Electron Beam Melting (EBM) is a metal powder bed fusion (PBF) process in which the heat source is an electron beam. Differently from other metal PBF processes, today, EBM is used for mass production. As-built EBM parts are clearly recognisable by their surface roughness, which is, in some cases, one of the major limitations of the EBM process. The aim of this work is to investigate the effects of the orientation and the slope of the EBM surfaces on the surface roughness. Additionally, the machine repeatability is studied by measuring the roughness of surfaces built at different positions on the start plate. To these aims, a specific artefact was designed. Replicas of the artefact were produced using an Arcam A2X machine and Ti6Al4V powder. Descriptive and inferential statistical methods were applied to investigate whether the surface morphology was affected by process factors. The results show significant differences between the upward and downward surfaces. The upward surfaces appear less rough than the downward ones, for which a lower standard deviation was obtained in the results. The roughness of the upward surfaces is linearly influenced by the sloping angle, while the heat distribution on the cross-section was found to be a key factor in explaining the roughness of the downward surfaces.

Traditional sintering processes are carried out to achieve complete material densification. In an electron beam powder bed fusion (EB-PBF) process, the same sintering mechanisms occur but only with the aim to form small connections between the particles (necks). A proper neck formation is central for the EB-PBF process because, among other effects, ensures the thermal stability of the process and helps to avoid smoke phenomena. This work presents a numerical study of neck formation under the EB-PBF processing conditions. A new type of modelling is introduced for the temperature sintering load and included in a phase-field model, which simulates the neck growth during the EB-PBF process of Ti6Al4V powders. The model was validated with an ad-hoc experiment, which provided a deviation with respect to the estimated neck diameter of about 9%. The deviation was investigated by reasonably varying the processing conditions. The results showed that the thermal history, the process time scale (including also the cooling phase), and the geometrical characteristics of the particles significantly affected the sintering rate and neck radius.