In-situ detection of powder bed defects in laser powder bed fusion using 3D surface normals and depth mapping

Abstract

Laser Powder Bed Fusion (LPBF) is one of the most extensively studied metal additive manufacturing processes. Owing to its capabilities in achieving high manufacturing precision, excellent surface quality, and producing complex geometries, it has gained widespread adoption in industries including shipbuilding, automotive, and aerospace. During the LPBF process, powder bed defects are among the most common types of manufacturing defects. However, researchers face major challenges due to the scarcity of in-situ monitoring methods and limited diversity in monitoring data. Optical measurement techniques offer high precision, efficiency, and non-contact operation for in-situ LPBF monitoring. This study proposes an in-situ powder bed monitoring method for LPBF based on a dual-sensor fusion of photometric stereo and structured light measurement. From the raw images captured by sensors, normal maps and depth difference maps are computed, and an enhanced image is synthesized. Multiple quantitative metrics are used to evaluate the effectiveness of the synthesized images in visualizing powder bed defects. The results show that, compared to grayscale images, the synthesized images exhibit significant enhancements of over 65.5 %, 39.8 %, and 147.0 % in entropy, average gradient, and variance, respectively, demonstrating the effectiveness of the proposed fusion strategy in enhancing defect visualization. This achievement provides researchers in the LPBF field with richer in-situ monitoring data and contributes to further defect reduction and improvement in manufacturing precision.