz, and a pixel size resolution of 4.1 µm.

    Thanks to the X-ray absorption contrast between Cu and Al, the mixing between both elements was observed during AM showing complex fluid mechanisms within the melt pool such as Marangoni flow and vortexes. In addition, based on the X-ray absorption contrast, quantitative projected compositions were extracted allowing to study its evolution. It was observed that locally the average Cu composition increases and reaches a plateau while its distribution passes by a maximum before decreasing due to the mixing occurring in the melt pool. In addition, to better understand the mechanisms occurring during the mixing, these operando results are supported by tomography experiments and computer simulation.