Menü des aktuellen Bereichs:



Why Not Double Hybrid?

Thanks to an ambitious team of experienced senior scientists, who forward independent research and also supervision, I found some time for my own research. Developing new numerical models is a pleasing luxury beside all this administrative and organizing stuff that comes along with my position as head of our Department.

Submerged Entry Nozzle (SEN) flow behaviour is crucial for continuous casting of slab steel since it determines the mould flow pattern. Experiments indicate the existence of highly turbulent secondary vortices in the deflection zone of a bifurcated SEN, which attract gas bubbles at their rotational axes. In addition cyclically detaching gas volumes are formed at the upper port region at higher gas flow-rates.

We addressed this complex multiphase flow by means of hybrid finite volume and lattice-Boltzmann based turbulence models, which, in turn, is coupled to a hybrid discrete and continuous bubble model. In contrast to conventional finite volume based turbulence models on coarse grids, the embedded lattice-Boltzmann simulation is able to capture the highly unsteady dynamics of the secondary vortices. Surprisingly, due to the high efficiency of the lattice-Boltzmann code, this tremendously increased flow resolution comes along with only 9% more computational time.

In our simulations small discrete gas bubbles are traced in a Lagrangian frame of reference. In agreement with experiments bubbles are accumulated in the axes of the secondary vortices, forming characteristic bubble threads. If bubbles coalesce and eventually exceed grid resolution we switch to a continuous representation of large bubble surfaces. The resulting double hybrid simulation model is able to efficiently picture the complex multiphase flow phenomenon occurring in gas stirred SEN operation.

Fig. 1: Embedded lattice-Boltzmann based tracking of discrete bubbles picturing bubble threads at secondary vortex axes (coloured by their vorticity)

Fig. 2: Double hybrid simulation of large bubble formation due to secondary vortices (1) and port recirculation (2)

Fig. 3: Experiment © Maria Thumfart, JKU