(578c) Investigation of Single and Multiphase Fluid Flow Behavior in Stirred Tanks By Means of CFD and Radioactive Particle Tracking

The understanding of the flow patterns and local hydrodynamics in stirred tanks is important for process design and scale-up.  Thanks to the availability of faster and faster computers, the industry has shown in recent years an increasing interest in the use of CFD models for the simulation of single and multiphase fluid flows. However, the closure rules inherent to such complex models (e.g. for turbulence and phase interactions) often lead to uncertainties and further justify the need for experimental data obtained through reliable validation techniques. The opaque nature of the vessel walls or the fluids themselves renders inefficient conventional measurement techniques such as optical and laser-based methods (e.g. LDA, PIV). On the other hand, the so-called Radioactive Particle Tracking (RPT) technique is a suitable method that has been used to provide extensive data about liquid flow patterns for both single and multiphase flow systems. It is based on the detection of the Lagrangian motion of a radioactive tracer inside the vessel. Our research group has employed this technique for many years to provide local hydrodynamic data in various chemical reactors and mixing processes. 

This work focuses on the assessment of CFD simulations of single and multiphase turbulent fluid flow in a laboratory-scale stirred tank using RPT data. First, the effect of impeller type, rotational speed and clearance on the flow patterns and agitation behavior of single-phase systems are investigated.  Three types of impeller are considered: a Rushton turbine, a Pitched blade turbine (PBT) in down-pumping mode and a concave disc impeller (CD-6). Next, the mixing behavior of gas-liquid flow in the loaded and completely dispersed regimes with the aforementioned impeller types are investigated numerically with a Eulerian-Eulerian approach and experimentally using RPT. Finally, the importance of developing multiscale models for such complex flows is discussed.