INTERACTION PARAMETERS FOR MONTE CARLO SIMULATION OF THE GRAPHITE/Fe-C MELT INTERFACE: A PHASE TRANSITION STUDY Rita Khanna and Veena Sahajwalla School of Materials Science and Engineering, The University of New South Wales, Sydney, NSW 2052, Australia (Received April 21, 1998) (Accepted December 16, 1998) Introduction The dissolution of carbonaceous materials such as graphite, coal, char etc. in molten iron is one of the key steps in a number of iron and steel making processes and has been the subject of a large number of investigations (1– 4). While most of these studies, both experimental and theoretical, have provided information about the reaction kinetics and the factors affecting the carbon dissolution rate, the atomic level understanding of the processes occurring at the melt/carbon interface is still far from complete. Long-range order, interfacial orientation, various binding energies, temperature etc. are some of the important parameters which are expected to affect the interfacial phenomena. In a recent study, Wu and Sahajwalla (5) have looked at the wetting of solid graphite by Fe-C melts under controlled conditions using sessile drop method. The contact between graphite and melt resulted in non-equilibrium reactive wetting and involved transfer of carbon from the solid to the liquid and iron transfer from the melt to the solid. In an earlier paper (6), we developed a theoretical model of the graphite/Fe-C melt interface and carried out a Monte Carlo simulation of the interfacial region. The atoms in graphite and Fe-C melt were arranged on a rigid hexagonal lattice (space group: P6 3 /mmc). Pair-wise short-range interaction was assumed between the atoms. While the interactions were aniso- tropic in graphite, they were chosen to be isotropic in the liquid phase. Simulations were carried out as a function of temperature, carbon content of the melt and orientation of the interface. Preliminary results of these simulations showed a good qualitative agreement with key experimental trends. Before attempting a quantitative comparison with experimental data, we need to look closely at some of the assumptions used in this simulation. The main assumption of the model is regarding the structure of the liquid phase. While considering the structure of a liquid phase without having access to direct experimental information, one looks at the structures of the crystalline phases in the same alloy system. There have been a few studies of the Fe-C liquid phase where the atoms were assumed to occupy rigid lattice sites. In the interstitial model (7), carbon atoms occupy the octahedral interstitial sites with Fe atoms arranged on a regular fcc lattice. Two-sublattice models using defects and associated solution model postulating molecular-like aggregates have also been used with varying degrees of success (8,9). While considering the graphite/Fe-C melt interface, we had assumed that the atoms in the melt were arranged on a hexagonal lattice. A cubic structure of the melt would cause unnecessary complications of boundary mismatch across the interface. Can this model accurately describe the properties of the Pergamon Scripta Materialia, Vol. 40, No. 11, pp. 1289 –1294, 1999 Elsevier Science Ltd Copyright © 1999 Acta Metallurgica Inc. Printed in the USA. All rights reserved. 1359-6462/99/$–see front matter PII S1359-6462(99)00073-1 1289