GONZAGA, L. M.; http://lattes.cnpq.br/9723940595359046; GONZAGA, Lígia Mara.
Resumo:
The particle size distribution is one of the most important variables in mineral processing,
and there are several direct and indirect methods for its determination. However, the results
generated by different methods necessarily diverge when compared or when there is a
crossing of data. In this work, particle size distributions, obtained using the cyclosizer and
conventional sieving, were measured in order to define an empirical distribution function,
capable of determining the particle size distribution by sieving using measurements made in
the cyclosizer, using the sieving hybrid between the techniques used. A hybrid sieving was
carried out which allowed the correlation of two diameters obtained by different techniques
using the empirical GaudinSchuhmann
distribution function and the Fuorier series. The
evaluation of the adequacy of the models to the series was carried out using the Root Mean
Square Error (RMSE) and Rsquare
(R2) error measures. For all investigated databases, the
adjustment by the Fourier model was better compared to the one obtained by the GaudinSchuhmann
model. Given the comparison, it was possible to observe that between the ranges
from 75 to 14 μm there was a better fit between the techniques used, the empirical
distribution of GaudinSchuhmann
represented very well the finer particle size ranges,
whereas the Fuorier series presented good adequacy both in thin band as well as thicker
band. It is noteworthy that the method is general and can be applied to irregularly shaped
particles.