Modeling of Fluid Flow, Mixing and Inclusion Motion in Bottom Gas- Stirred Molten Steel Vessel Development of a Process to Continuously Melt, Refine, and Cast High-Quality Steel Lifeng Zhang, Brian G. Thomas University of Illinois May 10, 2004 Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 1
Acknowledgements - Continuous Casting Consortium at UIUC - UMR (Prof. Kent Peaslee, Prof. David Robertson, Mr. Jorg Peter) - Department of Energy (DOE) - Fluent Inc. - National Center for Supercomputing Applications at UIUC Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 2
11 10 9 8 height or width (meters) 7 6 5 4 3 2 30 min 50 t 0.15%C melt Conceptual layout for 110 tph process 2890ºF measure + 12 min + 12 min + 21 min = 75 min 20 t CO, Ar CaO, MgO measure Deoxidizer, CaO, MgO? heated holding ladle (for EAF or tundish), or pig the steel O 2 + Ar 1 oxidize Ar Ar Ar Ar Ca wire 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 mold Univ. of Illinois at Urbana-Champaign Metals length Processing (meters) Simulation Lab L. Zhang, 2004, 3 Figure from Jorg Peter, UMR 20 t 2910ºF 2880ºF Vessel II? reduce Ar alloys 2870ºF measure heating rods 27 t wire feeding adjustment Top view looking at steel surface Side view homogenize & float Ar 7 t 23 t 2830ºF measure
Geometry of Vessel II Inlet Outlet Launder Inlet Launder Vessel II Outlet SEN Gas Injection inlets: Position: 1/4 and 3/4 diameter of the bottom Din=0.2m Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 4
Geometry of Vessel II Inlet (Din=0.14m, Dout=0.2m, Submergence depth:0.15m Simulated inlet SEN length: 0.5m Distance of inlet center line to left wall: 0.2m Vin=0.23m/s ε Kin=0.01m2/s2 in=0.06m2/s3) Outlet Launder: Thickness:0.3m 1.1m Inlet Launder: Left depth: 0.3m Right depth:0.6m Length: 1m Thickness:0.3m 0.4m 0.3m 0.9m 0.7m 0.3m Vessel II (Din=1.4m, Depth: 2.0m) 0.4m Outlet SEN (Din=0.14m, Dout=0.2m, Length: 1.1m, no slide gate Distance of outlet center line to right wall: 0.2m) Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 5
Geometry of Vessel II Inner diameter: D=1.4m Steel bath height: H steel = 2.0m Slag depth: H slag =0.2m Diameter of gas porous plug: 0.2m (two plugs) Inlet nozzle inner diameter: D steel, in =0.14m Inlet nozzle outer diameter: 0.2m Inlet submergence depth: 0.15m Inlet SEN length: 1.1m Modeled inlet length: 0.5m Distance from inlet center line to right side wall: 0.2m Molten Steel temperature: T o =1900K Inlet and outlet launders on opposites of vessels and directed radially Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 6
Input Flow Conditions Steel flow rate: 99.5 tons/hour Steel density: 7020 kg/m 3 Mean inlet velocity: 0.232 m/s Cold argon gas flow rate: Hot argon gas flow rate: Mean bubble injection velocity: Mean bubble size: 0.49 m 3 /min 1.314 m 3 /min 0.35 m/s 48mm Discrete multiphase transient flow input: 19 bubbles/0.1s at each of the two gas inlets at the bottom Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 7
Fluid Flow and Particle Motion with Argon Injection Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 8
Fluid Flow with Argon Injection Fluid flow velocity Gas concentration Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 9
Fluid Flow with Argon Injection Turbulent Energy Turbulent Energy Dissipation Rate Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 10
Typical Trajectories of Alumina Inclusions (300µm, 3500 kg/m 3 ) (With Argon Injection) Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 11
Transient Fluid Flow with Argon Injection 500 Bubbles injection per seconds Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 12
Fluid Flow and Particle Motion without Argon Injection Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 13
Fluid Flow Velocities Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 14
Typical Trajectories of Tracer Particle (Without Random Motion) Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 15
Typical Trajectories of Particles (With Random Motion) Tracer particles Alumina Inclusions Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 16
Fluid Flow, Particle Motion, and Mixing in a Argon-Stirred Ladle (300 ton, 4.5m height, 0.5 m 3 /min argon injection, bubble size: 33mm) Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 17
Fluid Flow speed (m/s) gas concentraction (kg/m3) 4.5 0.0 0.3 0.5 0.8 1.0 1m/s 0.0001 0.0006 0.0040 0.0251 0.1585 1.0000 4.5 4 4 3.5 3.5 3 3 2.5 2.5 (m) (m) 2 2 1.5 1.5 1 1 0.5 0.5 0 0 2 1.5 1 (m) 0.5 0 0.5 1 1.5 2 (m) 2 1.5 1 (m) 0.5 0 0.5 1 1.5 2 Velocities Gas Concentration Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 18
Particle Motion and Mixing Time 4.5 4.0 3.5 Calculated gas column shape (0.001 isoline of the gas volume fraction) 7 5 2 1000 50t ASEA-SKF 50t Ar-stirred 58.9t Ar-stirred 6t Ar-stirred 200RH 65kg Water model τ=523ε - 0.4 3.0 2.5 2.0 1.5 50µm 100µm 1.0 300µm 0.5 0.0 6 4 3 1-1.5-1.0-0.5 0.0 0.5 1.0 1.5 (m) Inclusion Trajectories Mixing time, τ (s) 100 10 3456 2 1 Current 300tonne Ar-stirred ladle 0.1 1 10 100 1000 Stirring power, ε (Watt/ton) Mixing Time Average Moving path length before reaching top surface: - 100µm inclusion: 47.0m (285s); - 33mm bubble: 5.0m (3.9s) Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 19
Dissolution of Aluminum Alloy in Ladle 0s 3s Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 20
Overall process summary 10s 50s Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 21
Conclusions 1. The developed Lagrangian-Lagrangian multiphase model can predict the fluid flow, inclusion motion and mixing in vessels of this continuous steelmaking process 2. Without gas injection - Generating short circuiting flow pattern (from the inlet launder quickly and directly to the outlet launder), very detrimental for the process. - The inlet jet impinges strongly against the shallow bottom of the inlet launder, so possible splashing and erosion of the refractory bricks - Possible improvement: making the outlet launder not align directly across the vessel from the inlet. Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 22
Conclusions 3. With gas injection - Generating a strong recirculation in the whole height of the vessel. - Particles recirculate a long time,good for reactions. 4. Useful researches for this project (now and future): - Solute mixing (such as alloy dissolution) (done by L. Zhang and J.Aoki for gas stirred ladles) - Reactions such as deoxidation, impurity element and inclusion removal, depending on mixing phenomena at the interface of slag / metal and gas/metal. - Modeling inclusion nucleation, growth, collision (done by Lifeng ) and interact with gas bubbles (done by Lifeng and Jun Aoki in gas stirred ladles) - Investigating the emulsification of the top slag Univ. of Illinois at Urbana-Champaign Metals Processing Simulation Lab L. Zhang, 2004, 23