The first difference is in dealing with the breakage distribution
function. Austin’s group used Eq. (4) to describe the breakage distribution
function, with its model parameters fitted to the mill feed
and product data. The CSIRO group used the Kelsall function as a
default breakage distribution function that was determined from
experiments with a tumbling mill of 300 mm diameter (Reid
and Stewart, 1970). At the JKMRC it was argued (Narayanan and
Whiten, 1983) that the appearance function aij is ore specific and
it is related to the specific energy. The appearance function of each
ore was determined with a pendulum tester (Narayanan, 1985).