Menü des aktuellen Bereichs:



Bubble Stirred Flows

Bubble stirred flows or bubbling beds are commonly known for several industrial processes in multiple fields for diverse purposes (e.g. thermal and species homogenisation).

However, the research done within this work is based on the steel production, where the hot metal desulfurization represents the main background. This process is a major step in refining hot metal and is executed as part of the converter process or within the ladle. By stirring the liquid metal utilising the upward movement of entrained bubbles the melt is homogenised and impurities are transported and accumulated within a buoyant slag layer. This induced turbulent flow is a key parameter for increasing the effectiveness of mixture while reducing negative effects like wall-erosion.

For experimental analysis the hot metal, the introduced stirring gas, and the overlaying slag are commonly substituted by a ‘cold’ system composed of water, pressurized air and a dedicated slag analogon (for example oil). This enables the usage of optical measurement techniques like Particle Image Velocimetry (PIV). Although the setup depicted in Figure 2 would imply a rather simple measurement environment, two phase flows and especially bubble columns with higher air fraction represent a complex task for global measurements. Frequently occurring problems caused by the rising bubbles are the deformation of the laser plane due to reflection/refraction, intense bubble reflection benefits blooming effects, or shadowing and occlusion phenomena to name just a few.

Since the turbulent flow field plays a significant role and proper analysis of turbulent quantities within their full spectrum is related to a high temporal and spatial resolution a multiscale setup was realised to monitor the global flow field as well as a special area of interest. The global flow field is limited in resolution due to the big dimensions and is used for analysing the overall mixing behaviour while the magnification cam acquires at higher resolution in space and time. This increase is used to depict turbulent parameters in more detail at dedicated locations of interest so these parameters can be compared with CFD results more precisely.

Fig. 1: Turbulent Analysis tool (TurbAn) for data-source independent interpretation of vector fields.

As this project should also provide multiple data-sets for validating novel CFD models a tool for analysing both, experimental- and CFD- generated data, the same way was developed. This turbulent analysis tool (TurbAn see Fig 1) is capable of computing standard turbulent parameters like ensemble means, fluctuating velocities, kinetic energies, turbulent length and time scales, vorticity, etc.

Furthermore, built in functions allows the user to check quickly for statistical convergence of the provided data, peak locking errors within PIV calculations, or generating visualisations of desired parameters in a static way or for time dependent analysis within a movie.

Recent efforts are made for implementing a vortex detection and tracking algorithm to enhance the functionality even further while maintaining an ‘as simple as possible’ data presentation for fast interpretability.

Fig. 2: Experimental setup with optical discrimination of multiscale image acquisition

(Bernhard König)