Magnetic resonance imaging and numerical modelling of hydrodynamics in vibrated bubbling fluidized beds
Fluidized beds are widely applied in several industrial fields, such as chemical engineering, energy conversion, and pharmaceuticals production. Vibrated bubbling fluidized beds are typically used in the food industry for drying of powders due to the resulting enhanced particle movement. Despite their widespread use, our fundamental physical understanding of the hydrodynamics occurring within fluidized beds is still limited and based on empirical or semi-empirical relations, that are valid only within narrow parameter ranges, making the scale-up of the apparatus design unreliable and expensive. Typical analysis methods are either intrusive and thus interfering with the process or based on 2D data. In recent years, non-intrusive tomographic techniques are increasingly used to study fluidized beds. Magnetic resonance imaging (MRI), a technique that has been mainly applied in the medical field, is particularly suited for obtaining spatially and temporally resolved dynamic information from the interior of fluidized beds. Besides experimental investigations, numerical modelling of fluidized beds has become more and more important during the last years due to increased computational power.The main aim of this project is to combine expertise in fluidized bed technology and novel simulation approaches with innovative real-time magnetic resonance imaging. An MRI system granted by DFG as part of a major instrumentation application (INST 153/152-1 FUGG) is currently being installed at TUHH. The MRI system will feature a vertical bore orientation of a diameter similar to those of clinical systems and a magnetic field strength of 3 Tesla. The system will employ powerful magnetic field gradients and radiofrequency hardware allowing high spatial and temporal resolution. Fluidized beds with and without vibration and different particle types will be investigated to get a detailed and thorough understanding of the effect of particle properties and fluidization parameters on the 3D hydrodynamics. The 3D MRI data obtained in this project are compared with frequently used correlations and with the results of intrusive probe measurements and image data obtained in pseudo-2D apparatuses. In addition to these experimental investigations, computational fluid dynamics simulations coupled with the discrete element model (CFD-DEM) will be performed. The MRI data are again used to test and validate these numerical models and benchmark their prediction accuracy of fluidization hydrodynamics. A specific focus here will be on the validation of frequently used models for drag forces and turbulence as well as coarse graining approaches. The vibration will be modelled by moving geometry and mesh. The real-time MRI data of vibrated and non-vibrated fluidized beds will allow us to validate the accuracy of existing correlations or develop new and more accurate correlations with the unparalleled detail in spatial and temporal resolution in the novel MRI system at TUHH.