| Abstract Scope |
Multi-principal-component alloys (MPCAs) are an emerging class of engineering materials with good high-temperature phase stability and the potential to experience sluggish diffusion characteristics. These alloys have primarily been investigated as they pertain to structural or refractory applications, but the aforementioned properties render them an interesting material class for the exploration of filler materials for metallurgical joining challenges. Previous work has demonstrated the feasibility of an MPCA, with the composition Mn35Fe5Co20Ni20Cu20, to be successfully employed as a filler material for brazing Ni-base alloy 600.
To inform the observed elemental partitioning in brazed joints, the segregation behavior of this MPCA filler was investigated during and after non-equilibrium solidification. This investigation was performed via in-situ synchrotron x-ray diffraction (XRD) and postmortem characterizations on an autogenous laser weld. In the postmortem studies, Fe- and Co-rich dendrites were found to be present alongside Cu- and Mn-rich interdendritic material. Through analysis of the in-situ XRD patterns taken at high temporal resolution, four distinct stages in the solidification and cooling process were clearly identified. These are (1) dendrite solidification, (2) solidification of interdendritic material, (3) a period of solid-state interdiffusion between the two disparate compositions, and (4) marginal interdiffusion during final cooling. The diffraction patterns indicated the presence of only FCC crystal structures during all stages.
Temperature estimates made by monitoring the evolution of the diffraction peak indicated that Stage 1 was complete at approximately 1073°C, Stage 2 terminated at 943°C, and Stage 3 terminated at 520°C. A Scheil solidification simulation performed in ThermoCalc CALPHAD software indicated that the termination temperature for Stage 1 corresponded with a solid fraction of approximately 0.65. Using this metric as a boundary between dendrite and interdendritic solidification, the average compositions predicted by the Scheil model to solidify during Stage 1 and Stage 2 were determined. These compositions agreed qualitatively with the observed postmortem segregation. A hard-sphere atomic model was employed to estimate the FCC lattice parameter of each predicted composition. This approach gave a calculated difference in lattice parameter of 0.0279A between fully solidified dendrites and interdendritic material, agreeing with the XRD data taken at the end of Stage 2. A subsequent simulation performed using ThermoCalc’s DICTRA module predicted that solid-state diffusion would moderate this disparity in lattice parameter to 0.0249A by the termination of Stage 3, again agreeing with the in-situ XRD data at that time. Rietveld refinements were performed on the in-situ diffraction data from Stage 3, using imitation structures to accommodate the disordered MPCA crystal structure, and revealed the dendrite fraction evolution during cooling. A good agreement was achieved between the refinement results, ThermoCalc simulation, and postmortem metallurgical characterization. |