A new hydraulic particle classifier: Experimental investigation and modeling

In the present work, a new particle classifier equipped with a novel symmetric quintuplet nozzle for washing stream has been introduced and evaluated experimentally. The nozzle tips has been designed in such way to create strong swirling flow, due to the induced centrifugal upward motion of wash flow. The effect of different operating variables including wash flow rate, feed flow rate, and split ratio on the performance of the classifier has been investigated. A 3-level response surface design has been applied considering the cut-size and the imperfection coefficient as the responses. Also, the regression models has been proposed according to statistical design of experiments to take into account the effect of key operating parameters on the performance of the classifier. The experimental results show that the wash flow rate is the main parameter affecting the performance of this new equipment. Furthermore, the range of the imperfection coefficient for the classifier was between 0.2 and 0.4 for water/sand as a typical system which illustrates a good classification efficiency. This new particle classifier is able to work in a wide operating window. Moreover, the particle short-circuiting has been reduced in this new design.