Optimization of the classification process in the zigzag air classifier for obtaining a high protein sunflower meal – Chemometric and CFD approach
MetadataShow full item record
In this study, sunﬂower meal is ground by a hammer mill after which air zigzag gravitational air classiﬁeris used for separating sunﬂower hulls from the kernels in order to obtain protein rich fractions. Three hammer mill sieves with sieve openings diameter of 3, 2 and 1 mm were used, while three air ﬂows (5, 8.7 and 12.5 m3/h) and three feed rates (30%, 60% an 90% of bowl feeder oscillation maximum rate) were varied during air classiﬁcation process. For describing the effects of the test variables on the observed responses Principal Component Analysis, Standard Score analysis and Response Surface Methodology were used. Beside experimental investigations, CFD model was used for numerical optimization of sunﬂower meal air classiﬁcation process. Air classiﬁcation of hammer milled sunﬂower meal resulted in coarse fractions enriched in protein content. The decrease in sieve openings diameter of the hammer mill sieve increased protein content incoarse fractions of sunﬂower meal obtained at same air ﬂow, and at the same time decreased matchingfraction yield. Increase in air ﬂow lead to the increase in protein content along the same hammer mill sieve. Standard score analysis showed that optimum values for protein content and ratio of coarse and ﬁne fractions have been obtained by using a sieve with 1 mm opening diameter, air ﬂow of 12.5 m3/hand 60% of the maximum feeder rate. Fraction ratio and protein content were mostly affected by the linear term of air ﬂow and the sieve openings diameter of the hammer mill sieve in the Second Order Polynomial model. The main focus of CFD analysis was on the particle simulation and the evaluation of the separation efﬁciency of the zigzag classiﬁer.