About this Abstract |
Meeting |
2021 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2021)
|
Symposium
|
Special Session
|
Presentation Title |
Empirical and simulated x-ray computed tomography probability of detection analysis for additive manufacturing and calibrated defect artifact development |
Author(s) |
Felix H. Kim, Adam L Pintar, John Henry J Scott, Nikolai Klimov, Sarah Robinson |
On-Site Speaker (Planned) |
Felix H. Kim |
Abstract Scope |
X-ray computed tomography (XCT) is an emerging industrial non-destructive testing (NDT) technique for additive manufacturing (AM). It shows great promise, in part because it is less susceptible to the complex part geometry and rough surfaces of AM parts that are problematic for other NDT techniques. The reliability of an NDT inspection process can be assessed by a probability of detection (POD) study, and the practice is expected to enable wider adoption of industrial XCT inspection. An empirical POD curve was estimated using calibrated artifacts and a statistical model incorporating reference measurement uncertainty. Empirical determination of a POD curve, however, can be costly in both time and money, which limit the number of practically-viable XCT measurements. For instance, it is not practical to study the effect of defect locations for wide ranges of defect sizes empirically with limited artifacts and measurements. In this presentation, the complete XCT simulation process will be discussed, and the empirical and simulated data will be compared. We explore the possibilities of incorporating simulated defect conditions that were not evaluated in the experiments to estimate a more comprehensive POD curve and using the simulation results to improve defect artifact designs. Finally, based on this study, we are developing improved calibrated defect artifacts using deterministic nanofabrication approaches, and the progress will be discussed. |
Proceedings Inclusion? |
Definite: Post-meeting proceedings |