Toward real-time chemical mapping during laser powder bed fusion: Robust in-situ spectroscopy and 3D reconstruction

Laser powder bed fusion of metals (PBF-LB/M) is an established additive manufacturing (AM) technique enabling material-efficient production of complex metal components. However, current PBF-LB/M systems cannot monitor or control local chemical composition during fabrication, relying instead on ex-situ analyses after printing to assess material quality, which is incompatible with real-time quality assurance in industrial production. This study applies optical emission spectroscopy (OES) to monitor the chemical composition of parts during fabrication, with a focus on identifying and quantifying key influence factors such as laser energy input, alloy composition, integration time, and analysis position exemplified for magnet (Nd-Fe-B) and Scalmalloy (Al-Mg-Sc) printing. By comparing spectra from powder and bulk samples and evaluating characteristic peak intensity ratios, global spectral effects are effectively minimized. The machine-integrated, optically off-axis OES enabled reproducible measurements with composition sensitivity in the sub-at% range. Thereby, a novel concept for the 2D/3D reconstruction of chemical composition during PBF-LB/M is demonstrated. This approach aligns spatial and temporal OES data with the laser scan path through synchronizing the spectrometer acquisition interval with the machine clock, generating spatially resolved composition maps. Proof of concept is presented through layer-wise 2D mapping, with OES-derived data validated by ex-situ EDX analysis, with layers digitally stacked into a 3D reconstruction. This method lays the groundwork for robust in-situ chemical mapping during PBF-LB/M and supports the advancement of digital twins by a material vector. By enabling real-time monitoring and process control, it will open new pathways for quality assurance and adaptive manufacturing strategies in industrial settings.