Real-time corner height estimation for multi-layer directed energy deposition using laser line scanner, vision camera, and artificial neural network

Abstract: Metal additive manufacturing (AM) technologies, such as laser direct energy deposition (DED), have been widely used in recent years because of their powerful capability for manufacturing near-net-shaped complex components for industrial applications in various fields. However, the current track geometry inaccuracy in the DED process, especially at corners with sharp turns, is a key barrier to the adoption of this advanced technology. In this study, a real-time corner height estimation technique is proposed for multi-layer track-with-corner deposition using an artificial neural network (ANN) in which measurements are obtained with a vision camera and a laser line scanner.