Master of Science (MS), Wright State University, 2021, Computer Science

The additive manufacturing (AM) field is striving to identify anomalies in laser powder bed fusion (LPBF) using multi-sensor in-process monitoring paired with machine learning (ML). In-process monitoring can reveal the presence of anomalies but creating a ML classifier requires labeled data. The present work approaches this problem by printing hundreds of Inconel-718 coupons with different processing parameters to capture a wide range of process monitoring imagery with multiple sensor types. Afterwards, the process monitoring images are encoded into feature vectors and clustered to isolate groups in each sensor modality. Four texture representations were learned by training two convolutional neural network texture classifiers on two general texture datasets for clustering comparison. The results demonstrate unsupervised texture-driven clustering can isolate roughness categories and process anomalies in each sensor modality. These groups can be labeled by a field expert and potentially be used for defect characterization in process monitoring.

Committee: Tanvi Banerjee Ph.D. (Advisor); Thomas Wischgoll Ph.D. (Committee Member); John Middendorf Ph.D. (Committee Member) Subjects: Computer Science; Materials Science

Doctor of Philosophy in Materials Science and Engineering, Youngstown State University, 2019, Department of Mechanical, Industrial and Manufacturing Engineering

Metal laser powder bed fusion (LPBF) Additive Manufacturing (AM) affords new design freedoms for metallic structures with complex geometries. The aerospace industry has identified the inherent benefits of AM not just in terms of shape creation but also with regard to producing replacement parts for an aging fleet of aircraft. However, for these parts to be deployed, the quality must be well established given the lack of heritage for this manufacturing process. As additive manufacturing is executed layerwise, opportunities exist to non-destructively verify the fabrication in situ with a qualify-as-you-go methodology. A proposed solution is presented, which utilizes a pair of low cost, high speed cameras that are integrated and synchronized together to provide stereovision in order to identify the size, speed, direction and age of spatter ejected from the laser melt pool. The driving hypothesis of the effort is that the behavior of spatter can be reliably measured in order to determine the health of the laser process and ensure that spatter is not contaminating the build.

Committee: Brett Conner PhD (Advisor); Michael Crescimanno PhD (Committee Member); Eric Macdonald PhD (Committee Member); Virgil Solomon PhD (Committee Member); Jason Walker PhD (Committee Member) Subjects: Aerospace Materials; Materials Science