Dr. Stefan Fleckenstein
Modeling and simulation of molecular mass transfer at deformable fluid interfaces
Mass transfer processes across deformable interfaces play an already crucial and further increasing role in industrial applications. These range from carbon dioxide separation in power plants to widely used bubble column reactors in chemical engineering. Process intensification leads to a growing need for precise understanding of the underlying chemical and hydrodynamic phenomena, which have not yet been tackled to full extent. Experimental methods are often not capable of providing the local data necessary for an in depth understanding. In contrast, numerical methods have seen a significant upswing due to refined techniques and an increase of computational power in recent years. These can fill the aparent need for local data and thus complement empirical research.
One of the goals of this project is to deepen the understanding of local processes adjacent to deformable interfaces. A key tool to gain this insight are realistic simulations of species transport and reaction, e.g. in the wake of rising gas bubbles. Moreover, for industrial mass transfer applications the hereby obtained results must be distilled to applicable rules for real world situations. In the course of these investigations we therefore aim at providing reliable correlations for (among others) Sherwood numbers at relevant Schmidt numbers. To do so, accurate and rigorously derived mathematical models for mass transfer of multicomponent chemical mixtures have to be used and (where not available) developed. In order to make these utilizable for the chosen finite volume approach, the problem has to be reformulated in a suitable way as a set of conservation laws. To simulate the dynamics of this system of conservation equations, we use and extend the volume of uid code FS3D that has been developed at Stuttgart University's ITLR and the CSI. The employed numerical method uses a two-scalar approach for the transport of species, which allows the retention of a sharp molecular concentration distribution in case of non transfer components. Core to this approach is the storage of existing species in both dispersed and continuous phase in different variables and modeling the transport through the interface by additional terms. The numerical methods that are needed to facilitate the simulations will continuously be developed so as to incorporate not yet considered but important physical phenomena and to resemble more closely the underlying models. These methods will then be validated by comparison to experiments where the necessary local and integral information is accessible.
Comparison of numerical and experimental mass transfer data. Left: Comparison of concentration profiles at varying angles, Sc = 330. Right: Comparison of concentration fields around rising bubbles. Numerical simulations were done at CSI, experiments by Prof. M. Schlüter, TU Hamburg-Harburg.
I would like to thank the German Research Foundation (DFG) for the financial support within the scope of PAK 119.