Iterative learning control for power profile shaping in selective laser melting

Selective laser melting (SLM) can be used to manufacture functional metal parts with complex geometries that cannot be produced by traditional manufacturing methods. However, SLM process control cannot yet guarantee the end part quality required for critical applications. The application of model-based control strategies to SLM is complicated by both the closed architecture of industrial SLM machines and the lack of suitable control-oriented process models. In this paper we (1) present an open-source SLM printer that allows implementation of the on-the-fly power adjustment and (2) use a data-driven method, iterative learning control (ILC) to learn the suitable laser power profile using the melt pool emission measurements from a coaxial camera. We demonstrate the effectiveness of the proposed ILC approach through experiments on the open-source SLM machine.