## Year

1990

## Degree Name

Doctor of Philosophy

## Department

Department of Materials Engineering

## Recommended Citation

Gupta, Govind Sharan, Mathematical modelling of smelting of iron oxide: carbon composites in an induction furnace, Doctor of Philosophy thesis, Department of Materials Engineering, University of Wollongong, 1990. https://ro.uow.edu.au/theses/1509

## Abstract

Brown coal is a widely available low rank coal which has not normally been used in Metallurgical Processes. However, because of its high Hydrogen content, it is a good reductant as well as a part of the source of heat in conjunction with top blown Oxygen/Air. To carry out an experimental evaluation of the optimum smelting conditions in an induction furnace would be expensive in money and effort. Hence, the reason for modelling.

In the Induction Smelting Process self-fluxed composite pellets (made of Iron ore + Brown Coal + Lime) are melted in a single phase induction furnace in conjunction with top blown Oxygen. Therefore, the process has three distinct regions, which are to be modelled: A. Electromagnetic Stirred Region, B. Liquid-Solid Interaction (Bed) Region and, C. Gas Jet and Solid Interaction Region. Each region involves the consideration of non-steady state energy and species balances with momentum transfer.

Region (A) has been solved using the Navier-Stokes and Energy balance equations with Maxwell's equations. To calculate the body force a numerical solution has been developed.

Region (B) has been solved using the momentum, heat and mass transfer equations. A new approach, known as the 'Cell Model', has been adopted to solve this region. All energy, species and momentum balance equations are solved for a typical particle inside a cell, which later gives the response of total system (i.e. the Bed Region) by multiplying the number of particles in the system by the single particle response.

Region (C) has been described by energy, species and momentum balance equations. To calculate the velocity in this region an analytical solution has been adopted.

All three regions have been connected by doing overall heat and mass balances for the entire process. Many complex partial differential equations have been formulated with appropriate boundary conditions to define each region. The governing equations, which are highly non-linear and complex in nature, have been solved by the Finite Difference Method using various techniques, depending on the nature of the individual differential equation. The solution of these equations gives considerable insight into the liquid iron and oxygen temperature and velocity, and FeO and Oxygen concentrations inside the furnace at different times. Also, it predicts the effects of numerous parameters on the entire process such as pellet size, gas jet velocity, preheating of pellet, energy consumption, effect of composition etc.

Computer predictions have been compared with an experiment for the temperature and composition. A velocity measurement device has also been developed to measure the liquid iron velocity at high temperature. Computed and experimental results show resonable agreement.

Although the model has been developed for a particular process, the treatment is general and could be readily extended to any other related process. Also, the solution in each region can be applied separately to define any other related process after a few minor modifications, e.g. the region (A) solution can be applied to various types of induction smelting/refining processes and, the region (B) solution can be applied to different packed-bed processes.

**Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.**