High-fidelity DEM-CFD modeling of packed bed reactors for process intensification

A high-fidelity DEM-CFD model is newly proposed for process intensification of packed bed reactors. The discrete element method (DEM) is employed for simulating random packing under gravity with hundreds of spheres in a cylindrical tube. It is verified that the DEM is capable of constituting a packed bed whose voidage well agrees with the literature according to particle-to-tube diameter ratio. The particles in or near contact are cylindrically bridged to reduce the fine and skewed computational cells around the contact point. A computational domain is then represented by connecting gaps between spherical particles each without geometrical simplification. It is shown that the pressure loss through the bed sufficiently agrees with a correlation that takes into account the particle-to-tube diameter ratio. Subsequently to the validation, the model capability for process intensification is conceptually demonstrated by specifying arbitrary boundary condition on each particle. Particles simulating inert are mixed among hot catalytic particles in laminar and random blending manners. It is confirmed that the blending style significantly affects the temperature distribution in the bed and hence it would be a design factor to be optimized by the high-fidelity DEM-CFD model.