Conclusions to Chapter 5

1. Taguchi’s methodology provides an alternative to standard factorial design. Its peculiarity resides in the fact that the selection of the experimental design is made from examination of linear graphs, allowing the investigation of the desired interaction effects, based on process knowledge. The physical phenomena on which “classical” electrostatic separation processes are based are known well enough to facilitate the set up of linear graphs for representing the input variables and the predicted interactions between them.

2. The two experimental designs presented in sub-chapter 5.1 clearly proved that the linear-interaction models of the electrostatic separation process could reflect the effects of the main factors in a manner that is satisfactory to most cases of practical interest. At the same time, they pointed out the limitations of Taguchi’s method, which deliberately and sometimes arbitrarily neglects some of the interactions that might be nevertheless relevant to the process.

3. Taguchi’s methodology proved to be a powerful tool for the robust design of electrostatic separation processes, within which there is an inherent variability due to fluctuating ambient conditions.

4. The impact of environmental conditions on the outcome of electrostatic separation is an issue that cannot be occulted at a time when the robust process design becomes a prerequisite of the industrial application of any new technology. Through the proper selection of the operating conditions of the electrostatic separation installation, the process can be made less sensitive to ambient humidity variations, thus avoiding the costly eventualities of product rejection and/or re-treatment. This will hopefully contribute to the design of more robust processes, and answer to the strong industrial demand for reduced sensitiveness of the electrostatic separation efficiency to such variations in the operating conditions of a given installation.

5. High-voltage and roll-speed are two variables that are easily controllable in any commercial electrostatic separator. From this point of view, the procedure described in sub-chapter 5.3 can be readily applied in an industrial environment. The operator should adjust the high-voltage and the roll-speed at the levels recommended by Taguchi’s methodology or by the interaction analysis approach.

6. However, special attention should be given to the preparation of the processed materials: the outcome of the electrostatic separation process is influenced by the metal content of the granular mixture to be sorted. Laboratory and pilot-plant researches previously described in Chapter 4 are in progress for the monitoring of this parameter.

7. Un-expensive technical solutions are available for the continuous measuring of ambient temperature and humidity. However, further work is needed to asses more accurate empirical models of the process, so that to make possible the adjustment of the control variables in response to any variation of the ambient humidity and temperature.