Article ID: ISIJINT-2018-141
This paper developed a quantitative understanding of the effect of ore properties on granulation effectiveness for different iron ore types based on the mechanistic model proposed by J.D. Litster et al. The granulation effectiveness was measured by the magnitude of x0.5, which was postulated to be a function of the layer binding strength. Water was adopted as the only binder for all granulation tests. A wide range of moisture contents, up to the optimum moisture content that yielded an optimum permeability, was tested for each ore type. The typical structure of granules made from different ores was investigated to reveal the effect of mineralogical properties of ore particles on the growth of granules. The results indicate that the granulation effectiveness x0.5 varies markedly with ore type and moisture content. Feed particle size distribution and moisture for granulation are paramount factors determining the granulation effectiveness x0.5, which has a single relationship with the mass mean size of layering particles independent of ore type. The mass mean size of layering particles is dependent on the inherent size distribution of raw materials and the mass fraction of layering particles, while the latter is linear to available moisture content. The model works well for some ore types except for those with unreasonable size distributions: 1) excessive intermediate particles meanwhile lacking of fine particles; or 2) excessive ultrafine particles. An effective criteria closely related to layer binding strength, (1-ε)S/εd, is proposed to determine the applicability of the model.