This work addresses the characterization of the degree of mixing to improve
quality and efficiency. The objective was to establish a new method to eva
luate the extent of mixing using a binary component batch system. Batch mix
ing was studied in a two-component system of Newtonian oil with a viscosity
of 3.5 Pa s and suspended particles with a mean diameter of 200 mum. The e
xperiments were performed with three particle loadings: 10%, 20%, and 30% b
y weight. Mixing was monitored as a function of number of revolutions using
magnetic resonance imaging (MRI). Images were acquired from initially segr
egated material to fully uniform material. They were analyzed statistically
in terms of center and spread characteristics, coefficient of variation, l
ength scale, and mixing intensity. As mixing proceeded, the signal intensit
y distributions changed dramatically at approximately 500 revolutions for a
ll three particle concentrations. The length scales decreased as mixing pro
ceeded by factors of approximately 1.5, 3.6, and 6.3 for 10%, 20%, and 30%,
respectively. Mixing intensities decreased from 1 to less than 0.05. Mixin
g intensity (I) versus log(no. of revolutions) showed a linear relationship
with coefficient of determination values (R-2) greater than 0.93 and slope
values in a relatively narrow range. (C) 2001 Elsevier Science Ltd. All ri
ghts reserved.