Today, furnace design still proceeds mainly by extrapolation from existing furnaces. Investigating existing furnaces shows the dangers of underestimating the impact of apparently small modifications: e.g., larger pits to accommodate higher anodes can result in furnaces with substandard performance. Side effects such as soot creation have then to be accepted. This paper presents an approach to bake furnace design which completely eliminates extrapolation from existing furnaces. The same approach can be used to estimate the optimization potential for existing furnaces.

The correct evaluation of the stationary metal-bath interface in aluminum reduction cells is still a source of discussion and controversy. The objective of this paper is to present some calculations of the interface performed by different methods, used in software packages and to compare them with measured metalbath interface profile in a real cell. Our comparison includes Shallow Layer approach and 3D multiphase methods, such as: homogeneous and inhomogeneous volume of fluid (VOF) as well as floating grid method. Different boundary conditions on the top of bath channels are tested in VOF methods. Finally, a simple generic magnetic field and vertical current density are proposed for software benchmarking of this problem.

No abstract

No abstract

No abstract