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 |
Multi-scale Shape Agnostic Flaw Detection and Monitoring in Laser Powder Bed Fusion using Heterogeneous In-process Sensor Data |
Author(s) |
Benjamin D. Bevans, Chris Barrett, Tom Spears, Aniruddha Gaikwad, Alex Riensche, Ziyad Smoqi, Scott Holliday, Prahalada Rao |
On-Site Speaker (Planned) |
Benjamin D. Bevans |
Abstract Scope |
The goal of this work is the flaw free production of metal additive manufactured (AM) parts made using the laser powder bed fusion process (LPBF). As a step towards this goal, we develop and apply machine learning approaches to detect multi-level flaws ranging from the micro-scale porosity, meso-scale layer deformation, and macro-scale flaws, using data acquired from multiple sensors targeted at each level. Data from a near infrared meltpool monitoring- , high speed- , and optical powder bed imaging-cameras are analyzed using graph theory to extract monitoring features. Then machine learning models are trained to detect the onset of multi-scale flaw formations and are tested across different part geometries and build plates. The approach detects porosity, layer-level distortion (warpage), and geometry (scan path) related flaws with statistical fidelity exceeding 90% (F1-score). Thus this work takes the first step towards shape agnostic, multi-sensor, detection of multi-scale flaws in LPBF. |
Proceedings Inclusion? |
Definite: Post-meeting proceedings |