Light barrier-based system to monitor welding fumes in powder bed fusion of metal with laser beam

Abstract Additive manufacturing processes, such as laser powder bed fusion of metals (PBF-LB/M), offer significant advantages to produce complex components. However, ensuring reproducible part quality remains challenging. Consequently, in-situ process monitoring is imperative to enhance process stability and minimise post-process quality assurance endeavours. The presence of welding fumes has been demonstrated to directly impact the quality of the resultant component, yet their highly dynamic behaviour complicates reliable detection. The present work proposes an in-situ welding fume monitoring approach based on a photodiode light-barrier system, in which fume particles intersect the optical path and induce measurable signal changes. The impact of typical process-related disturbances, including build-chamber illumination, laboratory lighting, powder recoating, laser exposure and process emissions, is systematically investigated under realistic PBF-LB/M conditions. The findings indicate that neither laboratory nor build-chamber lighting, nor laser activation in isolation, results in a quantifiable signal alteration. Conversely, the presence of powder turbulence during the process of recoating gives rise to a marked reduction in signal intensity, thereby facilitating unambiguous identification of this particular phase of the process. In the context of welding operations accompanied by fume generation, a discernible signal attenuation of A fume = − 13.47 is observed. Conversely, signals recorded with a laser power P L = 400 W and P L = 0 W exhibit a high degree of similarity in the absence of welding fumes. The findings demonstrate that the observed signal attenuation is directly linked to welding fume occurrence rather than to laser operation. The study establishes the fundamental feasibility and robustness of optical welding fume detection using a light-barrier concept and provides a quantitative basis for future sensitivity analyses and industrial implementation.