Distinct element method simulation of mechanical properties of material layer of pellet belt roasting machine

Abstract

The thick layer and graded feeding technology of a belt roasting machine is an effective method for improving the production efficiency and quality index of pellet production, and a reasonable design of the mechanical structure of the layer is the basis for optimizing the heat and mass transfer performance of the layer. Janssen effect and von Mises yield criterion were used to establish a simplified mathematical model describing the elastic and plastic deformation of the green pellet under the action of an external force. The mechanical characteristics of extrusion, contact, and elastic–plastic deformation between green pellet particles in the material layer of the belt roasting machine were modeled using EDEM software. For a green pellet size of 12 mm, as the layer height increases from 300 to 1000 mm, the maximum vertical pressure on the pellets increases from 11.64 to 24.01 N, and the porosity decreases from 27.04% to 22.01%. As the layer height increases, the contact between the green pellets becomes more intense, and the force chain structure of the layer becomes more stable; the Janssen effect is observed when the layer reaches 700 mm. The compressive strength of the green pellets is linearly related to the particle size, and the compressive strength increases with an increase in particle size. At a layer height of 600 mm, as the particle size of the green pellets increases from 8 to 20 mm, the maximum vertical pressure increases from 7.54 to 44.16 N, and the porosity increases from 23.20% to 31.47%, while the yield per unit of the layer decreased by 12.1%. Small particles have a more stable force chain structure, larger comparative area, and higher production efficiency; however, their compressive strength is lower. Large particles have higher compressive strength and good permeability in the layer, but the production efficiency is relatively low. In actual production, a variety of factors should be integrated to optimize the feeding, and a multi-granularity graded feeding is the most ideal feeding.