Dr.-Ing. André Weiner
Direct Numerical Simulation of multi-physics reactive mass transfer at single and multiple bubbles
In chemical process engineering one of the main challenges is the conversion of chemical species into the desired product with high yield and selectivity. For many processes this involves mixing and reacting of gaseous species in a liquid bulk phase. Such systems are typically realized through bubble column reactors, in which the gas is sparged in form of bubbles into a liquid phase. Time and length scales of mixing and chemical reactions are mainly influenced by the hydrodynamics within the rising bubble swarm and physical phenomena around and on the gas-liquid interface (boundary layers of species and velocity, surfactants, species transfer resistance, etc.). These phenomena and there interaction are so far not well enough understood, which is why the prediction of yield and selectivity for bubble column reactors is very limited. Direct Numerical Simulation (DNS) is currently the only tool capable to obtain information about local transport processes in order to achieve a better understanding of the integral behavior of such systems.
- Development and validation of reactive subgrid scale models
Chemical reactions of the transfer species together with bulk species reduce the boundary layer thickness drastically which results in an enhanced mass transfer of transfer species. In the context of DNS, this leads to infeasible requirements on the mesh resolution. In subgrid scale modeling, simple ODEs are solved to improve the results for species transfer in interface vicinity on coarser grids .
- Reactive species transport at freely rising bubbles with (partly) immobilised gas-liquid interface
For representative prototype reactions, DNS of bubbles ranging from one to five millimeters will be conducted. Particular attention is payed to substance systems with industrial relevance and realistic material properties (realistic Schmidt numbers). A model for considering surfactant influence in Volume of Fluid (VoF) simulations will be used .
- DNS of reactive mass transfer considering volume change locally
The model for multi-component mass transfer with local volume change by Fleckenstein and Bothe  has to be extended in order to allow for chemical reactions. A standard test case for species transfer with local volume change will be developed.
- DNS of chemisorption of CO2 from Taylor bubbles and development of a benchmark
Taylor flows are of practical relevance for example in chemical (micro) process engineering because of the efficient heat and mass transfer between Taylor bubble and liquid slug. Planned are DNS of oxygen physisorption from a multi-component gas bubble and carbon dioxide chemisorption considering volume change locally. By comparing to experimental results from SPP 1740, a benchmark case for computational reactive mass transfer will be implemented.
- Solving of transformed species equations for mitigating the high Schmidt number problem
Subgrid scale modeling allows to save roughly one level of refinement in DNS (e.g. 20µm instead of 10µm cell size) . The length scales of species transfer in realistic substance systems can be one oder of magnitude or more below the hydrodynamic scales, which is why subgrid scale modeling alone is not enough for accurately simulating mass transfer. A combined approach of transformed species equations and subgrid scale models has potential to mitigate the high Schmidt number problem to a reasonable level.
- Influence of bubble-bubble interactions on physical and reactive species transfer
Having in mind the scale up process in chemical reaction engineering, the next step after single bubble investigations is to focus on bubble-bubble interactions and pseudo bubble swarms. The point of interest here is the interaction of boundary layers, whereby coalescence is not taken into account.