The effect of mineralogy and microstructure on sinter solid-state reduction

Abstract

 Ironmaking in the blast furnace is a rather complex process, governed by the generation of multi-phases and fluid-flow conditions. It includes the reduction of iron ore pellet or sinter material towards producing metallic iron. The metallurgical behaviour of sinter and its reduction rate are controlled by its initial physical and chemical properties, as well as the blast furnace gaseous conditions. To improve and optimize the metallurgical performance of sinter for ironmaking, it is necessary to understand the fundamental mechanisms of phase equilibria, microstructural transformations and reaction kinetics as relevant to different sinters in the upper shaft down to the reserve zone of the blast furnace. The objective of the present MSc research is to study the sinter solid state reduction reactions that take place in terms of thermodynamics and metallurgical kinetics when varying the initial material composition under certain temperature and gas atmosphere conditions.
 Six pilot scale sinter materials with strongly variable bulkmineralogical and chemical composition were sized down to fractions of 250-500μm,being a grain range adequately fine to isolate microstructural effects, whileexcluding the impact of meso-porosity and mechanical fractures (>1 mm), as practically aspossible. A industrial sinter sample, obtained from the production lineof Tata Steel Ijmuiden was included in order to verify potential differences inreduction. The starting materials were characterized with XRD, XRF, sizedistribution and BET measurements.
 Thirty six (36) isothermal reduction experiments were performed in the TGA and GERO furnace under two set of conditions; i) T=750^oC/XCO=0.55/N₂=0.5, ii) T=950^oC/XCO=0.65/N₂=0.5 simulating point-conditions of the BRASS test, as relevant to the stability field of Wüstite and solid state reduction. The reduction experiments were interrupted at different reduction times. The 36 reduced samples were further analysed with XRD for phase quantification to then be placed in polished sections for microscopical analysis. The materials were examined under Reflected Light Optical (LOM) and Scanning Electron (SEM) microscopes in order to verify microstructural changes, visualize phase transitions and identify existing stable and meta-stable phases. In total 396 microscopical images were produced. The results of the experiments were compared with thermodynamic models, which show which phases and reduction degree are theoretically expected at equilibrium.
 Hereby,based on research results, the study attempted to give answers to the initialreseach questions in order to confirm or deny the research hypothesis. The mainfindings of the study were mostly qualitative, referring to the mineralogicalchanges observed during reduction and their impact on reduction kinetics.Results verified that starting composition and mineralogy influencesreducibility kinetics, while the way minerals impact reducibility is dependedon the imposed conditions. It was observed that at T= 950^oC/XCO=0.65/N₂=0.5, the differences between the reduction rates of low and high basicitysamples become smaller, due to enhancement of the relative reduction progressof minerals like precipitated Magnetite. At T=750^oC/XCO=0.55/N₂=0.5,the transition of Magnetite to Wüstite occurred in different stages ofreduction progress, amongst samples. In addition, the effect of mineralogy and microstructure could not be distinguished ofthat of open particle porosity.

 Findings obtained from the XRD analysis andmicroscopy verified that Hematite is the most pronemineral to reduction followed by SFCA, while Magnetite stays stable for longer.Even within the samemicrostructure, Hematite is clearly reduced to a greater distance from theparticle exterior than the Ca-ferrites surrounding it, and its normalizeddecrease in the XRD analyses from its initial concentration is clearly fasterthan that for any of the Ca-ferrites (SFCA, CF2). Moreover, the relative reduction of Hematite and SFCA differsbetween the two sets of conditions; More reducing conditions converge thereduction progress of the two phases, due to SFCA’s greater reduction. The reduction fronts of Hematite and SFCA converge in a single sinter particlewith higher temperature conditions.Two types of unreacted SFCA wereidentified under the microscope, one Fe-rich SFCA and one Ca-Al-rich SFCA. SFCA1 starts reducing into a multiphase intergrowth, comprised by an Fe-rich pathand a Ca-Si-Al -rich path, while the high Ca type SFCA demonstrates one -to-onephase transition.

The reduction experiments inthe GERO showed that longer-time experiments always give the same samplesequence based on reduction progress; WCS108 is the most reduced, followed byWCS90, WCS86, the industrial sinter MH1785, WCS62 and WCS94, whilst noparticular differences were observed in reduction behavior of the industrialsinter MH1785/21 compared to pilot-pot sinter samples. In addition, the comparisons between the reductionexperiments and the thermodynamic predictions showed that there is mostly qualitative agreement, but results differ quantitatively.

Finally, based on the  findings of the research work conducted, the study hypothesis was confirmed; Sinter microstructure and mineralogy influences sinter solid statereduction.