CFD-DEM simulation of a circulating fluidized bed using a particle location based multi-level coarse-graining approach
Valerie Brandt, J. Grabowski, Nico Jurtz, Matthias Kraume, Harald Kruggel‐Emden
Abstract
Circulating fluidized beds (CFB) comprising of riser, cyclone, and standpipe subsystems are frequently used industrially. The coupling between Computational Fluid Dynamics (CFD) and the Discrete Element Method (DEM) is a popular way to simulate CFBs. To speed-up CFD-DEM, coarse-graining (CG) is a numerical technique that bundles multiple particles into a representative grain to reduce the computational effort. Employing conventional CG using a global grain size often leads to inefficient computations, especially when the complexity of fluid-solid flow structures varies locally between different subsystems. Multi-level coarse-graining (MCG) allows to alter the grain size locally to model specific regions of the CFB with an adapted resolution. Thereby, the challenge in the context of the CFBs is that specific zones requiring a finer resolution of the solid phase than others are not initially apparent. Addressing this aspect, a novel framework was developed in which subsystems of the CFB are analyzed individually with conventional CG to determine optimal grain sizes for each subsystem. The application of MCG using optimal grain sizes in each subsystem of the CFB then outperforms conventional CG by providing same quality results with less computational effort. • Use of multi-level coarse-graining (MCG) in a circulating fluidized bed (CFB). • For this, subsystems of the CFB are analyzed with CG regarding their criticality. • Obtained results allow defining CG ratios and zones as part of a MCG simulation. • MCG simulation results are successfully benchmarked against experimental results. • MCG allows reducing computation times while maintaining accuracy.