Advances in Mathematical Modelling of Sintering Performance of Iron Ore Fines

The prediction of sintering performance of iron ore fines and the possibility of targeted optimisation of specific sinter properties are very important for both the iron ore industry and related research organisations. Physical pilot-scale sinter runs require large amounts of raw materials and are expensive, so it is critical to target large test work programs with many different sintering parameters effectively. To predict values for sintering performance parameters such as tumble index (TI), low temperature reduction disintegration index (RDI) and productivity, a comprehensive database of pilot-scale sintering experimental results has been established and empirical modeling conducted. Similarly to earlier work, physical, chemical and mineralogical characteristics of the iron ores have been considered. However, in contrast to the previous work where only one textural parameter was taken into account, namely the proportion of dense haematite ore texture, the current model includes the abundances of several different textures which were combined into different textural factors corresponding to different sinter properties. Coefficients for the variables within specific regression equations can provide a better understanding of the effect of the variables on the corresponding sintering performance. The modelling results have also been used for predicting the sintering performance of mixtures that were not part of the database used for establishing the models. Comparison of experimental and modeling results for the verification test data is also reported in this paper.