In situ and operando Monitoring of Laser Additive Manufacturing

Metal laser additive manufacturing (LAM) is increasingly being used for manufacture of high-value, safety critical components. However, it still suffers from a lack of in situ quality assurance and control. In this thesis, an interferometric monitoring technique called inline coherent imaging (ICI) is applied to LAM for high-speed (200 kHz) in situ morphology measurements and defect detection. The Fraser group powder bed fusion (PBF) machine was used to demonstrate manual layer-wise control of a 3D build of 316L stainless steel with a seeded unsupported region. Surface roughness was corrected by ablating raised areas and infilling depressed areas of each layer based on post-processing ICI measurements (54% decrease in max. S_a). ICI was then combined with high-speed synchrotron X-ray imaging for correlative monitoring of LAM. ICI was integrated into a directed energy deposition (DED) machine with a novel off-axis imaging beam delivery and micro-electromechanical systems (MEMS) scanner for alignment and positioning. ICI was used to monitor DED track morphology during thin-wall builds of nickel super-alloy CM247LC, an alloy used for turbine components susceptible to cracking under LAM conditions. ICI provided immediate diagnostics on the DED process, including material deposition rate, track width, and surface roughness and waviness. Humping (large surface waviness) in thin-walls reduces geometric accuracy and leads to increased residual stress in valleys, which can induce cracking. ICI was used to in situ detect humping and crack openings, simultaneous with X-ray observation of sub-surface crack growth at the Diamond Light Source (DLS). ICI was then integrated into the PBF test rig at the Advanced Photon Source (APS) for simultaneous ICI and X-ray imaging of PBF. Simultaneous radiography was used to validate ICI depth measurements from the keyhole during laser welding and PBF of aluminum alloy 6061. Even in a turbulent pore-generation mode, ICI depth measurements from the keyhole corresponded closely with the keyhole depth extracted from radiography (>80% within ±15 µm). A ray tracing simulation confirmed that outliers in ICI data are often the result of multiple reflections within the keyhole (57%). ICI depth signatures for bubble and pore formation were also identified, capturing keyhole pinching, “pull-up,” and spiking phenomena.
External DOI