Multi-data-driven model-based control to improve the accuracy of overhang structures in laser powder bed fusion

The additive manufacturing process of laser powder bed fusion (LPBF) has gained significant attention due to its potential to fabricate complex structures. In this paper, we propose an open-loop multi-data-driven model-based control (MMC) architecture to improve the accuracy and quality of overhang structures fabricated through LPBF. Overhang structures are particularly challenging due to thermal and structural complexities. Traditional methods often require support structures, which can increase costs and reduce efficiency. Our MMC architecture integrates data-driven models with real-time monitoring using photodiode measurements. The proposed MMC architecture consists of energy density modeling, infrared correction, and reference voltage modeling. The energy density model is established for both solid substrates and overhang structures to optimize laser power, while the infrared correction model compensates for measurement errors caused by melt pool position variations. The reference voltage model enables real-time process monitoring. Experimental results demonstrate the effectiveness of the proposed MMC architecture in producing high-quality overhang structures without the need for support materials. The architecture offers a significant advancement in LPBF process control by reducing 49.9% of residual material of overhang structures, enabling the fabrication of intricate geometries with improved accuracy and efficiency.