8th International Congress on Science and Technology of Ironmaking - ICSTI 2018

With the rapid development of iron and steel industry in China, the iron ore pellet as an essential raw material plays an important role in the blast furnace. Due to large-scale production of iron ore pellet, microscopic characterization is an important step in quality control. The structure and distribution of minerals in a pellet and the bonding types at point contacts between the grains strongly affect its metallurgical properties, like mechanical strength and reduction performance. Traditionally, pellets are characterized by 2D Optical Microscopy or Scanning Electron Microscopy. However, these approaches cannot be used to observe inner interconnection crystals in the volume space. Therefore, the mineral information is incomplete and the 3D morphology and spatial distribution of the minerals are unknown. In this work, a new 3D characterization methodology based on serial sectioning of digital microscopy and image analysis was developed, automating some steps of the characterization. The discrimination of the main solid phases such as hematite, magnetite and silicates, and of pores can be achieved by threshold segmentation in 3D space. This approach enables the acquisition of high-resolution images in a cost-effective manner and greatly assists in the further study of the microstructure of the minerals in an iron ore pellet. The results of this study will contribute to further understanding of the mineral morphology effects on the properties of pellet.