In the present article, the performance and the efficiency of ceramic filters for continuous steel filtration in an induction crucible furnace, which is part of the steel casting simulator facility located at Technische Universität Bergakademie Freiberg, is investigated numerically. In order to achieve this objective, a macro-scale simulation for the melt flow in the crucible is coupled with a pore-scale simulation for the flow inside the ceramic filter that is adequately resolved by its detailed filter geometry, obtained from computed tomography scan images. The considerable influence of the filter on the flow field is indicated from the present results. Moreover, the efficiency of the employed filter is also determined and compared for two pore densities.

In this article, a numerical study on the sensitivity, related to the performance of open-cell foams used for the depth filtration of liquid metals, on two characteristic morphological properties is presented. Therefore, simulations of fluid flow and particle transport inside an artificial foam structure are carried out, whose porosity and strut shape is varied within a certain expected range. For comparison purposes, however, the simulations are also performed for three typical ceramic foam filters (CFF) with pore densities of 20 and 30 PPI, whose geometries are obtained from CT scans. In order to allow for a comparison between the different structures, a reference length is introduced that relies upon the actual ratio of pores per volume. The evaluation is mainly based on the comparison of the hydraulic tortuosity, the viscous and the inertial permeability coefficients as well as the initial filtration coefficient for alumina inclusions, with their size ranging from of 10 to 40 mm at process conditions typically encountered during the aluminum filtration. It is shown that the ratio of filtration coefficient and pressure drop increases with the porosity, while the material distribution between the struts and the joints is less influential. Finally, the article also provides information on the anisotropy of CFFs and on the transition behavior from steady to unsteady flow in open-cell foams.

No abstract

This article demonstrates an approach for the numerical modeling of open cell ceramic foams, in order to investigate the thermo-mechanical behavior during metal melt filtration processes. The described methods include the creation of the geometric models, the fluid-dynamic simulations of the melt flow using the lattice boltzmann method, and finite element simulations of the ceramic foam under general loads. The paper mainly focuses on the representative geometry generation process, such that the modeled geometry replicates the real foam as far as possible, since high quality models are essential for the quantitative analyses to be reliable. The present parametric studies demonstrate the manner, in which the foam properties may be influenced by changing the geometric and the topological properties, in order to improve the behavior during filtration processes.

In the present study, an alternative approach for the numerical investigation of short-term depth filtration of liquid metals within ceramic foam filters (CFF) is proposed, which is expected to drastically reduce the computational effort of simulations. In this methodology, the flow field is solved in a repeating periodic element of the filter structure, while the inclusions are tracked on an unfolded flow field with much larger dimensions. In order to demonstrate the performance and utility of this approach for parametric studies, the influence of different parameters on the depth filtration of liquid metal within CFF is investigated for the transient, laminar flow of liquid metal through an idealized two-dimensional filter structure. The fluid flow is numerically solved using the lattice-Boltzmann method and the trajectories of the metal inclusions within the filter are calculated using a Lagrangian approach through one-way coupling. The filtration efficiency is evaluated for inclusions of different size and density ratio and its dependence on different process conditions is analyzed, along with the spatial behavior of the filtration process. Regarding the inclusion properties, the results show that the filtration efficiency is significantly influenced by the size of the inclusions and, in case of large particles, also by the density ratio. Further, the filter porosity affects the filtration process while the direction of gravity is found to be unimportant.