Pressure drop investigations in packings of arbitrary shaped particles

Unstructured packings consisting of arbitrary shaped particles are widely used in chemical, biochemical and petrochemical industries as well as in energy technology. In packings passed through by a fluid, pressure drops are of key concern and can be derived experimentally, calculated by empirical correlations or by numerical approaches. Among numerical approaches either resolved flow simulations or porous approaches are feasible. In the latter approach large systems can be addressed at reasonable computational expense. The fluid velocity is addressed as a spatially averaged quantity per fluid cell which is larger than a single particle. Information on the porosity must be provided by either experimental techniques or particle based methods such as the discrete element method (DEM). The DEM can be coupled with computational fluid dynamics (CFD) to a combined DEM-CFD approach and is then applicable to systems involving arbitrary shaped particles. As flow is not resolved in porous approaches information on the pressure drop must be provided by suitable submodels e.g. the combination of the drag force model by Di Felice (Int. J. Multiph. Flow. 20 (1994) 153-159 [1]) and the drag coefficient model by Holzer and Sommerfeld (Powder Technol.184 (2008) 361-365 [2]). As there is an ongoing discussion regarding the validity of these combined submodels, pressure drops in packings of spherical and non-spherical particles are derived by carefully performed experimental investigations as a verification. Numerically obtained results from the DEM-CFD are benchmarked against the experiments and available empirical correlations for the pressure drop. Results are in very good agreement for spheres. For complex shaped particles DEM-CFD simulations can be very flexibly applied. Simulations are generally in good agreement with experiments depending on the particle shape and size and are often better than empirical correlations which are usually tailored towards certain shapes and therefore limited in their usability. (C) 2014 Elsevier B.V. All rights reserved.