About this Abstract |
| Meeting |
MS&T25: Materials Science & Technology
|
| Symposium
|
Advancement of Measurement Technologies for Harsh Environments
|
| Presentation Title |
Infrared Video Imaging for In-Line Evaluation of Oxide Descaler Performance on Steel Strips |
| Author(s) |
Ry Karl , Jarrod Angove, Jonas Valloton, J. Barry Wiskel, Chad Cathcart, Tihe Zhou, Christopher Martin-Root, Hani Henein |
| On-Site Speaker (Planned) |
Jarrod Angove |
| Abstract Scope |
Surface oxides can form at various stages during thermomechanical control processing (TMCP). These oxides have been shown to shift the boiling curve on the surface of the strip, causing localized changes in cooling rates. As oxide growth and removal can be inconsistent, this may result in small regions with unexpected phase compositions. To address this, this work proposes a novel method for the visualization of surface oxides via infrared (IR) thermal imaging. This is done by applying a combination of two machine learning algorithms – Gaussian process regression and Gaussian mixture models – to identify oxides in IR videos. The proposed methodology is applied to a collection of IR videos taken during the production of multiple X70 steel strips. The resulting oxide distributions are used to identify possible causes of increased oxide coverage. |