Mathematical modelling tools for the optimisation of direct smelting processes

Direct smelting operations involve the strong interaction of a wide range of complex physico-chemical processes. Moreover, for such processes to be efficient, these interactions have to be optimised to yield the desired set of chemical reactions and exchanges of heat and mass amongst the variety of gaseous, liquid and solid phases. This paper focuses upon a consideration of Computational Fluid Dynamics (CFD) based models developed to represent the Hlsmelt(R) direct smelting process. The models are extremely sophisticated and at every stage have challenged the limits of CFD technology, as well as the adequacy of constitutive sub-models to represent the chemical reaction/combustion phenomena. From an unprecedented synergy amongst process metallurgists, experimental scientists and CFD modellers, sophisticated, comprehensive and well-validated models of the process have evolved. The paper highlights some of the key state-of-the-art CFD techniques developed for the models and the role of specially designed experiments in parameter estimation and validation against plant measurements.