RESEARCH ON THE MECHANISM AND PARAMETER OPTIMIZATION OF STRATIFIED SCREENING OF MILLET THRESHING MIXTURE BASED ON BRAZIL NUT EFFECT

To address the challenges of millet threshing mixtures, namely their light mass, small volume, minimal differences in component suspension speeds, and susceptibility to interference from impurities during sorting, which lead to high loss and impurity rates, this study applies the Brazil Nut Effect (BNE) to optimize the operating parameters of an air-sieve-type grain sorting device. The discrete element method (DEM) was used to numerically simulate the sieving and segregation processes of the millet threshing mixture, to clarify the grain population structure most conducive to effective sieving, and to identify the main factors influencing segregation and stratification as well as the appropriate parameter ranges. A bench-scale multifactor performance test was conducted to establish regression equations describing the effects of amplitude, stepped jitter plate length, crank rotational speed, and fan shaft rotational speed on loss rate and impurity rate. Using a multi-objective optimization method, the optimal parameter combination was determined as follows: amplitude of 26 mm, stepped jitter plate length of 450 mm, crank rotational speed of 580 rpm, and fan shaft rotational speed of 645 rpm. Under these conditions, the loss rate was 2.68% (with coiling loss below 0.5%) and the impurity rate was 3.95%. The results of this study provide a reference for optimizing the operating parameters of millet cleaning devices.