Optimization of Three-Dimensional Screening Surface Parameters: Multifactor Modeling and Industrial Validation

Abstract

The technological efficiency of bulk-material separation is determined not only by the area and geometry of the apertures but also by the dynamics of the multilayer material flow above the screening surface. Conventional approaches to separator optimization are primarily based on varying kinematic operating conditions, the inclination angle, or the absolute geometric dimensions of the working elements. However, such an approach does not ensure adequate coordination between the surface geometry, particle dimensions, and the structural thickness of the material layer, thereby limiting the transferability of the obtained results to other grain materials and industrial-scale machines.

 The aim of this study was to establish the nonlinear patterns of the combined effects of the relative height of spatially profiled elements, the number of grain-material sublayers, and the open-area ratio of the perforated surface on separation capacity at a specified separation completeness, as well as to verify the scalability of the identified parameters under industrial operating conditions. The scientific distinctiveness of the proposed approach lies in selecting optimization factors not as isolated absolute machine parameters, but as a system of interrelated geometric and structural characteristics. The height of the three-dimensional element was normalized relative to the characteristic thickness of the undersize component; the material-layer thickness was represented by the number of equivalent sublayers; and the geometric throughput potential of the surface was characterized by the open-area ratio. This approach ensured the physical interpretability of the factors and enabled the transfer of the identified relationships across equipment of different scales. The experiment was conducted using a maize grain mixture and a three-factor, three-level Box–Behnken design. The response variable was separation capacity, subject to the mandatory condition that the separation completeness of the impurity component was not lower than 95%. A second-order regression model was developed. Homogeneity of the experimental variances was confirmed using Cochran’s test, the significance of the regression coefficients was assessed using Student’s t-test, and model adequacy was verified using Fisher’s F-test at a confidence probability of 0.95. A rational parameter region was identified: a relative profiled-element height of 0.46–0.51, a number of sublayers of 4.5–5.5, and an open-area ratio of 0.270–0.285. The recommended central configuration was 0.49, 5.0, and 0.277, respectively. Industrial testing of the optimized screening surfaces on Petkus K-527 and BSH-100 separators during maize cleaning confirmed a 20–30% increase in throughput compared with conventional screens. For the BSH-100 separator, throughput increased from 57 to 72 t/h during maize cleaning and from 33 to 40 t/h during sunflower-seed cleaning. At the same time, the specific energy consumption decreased from 0.026 to 0.021 kWh/t for maize and from 0.045 to 0.0375 kWh/t for sunflower seeds.

The results demonstrate that the normalized coordination of profiled-surface geometry with the discrete structure of the grain layer represents a scalable principle for intensifying separation rather than a local effect limited to a particular laboratory-scale installation.