Experimental investigation and DEM-CFD analysis of Darcy–Forchheimer flows in randomly packed bed systems of wood particles

Abstract

Understanding the packing structure and pressure drop across a randomly packed bed of wood particles is essential for the design and control of wood drying, pyrolysis, and gasification processes. This study utilizes experimental and micro-scale simulation methods to explore fluid dynamics within packed bed systems of wood particles and glass spheres. Pressure drop and velocity data from experiments and simulations were fitted using Darcy’s law and the Darcy–Forchheimer law to derive key parameters: permeability (K) and Forchheimer coefficient (β), which were then compared with existing correlations. Experiments were conducted in packed beds of wood pellets (Re = 11.5 to 185.1) and glass spheres (Re = 28.3 to 455.7). The Discrete Element Method (DEM) was used to generate packed bed structures of cylindrical and spherical particles, corresponding to the experiments. Flow within the beds was modeled using the incompressible Navier–Stokes equations, with detailed analyses of streamlines and vorticity. CFD results indicated critical Reynolds numbers of 10.1 for glass spheres and 4.1 for wood pellets, marking the transition from Darcy to Forchheimer flow regimes. Beyond these values, the formation of vortices indicated nonlinear effects. Experiments showed that K values were $2.95×10^{-7} m^{2}$ and β values $1.22×103 m^{−1}$ for glass spheres; and for wood pellets, K values were $9.82×10^{−8} m^{2}$ and β values $3.04 × 103 m^{−1}$. Using experimental results as references, simulation errors were lower than those from the correlations. Specifically, for wood pellets, simulation errors were 13.54% for K and 10.20% for β, while correlation errors were 42.57% for K and 7.89% for β. This indicates that simulation results are more reliable than existing correlations.