Predictions of Porosity and Fluid Distribution Through Nonspherical-Packed Columns Richard Caulkin, Xiaodong Jia, Michael Fairweather, and Richard A. Williams Institute of Particle Science and Engineering, School of Process, Environmental and Materials Engineering, University of Leeds, Leeds LS2 9JT, UK DOI 10.1002/aic.12691 Published online July 7, 2011 in Wiley Online Library (wileyonlinelibrary.com). For beds comprised randomly arranged nonspherical particles, the prediction and understanding of the packing characteristics and subsequent fluid flow through the resulting porous media is a longstanding problem for chemical and process engineers. This paper presents the application of a digital modeling approach to particle packing, in which no more than elementary physical concepts are used, with the model using collision points to predict trends in bed structures of particles of different geometry. Lattice Boltzmann modeling (LBM), coupled to the output of the packing model, is used to subsequently assess velocity distribution through the generated structures. Sim- ulation results are compared with data available from the literature, as a means of model validation, where it is demonstrated that the combined approach of the digital packing algorithm and LBM provide a modeling capability that is of value to a range of engineering applications. V V C 2011 American Institute of Chemical Engineers AIChE J, 58: 1503–1512, 2012 Keywords: chemical reactors, packed bed, simulation, voidage, fluid velocity distribution Introduction From the many studies into packed beds, it has been established that bed structure within packed columns is essentially affected by four key parameters, namely, the par- ticles (shape and size distribution), the container (shape and the related tube-to-particle diameter ratio), the method of packing (rate, intensity, bed inlet, and outlet conditions), and the treatment of the final packing matrix (vibration, shaking, and compression). Packed beds contained within cylindrical vessels are widely used in the chemical and process industries, and because of their high surface area ratios, they are the domi- nant type of reactor used for industrial heterogeneous cata- lytic reactions. As such, a substantial body of work has already been undertaken in an attempt to quantify the struc- tural properties of packed beds and the transport processes through them to facilitate the design of more efficient unit operations. Numerical modeling using computational fluid dynamic calculations (CFD) 1–5 has advanced to the stage, where it now has the potential to become a valuable tool in the field of research and design for many disciplines, includ- ing that of chemical reactors. In recent studies, the genera- tion of the packing and the simulation of fluid flow have been combined with subsequent simulation of pore-scale tracer dispersion 6–8 or a subsequent simulation of a chemical reaction. 9,10 However, the majority of these studies have uti- lized the packing of spheres or spheroids in the consideration of flow characteristics. As packed beds are naturally associated with varying degrees of randomness, 11 influenced by the specific structural and geometric make-up, the internal free channels present within individual packed structures should not be disregarded when investigating flow distribution within packed col- umns. 12–15 There is great interest in practical models for use in the improved design of systems such as fixed bed Correspondence concerning this article should be addressed to R. Caulkin at r.caulkin@leeds.ac.uk. V V C 2011 American Institute of Chemical Engineers AIChE Journal 1503 May 2012 Vol. 58, No. 5