Validation of a Probabilistic Mathematical Model for Soybean Seed Separation in a Gravity Column

Abstract

In mathematical models of the separation process in a gravity column, the coefficient of bar vibration is frequently disregarded as seeds and impurities move along the curved surface of the gravity column, which, in turn, affects the quality of separation. It is important to note that the vibration coefficient directly depends on the design parameters of the comb bars in the gravity column. (Research purpose) The research aims to experimentally validate the theoretical findings from a probabilistic mathematical model elucidating the process of soybean seed separation within a gravity column. (Materials and methods) For laboratory research, a gravity column prototype with a single comb was developed. Parameters such as the effective length of the comb bars, the cross-sectional diameter of the bar, and the comb angle inclination were determined. (Results and discussion) It was found out that, under the condition of self-cleaning combs, the separation coefficient μ = 3.6 meters to the power of negative one. During bar vibration, the functions of the bending moment with vertical coefficients a = –30, b = –31.9, and horizontal coefficients a ̃= – 0.58, b ̃= –1.47 Pascal to the power of negative one per meter to the power of negative one, yield a fairly high statistical correlation between the developed mathematical model and experimental values (coefficient of determination 0.96). The effective length of the bars was determined to be within the range from 125 to 150 millimeters of the comb, with a cross-sectional diameter of the bars measuring 1.5 millimeters. (Conclusions) Experimental validation of the probabilistic mathematical model regarding the process of soybean seed separation in the gravity column was conducted. Further ways to intensify the separation process are aimed at finding a new design solution for the shape of the comb bars.