This is the authors' abstract. We don't add key points for this paper.
Additive manufacturing processes, such as laser beam melting, are now an integral part of industrial production. However, the development of the powder bed fusion-laser beam (PBF-LB) process is still ongoing in order to fully understand the complex interaction of the different parameters or to avoid process instabilities. As a result, process control and active parameter regulation are increasingly in the focus of current developments. A necessary prerequisite for achieving the goal of a controlled process is reliable defect detection as early as possible during a build job. In this paper a new approach is discussed: An equipment-independent image acquisition system combined with deep learning image processing. image of each layer is acquired after the build platform is recoated. The surface of the entire powder bed is evaluated by a CNN and the defect classification result is returned within seconds. The CNN is trained with a dataset of images from different real build jobs. Evaluating the entire powder bed provides the ability to detect and categorize defects with large extents. The perspective is a real-time process monitoring and control. Correction strategies are to be derived that interrupt the build process at the desired point and intervene in a targeted manner.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Kunkel et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: