Real-Time Non-invasive Velocity Field Measurement of Molten Iron Jet Discharged from Blast Furnace

Monitoring the internal state of a blast furnace during iron-producing operation is critical to maintaining a steady and efficient smelting process. Directly measuring the instantaneous casting rate at the blast furnace tapholes can provide process insights that would otherwise be unattainable. This research utilizes a machine vision camera with the image cross-correlation method to estimate the surface velocity field of the molten iron jet exiting from a real blast furnace during the casting process. An optimal method, the streamlineguided digital image correlation, is proposed specifically for the experiment target, the molten iron jet. The proposed method first estimates the streamline of the jet based on its edges and only applies the digital image correlation along the streamline to reduce unnecessary calculations. The proposed method is successfully integrated into a designed monitoring system for a blast furnace at United States Steel Corporation Gary Works with a refresh rate of 2.9/14.8 s with/without being boosted by graphic card. The measurement result successfully estimates the velocity field of the jet exiting the furnace and shows the proposed method could run 27 times faster than the conventional method and keep an absolute difference under +/- 1 pixel for both horizontal and vertical directions.