Study of the reactions between iron ore slag and refractories

For the fabrication of iron ore pellets at LKAB (Luossavaara-Kiirunavaara AB), one of the methods used is to sinter the pellets in a rotary kiln. This Kiln has a drift temperature of 1250°C in average. During this process a coating, causing the wear of the refractories materials, is formed, inducing expensive maintenance periods. The aim of this work is to find the reaction that takes place between the Slag and the bricks, in order to find the parameters that have to be changed to improve the bricks lifetime in the kiln. The use of bricks with higher alumina content is one of the solutions, but it is expensive. The influence of Alkalis and especially Na2O has been studied before in inert atmosphere. In this work it will be studied in air as well as the influence of the alumina content of the brick. Experiments were performed in air atmosphere, using powder samples that have been compressed into pellets. The devices used to characterize the reaction and the product of the reaction were: Kanthal furnace, X-ray diffraction (XRD), dilatometer, differential scanning calorimetry and thermogravimetry (DSC/TG) in-situ mass spectrometer, Scanning electron microscope (SEM) and Energy dispersive spectrometry (EDS). The XRD experiments permitted to show the presence of a glassy phase after a heat treatment of the samples containing alkalis at 1350°C. Dilatometer experiments gave more information about the influence of alkalis: it catalyzed the sintering reaction. The DSC/TG/spectrometer permitted to show three main reactions. One at 1175°C for the samples containing a low content of alkalis and a low content of Alumina. For the samples with addition of 10wt% of Na2CO3, and 20wt% and more of alumina to the brick, a reaction is observed at 1235°C and one with a maximum at 1300°C. The SEM pictures and EDS show the formation of different phases. Unfortunately it was only possible to identify with certitude the Hematite, the Corundum and a glassy phase. The reliability off the EDS measurements is low, due to the oxides used and a special correction of the data had to be done. The conclusions of the study are: - For the samples with a small amount of alkalis, hematite dissolves mullite at 1175°C and forms Fe2O3Al2O3 and a phase with a nepheline composition (Na2O.Al2O3.2SiO2) that melts at 1235°C. - For the samples with high content of alkalis, all the mullite is dissolved. Silica coming form the decomposition of mullite enters completely the melt and needles crystallize in the melt around 1300°C. - Brick with high alumina content start to react with the slag at 1235°C to form needles around 1300°C. In a future work, WDS or TEM will give a higher reliability than EDS, according detection of the phases.Validerat; 20101217 (root