|About this Abstract
||Materials Science & Technology 2019
||Additive Manufacturing Education
||Data Analytics for Metal Additive Manufacturing
||Amir Mostafaei, Nihal Sivakumar, David Crockett, Anthony D Rollett
|On-Site Speaker (Planned)
Producing defect-free parts with required microstructure and mechanical strength are necessary for the continued rapid growth of metal additive manufacturing (AM). The variation in feedstock powder, processing parameters during a build fabrication and post-processing can contribute to the reduction of defects. We are developing an image-oriented database to consolidate and analyze the data gathered from powder bed metal AM processes to identify the necessary parameters to produce defect-free parts. One issue was that so many different experiments had different parameters that consolidation into a single information file was difficult. The data gathering process had to be designed to include the parameters that were general to all experiments, and a filter search bar was introduced to allow users to search for specific key terms that return specific images. Ultimately, we aim to enable efficient machine learning on a wide range of materials, processes and imaging modalities.
||Definite: At-meeting proceedings