Using suitable scanning strategies, even single crystals can emerge from powder during additive manufacturing. In this paper, a full microstructure map for additive manufacturing of technical single crystals is presented using the conventional single crystal Ni-based superalloy CMSX-4. The correlation between process parameters, melt pool size and shape, as well as single crystal fraction, is investigated through a high number of experiments supported by numerical simulations. Based on these results, a strategy for the fabrication of high fraction single crystals in powder bed fusion additive manufacturing is deduced.

The current study evaluates the capabilities of electron-optical (ELO) in situ imaging with respect to monitoring and prediction of manufacturing precision in electron beam powder bed fusion. Post-process X-ray computed tomography of two different as-built parts is used to quantitatively evaluate the accuracy and limitations of ELO imaging. Additionally, a thermodynamic simulation is performed to improve the understanding of ELO data and to assess the feasibility of predicting dimensional accuracy numerically. It is demonstrated that ELO imaging captures the molten layers accurately (deviations <100 μm) and indicates the creation of surface roughness. However, some geometrical features of the as-built parts exhibit local inaccuracies associated with thermal stress-induced deformation (deviations up to 500 μm) which cannot be captured by ELO imaging. It is shown that the comparison between in situ and post-process data enables a quantification of these effects which might provide the possibility for developing effective countermeasures in the future.

The development of process parameters for electron beam powder bed fusion (PBF-EB) is usually made with simple geometries and uniform scan lengths. The transfer to complex parts with various scan lengths can be achieved by adapting beam parameters such as beam power and scan speed. Under ideal conditions, this adaption results in a constant energy input into the powder bed despite of the local scan length. However, numerous PBF-EB machines show deviations from the ideal situation because the beam diameter is subject to significant changes if the beam power is changed. This study aims to demonstrate typical scaling issues when applying process parameters to scan lengths up to 45 mm using a fourth generation γ-TiAl alloy. Line energy, area energy, return time, and lateral velocity are kept constant during the additive manufacturing process by adjusting beam power and beam velocity to various scan lengths. Samples produced in this way are examined by light microscopy regarding lateral melt pool extension, melt pool depth, porosity, and microstructure. The process-induced aluminum evaporation is measured by electron probe microanalysis. The experiments reveal undesired changes in melt pool geometry, gas porosity, and aluminum evaporation by increasing the beam power. In detail, beam widening is identified as the reason for the change in melt pool dimensions and microstructure. This finding is supported by numerical calculations from a semi-analytic heat conduction model. This study demonstrates that in-depth knowledge of the electron beam diameter is required to thoroughly control the PBF-EB process, especially when scaling process parameters from simply shaped geometries to complex parts with various scan lengths.