| About this Abstract |
| Meeting |
2024 TMS Annual Meeting & Exhibition
|
| Symposium
|
Additive Manufacturing Fatigue and Fracture: Towards Rapid Qualification
|
| Presentation Title |
An Investigation into Non-destructive Material and Part Qualification for Fatigue Critical Applications |
| Author(s) |
Alireza Jam, Shahryar Baig, Jia Liu, Shuai Shao, Nima Shamsaei |
| On-Site Speaker (Planned) |
Alireza Jam |
| Abstract Scope |
Additive Manufacturing typically induces varying degrees of defects/anomalies in the parts. The variation in defect morphologies and location can result in inconsistencies in structural integrity, especially under cyclic loading, which ultimately challenges the qualification of AM materials/parts in fatigue critical applications. This study attempts to shed light on the correlation between the geometrical features as well as locations of critical defects and fatigue behavior of laser beam powder bed fused parts by means of non-destructive techniques. A data-driven framework with computer vision and machine learning is used for the non-destructive qualification of AM materials/parts based on the defect criticality. Utilizing the 2D image data from fractography, as well as the 3D defect data from X-ray CT scans, the machine learning framework is trained to identify critical defect features by computer vision and model their relationships to fatigue performance. The prediction results are compared with experimentally obtained fatigue data. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Additive Manufacturing, Mechanical Properties, Characterization |