INTELLIGENT TECHNOLOGIES FOR DETERMINING QUALITY CHARACTERISTICS OF METAL PRODUCTS IN LAYER-BY-LAYER FORMATION PROCESSES

Abstract

The study examines intelligent technologies for assessing the quality of metal parts produced by layer-wise additive manufacturing. Based on an analysis of recent publications, the main defect types in laser powder bed fusion processes are summarized together with their links to powder morphology, melting regimes and microstructural evolution. It is shown that the combination of computer-vision and machine-learning methods enables automated powder characterization, in-situ melt-pool monitoring, detection of anomalous regimes and layer-wise porosity prediction. Particular attention is given to digital twins of additive processes, which reproduce thermal, mechanical and microstructural fields and support multi-scenario optimization of printing parameters before physical experiments are carried out. The feasibility of integrating physics-based simulation data, process-monitoring signals and microstructural characterization results into a unified intelligent quality-control system focused on safety-critical machine-building components is substantiated. A structural scheme of such a system is proposed, comprising modules for raw-material analysis, process planning, in-situ monitoring, defect prediction, adaptive control and post-process validation. It is demonstrated that the implementation of such systems provides a basis for self-learning additive manufacturing lines, reduces porosity levels and improves the repeatability of mechanical properties of parts. The scientific contribution of the work lies in consolidating current approaches to coupling digital twins with deep-learning models and in formulating a generalized architecture of an intelligent quality-assessment framework for metal additive manufacturing. The practical relevance is associated with the potential to reduce the scope of destructive testing, shorten the time-to-market for new products and increase the reliability of critical structural elements operating under demanding service conditions.