Skip to main content
To KTH's start page To KTH's start page

Mathematical and Experimental Study on Filtration of Solid Inclusions from Molten Aluminium and Steel

Time: Fri 2023-03-31 10.00

Location: Sefström, Brinellvägen 23, Stockholm

Video link:

Language: English

Subject area: Materials Science and Engineering

Doctoral student: Shahin Akbarnejad , Processer

Opponent: Professor Torbjörn Carlberg, Mittuniversitetet

Supervisor: Professor Pär Jönsson, Processer; Professor Dong-Yuan Sheng, Processer; Universitets lektor Björn Glaser, Processer

Export to calendar


Aluminum and steel have been the most produced metal and alloy, respectively, for many years. Their extensive use in various industries, their fundamental role in our everyday life, and their excellent recycling characteristics are the major driving forces for development of their production towards more sustainable processes. A successful integration and application of molten metal filtration from unwanted inclusions in production processes could result in reducing scrap, rework and would provide a cleaner molten metal which could lead to production of metallic materials with enhanced mechanical properties.  Filtration of aluminum melts by ceramic foam filters is an established process in aluminum industry. Ceramic filters are also used in steel foundries to remove inclusions from the melt prior to casting to the mold. However, the use of ceramic filters is either limited to specific types of alloys or casts or to specific filters with large pores and openings. As a result, utilization of ceramic filters in the steel industry has limitations in capturing inclusions, where specifically small size inclusions may not be captured.  

 This research work aims at contributing to the global effort in developing the molten metal production processes to become more sustainable and to increase the quality of the final product. To be specific, it is aimed at shedding more light into filtration applications and the use of ceramic filters for removal of solid non-metallic inclusions from molten aluminum and steel. Thus, permeability characteristics of single 30, 50, and 80 Pore Per Inch (PPI) alumina Ceramic Foam Filter (CFF) grades as well as stacks of three 30, three 50, and three 80 PPI alumina CFF grades were both experimentally and numerically obtained and studied. This provides the information needed to estimate the pressure required to prime and/or push the molten aluminum through the filters. The pressure could either be built up by gravitational or other forces. It has been shown recently that it is possible to prime such filters with electromagnetic forces and filter solid inclusions from molten aluminum. Lastly, physical refining of molten steel from solid alumina inclusions through monolithic extruded square-celled alumina ceramic filter was investigated and studied with a developed mathematical Computational Fluid Dynamics (CFD) model as well as the particle trajectories of inclusions in the size range of 1 to 100 [µm]. 

 The experimentally obtained permeability characteristics as well as the obtained pressure gradient profiles of the single 30, 50, and 80 PPI CFFs were compared to previous research findings from the literature. Overall, a good agreement between the current and previous findings was found. It was also shown that fluid bypassing should be avoided during permeability experiments, otherwise deviations as high as 60% may occur. It was also revealed that similar permeability characteristics for the stacked filters, compared to single filters, could be achieved. However, an about three times higher pressure gradient or pressure needs to be applied when using a stack of three identical PPI filters compared to using single filters. The numerical simulations also validated the experimental findings of the permeability experiments.

 The CFD simulations and particle trajectories of the solid alumina inclusions in molten steel through the monolithic alumina filter revealed that it was possible to capture all particles larger than 50 [µm]. However, it was not possible to capture all particles smaller than 50 [µm] due to the applied simulation approach as well as current simulation limitations in the software.