|About this Abstract
||2023 TMS Annual Meeting & Exhibition
||Advanced Real Time Imaging
||Prediction of Laser Absorptivity from Synchrotron X-ray Images Using Deep Convolutional Neural Networks
||Runbo Jiang, Joseph W. Aroh, Brian Simonds, Tao Sun, Anthony D. Rollett
|On-Site Speaker (Planned)
The quantification of the amount of absorbed light is essential for understanding laser-material interactions and melt pool dynamics in order to minimize defects in additively manufactured metal components. The geometry of a vapor depression, also known as a keyhole, in melt pools formed during laser melting is closely related to laser absorptivity. This relationship was observed by the state-of-the-art in situ high speed synchrotron x-ray visualization and integrating sphere radiometry. These two techniques create a temporally resolved dataset consisting of keyhole images and the corresponding laser absorptivity. In this work, deep convolutional neural networks including ResNet-50 and ConvNeXt-t are adapted to interpret an unprocessed x-ray image of a keyhole and predict the amount of light absorbed. Class activation map is used for visualizing where deep learning networks pay attention to make predictions. The high-dimensional features extracted is visualized using principal component analysis to identity the relationship between the keyhole geometry and laser absorptivity.