L-PBF High-Throughput Data Pipeline Approach for Multi-modal Integration

• DOI: 10.1007/s40192-024-00368-0
• Publication Date: 2024-07-19T16:01:43Z
• Authors: Kristen J. Hernandez; Thomas G. Ciardi; Rachel Yamamoto; Mingjian Lu; Arafath Nihar; Jayvic Cristian J...; Pawan K. Tripathi; Brian Giera; Jean-Baptiste Forien; John J. Lewandowski; Roger H. French; Laura S. Bruckman
• Keywords: Pyrometer; Feature (linguistics)

Abstract Metal-based additive manufacturing requires active monitoring solutions for assessing part quality. Multiple sensors and data streams, however, generate large heterogeneous sets that are impractical manual assessment characterization. In this work, an automated pipeline is developed enables feature extraction from high-speed camera video multi-modal analysis. The framework removes the need through utilization of deep learning techniques training models in a weakly supervised paradigm. We demonstrate pipeline’s capability over 700,000 frames. successfully extracts melt pool spatter geometries links them to corresponding pyrometry, radiography, processparameter information. 715 individual prints examined reveal areas exceeds 0.07 mm 2 pyrometry signal threshold (375 units) were more likely have defects. These processes enable massive throughput characterization techniques.