%0 Journal Article %T Numerical and experimental comparison of tracer particle and averaging techniques for particle velocities in a fluidized bed %+ U.S. Department of Energy [Washington] (DOE) %+ West Virginia University [Morgantown] %+ IFP Energies nouvelles (IFPEN) %+ École Polytechnique de Montréal (EPM) %+ University of Cape Town %+ University of Birmingham [Birmingham] %+ University of British Columbia (UBC) %A Xu, Yupeng %A Li, Tingwen %A Lu, Liqiang %A Tebianian, Sina %A Chaouki, Jamal %A Leadbeater, Thomas %A Jafari, Rouzben %A Parker, David, J %A Seville, Jonathan %A Ellis, Naoko %A Grace, John, R %< avec comité de lecture %@ 0009-2509 %J Chemical Engineering Science %I Elsevier %V 195 %P 356-366 %8 2019-02-23 %D 2019 %R 10.1016/j.ces.2018.09.034 %K fluidized bed %K particle tracking %K particle velocity %K numerical simulation %Z Chemical Sciences/Chemical engineeringJournal articles %X Particle tracking methods using emitted radiation are attractive for following motion in opaque systems such as granular materials. Leading examples are Positron Emission Particle Tracking (PEPT) and Radioactive Particle Tracking (RPT). The application of such techniques sometimes requires the use of tracer particles which differ in size, density and/or shape from the particles of interest. This study investigates the extent to which such differences affect the result of the study by using the open source MFIX-DEM software to model particle motion in the travelling fluidized bed experiments. The results are compared with previously reported experimental studies using both PEPT and RPT. Consistent numerical results were obtained for both PEPT and RPT tracer particles. In determining averaged velocities using such techniques, there is a choice to be made between averaging velocities of particles crossing a virtual plane over a period of time (the "face-average" approach) or those passing through a defined volume over time (the "volume-average" approach). The differences between results obtained with these two approaches are shown to be significant in this case, for both computation and experiment. %G English %2 https://ifp.hal.science/hal-02008311/document %2 https://ifp.hal.science/hal-02008311/file/Numerical%20PP.pdf %L hal-02008311 %U https://ifp.hal.science/hal-02008311 %~ IFP