SUMMARY

Additive Manufacturing has been seeing widespread adoption in a variety of industries over the years. However, due to the inherent uncertainties with these manufacturing methods, process monitoring and qualification is critical. In situ process monitoring has been seeing significant interest over the years, and, especially in metal powder bed fusion process such as electron beam powder bed fusion, imaging is commonly used to detect defects. This work looks at comparing three basic segmentation methods (manual thresholding, statistical thresholding, and gradient) used for detecting porosity in in situ infrared imaging in electron beam powder bed fusion. Samples were manufactured with a variety of focus offsets to induce porosity, and the segmented infrared images were compared to ex situ X-ray computed tomography scans, which served as a ground-truth reference for objective evaluation. Through this analysis framework, influential parameters for each of the three methods were analyzed for effect on performance in properly detecting porosity, then the three methods were compared to each other.