Thermodynamic properties for an alloy system play an important role in the materials science and engineer- ing. Therefore, theoretical calculations having the flexibility to deal with complexity are very useful and have scien- tific meaning. The Hoch-Arpshofen model was deduced from physical principles and is applicable to binary, ternary and larger system using its binary interaction parameters only. Calculations of the activities of Fe-based liquid alloys are calculated using Hoch-Arpshofen model from data on the binary subsystems. Results for the activities for Fe-Au- Ni, Fe-Cr-Ni, Fe-Co-Cr and Fe-Co-Ni systems at required temperature are presented by Hoch-Arpshofen model. The average relative errors of prediction are 7.8%, 4.5%, 4.9~ and 2.7%, respectively. It shows that the calcu- lated results are in good agreement with the experimental data except Fe-Au-Ni system, which exhibits strong inter- action between unlike atoms. The model provides a simple, reliable and general method for calculating the activities for Fe-based liquid alloys.
The component activity of Mn in Fe-C-Mn system as well as the component activities of C and Si in Fe-C-Si system was predicted by applying the pseudo-multicomponent approach of the molecular interaction volume model （MIVM） and the Wagner interaction parameter formalism （WIPF） respectively. The average relative errors between the predicted values of MIVM and the experimental data for the three components were 4.5 0%, 17.0% and 13.0 %, respectively, and those between the calculation results of the WIPF and the experimental data were 18.0K for Mn, 9.0% for C and 27.0% for Si. The results indicated that the MIVM method could better predict the component activity of carbonaceous iron-based solution. Based on the data in an actual blowing process, the MIVM method was applied to predict the component activities of C and V as well as the transition temperature of vanadium oxidization （TTVO） in Fe-C-V-Si quaternary iron-based solution, and a comparative analysis of the predictions against the ex- perimental data was carried out, with their average relative errors being 24.0% for C, 7.3% for V and 1.0% for TTVO respectively. On that basis, the TTVO at Panzhihua Iron and Steel （Group） Co. , Ltd. was estimated by the MIVM method and an expression that the TTVO changed with composition and temperature of iron solutions was obtained by multiple linear regression method. The research results showed that the estimated values were in good agree- ment with the practical data.
