A grinding simulation has been developed based on the population balance mo
del. In the chemical industry, it is usual that a range of products is mill
ed in the same grinding facility. Performing experiments with narrow fracti
ons or tracers to obtain grinding coefficients for each product would be ve
ry cost-intensive. Here, a method is used to determine grinding coefficient
s using algorithms based on only a few laboratory experiments. The data are
approximated with equations for the selection function and the breakage fu
nction, which allows the milling conditions to be calculated for a desired
particle size distribution. Grinding experiments with a stirred ball mill s
how that the method provides a good approximation in practice. (C) 1999 Els
evier Science S.A. All rights reserved.