According to the IFPRI-feasibility study of May 1982 on the evaluation of an electrostatic classifier, it has been shown, that it should be possible to build and operate a classifier in the size range between approximately 0.05 um and 2 um. This statement was valid, however, only if the simplifying assumptions made in the evaluation could be transfered to the actual system.
Approximately 18 months have now been spent on a preliminary system rainly consisting of the feeding and dispersion zones for the extremely fine feed material. and a separation zone following them to determine for instance the ion density available, and the charges transfered to the particles under varying conditions. A china clay with all particles below approximately 0.8 pm and a median particle size of the cumulation distribution by weight of approximately 0.27 pm has been chosen as feed material. The handling. feeding and the dispersion or deayqlomeration of this powder has turned out to be extremely difficult. Even measure- ments of the particle size distribution with a centrifugal sedimentation technique (ALpine SZF 400, Auqsburyl were not as reliable as anticipated. We consequently had to start an investigation of the centrifuge and we are still busy improving it. It might be necessary to additionally analyse the distri- butions obtained by image analysis. The new Quantimet 900, which we received only recently, using REM images could possibly overcome our problems, but will undoubtedly cause new ones. Full financial support to purchase the necessary equipment, high voltage supplies etc. had unfortunately not be granted for the first year. We received, however full funding several months ago and we are now able to finish the final set up and the production of the ccmplete classifier. Work should proceed now at a faster rate.