Multi-scale understanding of agglomeration and deagglomeration in fluidized beds
Due to the very good heat, mass and momentum transfer, fluidized beds are widely applied in several industrial fields, e.g. chemical engineering, combustion, pharmaceutical and food industry, whereby a liquid is often injected for granulation or coating purposes. When the particle is not growing by solidified liquid layers but by wetted particles that stick together, agglomeration takes place, resulting in black-berry shaped and porous particles. The agglomeration may either be desired e.g. for the size-enlargement of food powders, or undesired, e.g. in pharmaceutical coating processes. Alternatively, particle adhesion can occur for particles that undergo glass transition by increasing the temperature. If the process temperature is near the glass transition temperature, the particle local surface may become sticky or rubbery. Particle collision at these spots will subsequently form agglomerates. Independent of the underlying mechanism, the extent of agglomeration of the bed material will determine its further transport within the system and the product quality and morphology. Therefore, this project aims to provide a fundamental understanding of the micro-mechanisms of agglomeration as well as deagglomeration of agglomerates, which is so far missing in literature. These findings will allow a better prediction and control of fluidized bed processes and an increase of the process performance. A multi-scale experimental setup will be used and extended by numerical investigations.In a first step, single particle experiments will be performed. The experimental setup created in a former DFG program (HE 4526/9-1, HE 4526/9-2) will be extended to allow for particle agglomeration. The deagglomeration mechanism will be analyzed by collisions of agglomerates with walls and other particles. Experiments will be performed first at ambient conditions to focus on collision dynamics and statics and later on heat and mass transfer will be integrated. The investigations include model particles as well as visco-elastic materials. The findings from the single particle experiments will be transferred to a fluidized bed process. This will connect the macroscopic process parameters and their influence on the flow behavior dynamics to their influence on the micromechanics of agglomeration on different length and time scales. By comparing the flow patterns and the size distributions of the agglomerates by digital image analysis, a regime map will be obtained for different substance systems indicating the limits for particle agglomeration.CFD-DEM simulations will be used for the numerical investigations. To account for agglomeration and deagglomeration, the effect of liquid bridges will be included into the model. After validation with experimental data, the results from these simulations will be used to construct agglomeration kernels for the open-source dynamic flowsheet simulation tool DYSSOL, which has been developed during DFG Priority Program SPP 1679.