Monitoring ladle eye dynamics using multivariate statistical methods

Key conclusions include 1. Ladle eye size increases with gas flowrate and decreases with increasing upper phase thickness. 2. The eye size increases with increasing lower phase depth. 3. Multivariate Image Analysis (MIA), based on principal component analysis, was successfully applied to the cold model work of Krishnapisharody and Irons [5]. Predicted ladle eye size showed good agreement with experimental results for a variety of process conditions. 4. The MIA approach was able to provide rapid on-line estimates of the ladle eye area. 5. MIA was sensitive to lighting conditions. Image pre-processing was performed on images to improve model prediction results. 6. Preliminary work has proven MIA capable of successfully detecting ladle eyes within an industrial steelmaking ladle. 7. Work is ongoing to substantiate using MIA methods to monitor ladle eye dynamics under all industrial process conditions.