Atomistic Monte Carlo Simulations on the Melting Transition of Iron at Ambient Pressure

Atomistic computer simulations are playing an increasingly important role in high temperature processes, bridging the gap between theory and experiment, and taking cutting edge research into unchartered territories. A fundamental understanding of the melting behaviour of BCC iron is of great significance in steelmaking operations. We report Monte Carlo simulations on the melting transition of iron using three well known interaction potentials, namely the Rosato potential, the Mendelev potential and the Sutton-Chen potential. Depending on the potential used, the interactions between iron atoms varied from 1/r², as well as 1/r^8.14 at short distances, and the range of applicability was found to vary from 3.5 Å to 9.5 Å. These atomic level computer simulations were carried out using a range of simulation variables such as the size of the simulation cell, step size, boundary conditions and statistical ensemble.  The melting transition was identified clearly in all cases through discontinuities in energy and increases in local disorder. The simulation data however showed a significant scatter. Interaction potentials of iron based on system characteristics at 0 K were found somewhat limited for quantitative high temperature simulations (~1800K). Iron potentials developed for earth science applications were not found to be suitable for these investigations. These computer simulations therefore point towards a significant gap in the knowledge base in the atomistic modelling of molten iron, especially for steelmaking applications.