Design driven defects in laser powder bed fusion: correlation between in-situmonitoring data and ex-situ measurement
Examensarbete för masterexamen
Materials engineering (MPAEM), MSc
Belsure, Nikhil Gautam
Industrial adoption of Laser Powder Bed Fusion (LPBF) is hindered by lack of process reliability and stability and the resulting build part quality. Manufacturing defect-free parts is crucial for quality-critical industries such as aerospace. Prior research has proposed in-situ monitoring of LPBF as a novel tool for early defect detection and improving process reliability. Numerous studies investigate causal correlation between anomalies in in-situ monitoring and defects and establishing ground-truth for the anomalies observed in in-situ monitoring data. However, the detection and characterization of defects generated by challenging geometries is relatively underexplored. This study aims to correlate anomalies observed in LPBF in-situ monitoring systems to design-driven defects arising from challenging geometries. Further, these defects are characterized through ex-situ measurements using non-destructive evaluation (NDE), namely X-ray computed tomography (XCT), 3D scanning, and optical surface roughness measurement. Geometries with challenging design features such as thin structures and unsupported overhangs, were designed and printed to provoke defects such as geometrical deviations, surface roughness, and internal flaws. Monitoring data obtained from the LPBF platforms, including optical tomography (OT) images, spectral data, and powder bed images, were analyzed to identify indications of these defects. Distortions in thin geometries exhibited strong correlations to protrusion of the part through the powder bed after powder dosing. Hotspots that corresponded with short hatch vectors and downskin surfaces of overhangs lesser than 45° indicated surface defects. It was observed that hotspots did not always indicate the presence of defects. Sometimes hotspots appeared in areas or parts that were exposed in the prior powder bed layer. Correlating the shifting of thin geometries due to recoater influences with the monitoring data was possible with long exposure OT. The findings highlight the importance of considering geometry-induced defects in LPBF processes and provided a validation for the anomalies observed in monitoring data.
additive manufacturing , laser powder bed fusion , in-situ monitoring , Non-Destructive Evaluation , X-ray CT , 3D scanning , surface roughness , geometrical deviation