Numerical Simulations of Mould-Electromagnetic Stirring for Round Bloom Strands with Full Coupling Methods
Sprache des Titels:
Proceedings of 6th International Conference on Electromagnetic Processing of Materials
Electromagnetic in-mould-stirring (M-EMS) is a tool widely used in continuous casting of steel blooms for long products. M-EMS is often used to increase the flow velocity inside the liquid core: high flow velocities at the solidification front enhance the transition from columnar to equiaxed solidification. Also M-EMS helps to decrease the number of surface and subsurface defects, as well as to improve the reduction of inclusions and segregations. At the steel plant of voestalpine Stahl Donawitz M-EMS is used for the majority of products.
In this paper the electromagnetic stirring of steel during continuous casting is studied by means of numerical simulations. Because of the hazardous environment during the casting process measurements are very difficult or nearly impossible to perform. Physical 1:1 scaled models of the casting process with water are available and commonly used, but because of the low electric conductivity of water these models are not applicable for cases with electromagnetic stirring. Metals which are liquid at low temperatures are expensive and ? due to the lack of transparency ? measurements are again very complicated. Thus numerical simulations are nearly irreplaceable means to get detailed information of the flow field with electromagnetic stirring.
In these simulations the full coupling between flow field and magnetic field is considered with different methods. The feedback of the liquid velocities on the magnetic forces cannot be neglected, as will be shown. Both methods are using the commercial code FLUENT to calculate the flow field. The first approach uses the FLUENT MHD add-on module to calculate the electromagnetic field. In the second approach the electromagnetic field is calculated via the FEM code ANSYS Emag. The presented approaches are compared to each other and to other simpler methods which are used in the literature in view of the accuracy of the results, their performance and the computational costs.