Layering defects detection in laser powder bed fusion using embedded vision system

Additive Manufacturing (AM) is nowadays emerging as a new way to manufacture complex metallic parts which would be difficult or impossible to obtain using conventional manufacturing processes. However, as with any manufacturing process, monitoring and mastering the process is essential to satisfy the requirements. In AM, most of the defects considered are internal (mainly porosity or inclusions) and due to process capabilities, it is mainly a way to obtain a raw part. This paper focuses on evaluating internal defects in Inconel 718 parts obtained by a laser-based AM technology, i.e. Laser Powder Bed Fusion (LPBF). Some defects can be explained by the interaction between the layering system and the powder bed in the layering stage (keyhole induced by a lack of powder), by non-well melted areas during the melting stage (porosity or keyhole), or by part distortions in the solidification and cooling stages (cracking, warpage, base plate separation,…). In this study, an algorithm is developed to build a 3D Voxel-based model of manufactured parts, based on defects detected on post-layering and post-melting pictures using an embedded visible camera. This model is also compared to a corresponding 3D model of the build that was achieved by X-ray Computed Tomography (CT) scan data, to check its relevance. © 2021 CAD Solutions, LLC,.