Experimental determination of the dispersivity of (unrealizable) perfectly random packings of identical spheres.
By: Ulrich Scheven
From: Univ. Nova Lisboa
At: Complexo Interdisciplinar, Anfiteatro
[2009-12-09]
There is no such thing as a perfectly random pack of mono-disperse spheres, in any laboratory. Consequently tracer dispersion experiments in random packs of mono-disperse solid spheres exhibit quite considerable scatter in reported results, which is due to the inevitable presence of order-inducing boundaries, differing packing protocols, or non-uniform flow injection. By contrast, the conceptual random packs of theory and simulations with periodic boundary conditions are unaffected by these hard experimental realities, and quantitative comparison of theory, simulation, and experiment has therefore been elusive for decades. PFG-NMR experiments and a novel data analysis (Scheven et al., Phys. Rev. Lett., 99, 054502, (2007)) established the missing link between theory and experiment. The experiments and results will be reviewed. Then I will present a related re-analysis of published elution data obtained with traditional step-function tracer experiments by (Han, Bakhta and Carbonell, AIChE J. 31, 277 (1985)) and compare with the ideal random pack. The re-analyzed elution results are consistent with NMR, theory, and simulation (AIChE Journal, DOI: 10.1002/aic.11993).