Towards a multi-sensor monitoring methodology for AM metallic processes

Additive Manufacturing (AM) is a promising manufacturing technology as compared to subtractive processes, in terms of cost and freedom of manufacturing. Among the AM techniques, Direct Energy Deposition (DED) processes are dedicated to functional metallic parts manufacturing. The energy input can be provided either by laser or electric arc, and having its deposited material as wire or powder form. DED processes incur drawbacks from lack of reproducibility and important production losses, mainly because they are operated in open-loop. Consequently, process monitoring is investigated to control the manufacturing state in real-time and ensure acceptable final parts. Presently, lots of papers have designed single closed-loop controls for DED processes, controlling either thermal, geometrical, or material delivery aspects. Multi-sensors monitoring strategies are also increasingly proposed, as controlling only one criterion has shown some limitations. Nevertheless, the developed multi-sensor strategies still focused on one type of phenomenonmainly geometryand have been implemented for only one DED process. This paper presents a new methodology of multi-sensor and multi-physics monitoring dedicated to at least two DED processes. The first investigations focus on a coupling between thermal and geometrical control loops, considering global part's temperature and layer height for thermal and geometrical aspects respectively. At the end of this paper, perspectives are given for closed-loop corrections according to the precited descriptors. These perspectives will be implemented in further works.