Large Scale High-Fidelity DEM-CFD Modeling of Packed Bed Reactors for Process Intensification
Special Symposium - EPIC-1: European Process Intensification Conference - 1
EPIC-1: Poster Session (EPIC - Poster) - P1
Keywords: high-fidelity CFD modeling, packed bed reactor, DEM, heat transfer, pressure loss
Blending catalytic with absorbent and/or inert particles is one of the process-intensification approaches for packed bed reactors. The former is utilized for sorption-enhanced reactors and the latter for the control of temperature distribution and of hot-spots. The laminar or random blending is a possible way to constitute such packing mixtures provided that the particle sizes are comparable. The mixture fraction and particle size ratio are locally and globally significant factors to enhance the blending effect as well as the properties of the additional particles. A recent high-fidelity CFD model, which represents the bed by connecting gaps between many particles each shaped without geometrical simplification, is a potential tool to optimize designs. The precise reality and large scale of packing is crucial for its practical and reliable utilization. In the present study the frame of novel numerical approach for such demands is presented and the capability is conceptually demonstrated. Packed beds are modeled by allowing 220 to 880 spherical particles to fall randomly under gravity into a cylinder whose diameter is 4 times to 8 times larger than the particle diameter. The commercial code EDEM (DEM Solutions Ltd) has been utilized for simulating the packing, which is based on discrete element method (DEM). It is shown that the DEM is capable of constituting appropriate random packing whose voidage well agrees with the literature. The particles in or near contact are cylindrically bridged in the stagnant region to reduce the fine computational cells around the contact point leading to large mesh size that cannot be handled easily. It should be mentioned that the novel bridging method has attained the packing about 10 times larger than previous hi-fi models described by others in terms of the number of particles. The method also enables automatic mesh generation which is critical with such complex geometries. The validity of the novel hi-fi DEM-CFD modeling is further verified in terms of pressure loss and particle-to-fluid heat transfer compared to the widely accepted correlations by means of a commercial code FLUENT6 (Fluent Inc.). The laminar and random blending is demonstrated by specifying different boundary condition on each particle surface, which is also a characteristic capability of the present hi-fidelity model. The number and blending configuration of hot particles as catalytic particles in inert particles with zero heat-flux are varied and the influence on the temperature distribution in the bed are examined from the operational viewpoint, i.e., efficient heat removal from the hot particles, attaining a high average-temperature without hot-spot formation and so on. Finally discussed is the meshing strategy for further enlargement of packing scale, viz., the number of contained particles in the bed.
See the full pdf manuscript of the abstract.
Presented Wednesday 19, 13:30 to 14:40, in session EPIC-1 Poster Session (EPIC - Poster) - P1.
