Use of numerical and physical simulation on multi-phase flows and particle separation in tundishes as a contribution to increased cleanness of steels
An investigation of the melt flow in a steel mill is nearly impossible due to the high temperatures and a lack of optical accessibility. For this reason, physical and numerical simulations have to be done. Water can be used for the physical simulation because the kinematic viscosities of liquid steel and water are comparable, thus the flow of both fluids is similar. Considering the similarity laws, the results for water modelling can be transformed to the real process. A series of experimental tests have been carried out in a 1:3 scale water model of a single-strand tundish. The separation of particles in water, depending on their size, has been determined. The separation rate is estimated for particles with diameters of dP = 1–200 μm. Moreover the influence of major factors – like casting speed and flow regulator, on the separation rate of the particles is investigated. The physical modelling shows the dependence of the separation rate from these factors. The separation rate can be improved by decreasing the volume flow rate or by an impact pad installation.
Experimental results were used to verify the mathematical model, which is used to predict the non-metallic inclusion distribution and separation in steel, and in consequence to improve its purity. The results of the liquid flow field obtained for numerical simulations are found in a good agreement with the experimental tests results, obtained by DPIV technique. In terms of discrete phase modelling (particles in water), the standard boundary condition (trap/reflect) available in FLUENT code is not suitable to model complex physical phenomena occurring on the steel-slag (water-air) or steel-refractory lining interfaces. For this reason, the modified boundary conditions for particle separation on a free surface and at the bottom and side walls of the tundish — calculated from the equilibrium of gravity, buoyancy and drag forces acting on a particle in the fluid are used. Results obtained with UDF function give a similar distribution of the particle separation rate as an experimental study in the water model. Since the mathematical modelling results are finding in a good agreement with physical one, the model is used to predict the separation rate of the nonmetallic inclusions in steel bath. Based on that model, one can attempt to determine the optimal working conditions for the investigated tundish, which could yield to better steel quality.
The project was funded by the Deutsche Forschungsgemeinschaft (DFG) under the reference number PF 394/11–1.