Rapid surface defects detection in wire and arc additive manufacturing based on laser profilometer

•A rapid in-situ surface defects detection system via laser triangulation.•Converting 3D surface profiles to 2D topography image by non-linear normalization.•Classifing the pixels into different categories with supervised learning.•The system was capable of locating and identifying surface defects quantitatively. Wire arc additive manufacturing technology (WAAM) has become a very promising alternative for large-scale metal components manufacturing. Due to the unsatisfactory performance in solidification, surface quality monitoring has been a critical issue for WAAM. In this study, we set up a non-contact in-situ 3D laser profilometer inspection (3D-LPI) system to automatically monitor the visual surface defects. The 3D surface point cloud was converted to a 2D topography image firstly. Then the surface defects were identified after the classification of pixels using a support vector machine (SVM) model. The availability of the system was validated in the building process of different aluminum components. The results illustrated that the proposed novel methods can detect not only widespread bulge and collapsing defects but also small pore defects with pixel-level accuracy, which has great significance for the automatic quality evaluation and process control in WAAM.