Search

Abstract: Nuclear magnetic resonance (NMR) imaging was used to measure velocity and density profiles in 3-D conical hoppers fed from an open vertical silo. Discharge of a 200μm-diameter powder in both mass and plug flow was studied with hoppers of different half angles, of 10° and 80°, respectively. An analytical solution for compressible (variable density) mass flow in the 3-D axi-symmetric geometry was also developed following the procedure outlined in and . The density variation and velocity profiles obtained experimentally were compared to predictions of this theory for dense, compressible granular flows. We found, from both theory and experiment, that the powder has to exhibit significant dilation (compressibility) as it is accelerated through the constriction in the hopper. The degree of compressibility was found, experimentally, to be lower than that predicted by the mass flow hopper theory. The powder unexpectedly exhibited a boundary layer with a fully-rough boundary condition in the mass flow hopper. In the funnel-flow hopper, the expected “dead zone” was found around the orifice and extended about one diameter length into the silo. The centerline velocity increased according to an exponential function. [Copyright &y& Elsevier]

Predicting structural and flow characteristics of cohesive powder systems is a technical challenge in many powder technology applications. The discrete element method (DEM) is a promising approach to model the mesoscale structure and flow behavior of cohesive granular systems. Advances in experimental techniques such as X-ray computed tomography (XRCT) have enabled characterization of powder systems with sub-micron resolution. Combined with standard bulk powder characterization methods, this technique allows for mesoscopic structural characteristics of cohesive powder systems to be compared directly with DEM simulation. In this study, DEM and XRCT are used to characterize the structural characteristics of two moderately cohesive roughly spherical “model” powders – glass and stearic acid – under loosely poured and tapped consolidated conditions. The DEM simulations are based on a constitutive elasto-plastic model that makes use of the JKR theory to account for cohesive interparticle surface forces. Powder density and mesoscopic void structure obtained from DEM simulation are compared to characterization results obtained from XRCT image analysis. The results show that DEM can accurately predict the extent of tapping induced densification and mesostructure characteristics in cohesive powders, with superior agreement for stearic acid. [ABSTRACT FROM AUTHOR]