Low-Cost Image Collection and Compilation for Enhanced In-Situ Monitoring in Laser-Powder Bed Fusion

Laser-Powder Bed Fusion (L-PBF) additive manufacturing relies on precise control for quality assurance and reproducibility, necessitating advanced in-situ monitoring techniques. Presented is a novel low-cost approach to image collection and compilation to enable in-situ monitoring in L-PBF in ways not performed to date, i.e., at a lower cost. The research incorporates optical, near-infrared (NIR), and infrared (IR) imaging modalities to record and, in the future, analyze and detect defects during the L-PBF process. Off-axis cameras with different spectral sensitivities complicates the comparison of images due to differing image planes. A low-cost, calibration plate is used to create common image planes through perspective transforms that allows the overlay of images from differing cameras and camera orientations, a prerequisite to achieving defect detection. This calibration plate uses two materials with differing emissivity to allow calibration across differing modalities. This approach results in a maximum image error of only 1 mm. Data is collected using a custom, Python-based recording script, optimized for real-time monitoring and storage using the HDF5 file format. The findings suggest that this low-cost solution can be effectively implemented, providing the base work for future work in defect detection.