About this Abstract |
Meeting |
2022 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2022)
|
Symposium
|
2022 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2022)
|
Presentation Title |
Characterization of Defects within AM Fabricated Metal Components |
Author(s) |
Admas Kedebe, Richard Anarfi, Benjamin Kwapong, Kenneth Fletcher, Todd Sparks, Aaron Flood |
On-Site Speaker (Planned) |
Admas Kedebe |
Abstract Scope |
Characterization of defects within the volume of metal components produced via additive manufacturing (AM) processes is a critical component in both developing the process and qualifying it for end uses. X-ray tomography techniques offer a method for inspecting volumes of material rather than the 2d slices of conventional microscopy that may miss important data due to sampling frequency issues. For metal AM processes, both the spatial distribution and morphology of internal defects are meaningful. A common approach to AM volumetric data analysis is to use deep 2D convolutional neural networks (CNN). However, dealing with the individual slices independently in 2D CNNs discards the depth information which results in poor classification performance. In this paper, we propose to build a 3D CNN to predict defects in fabricated additive manufacturing metal components. Results from experiments with simulated models and from a DED process show that our model successfully classifies defects. |
Proceedings Inclusion? |
Definite: Post-meeting proceedings |