Application Of The Molecular Interaction Volume Model And HA Model To Predict The Enthalpy Of Formation Of Liquid Alloys

In this thesis, the enthalpy of formation of 36 binary liquid alloys has been predicted by the molecular interaction volume model (MIVM) and the Hoch-Arpshofen (HA) model respectively. The results show that the prediction effect of the models is acceptable in both the positive deviation systems and the negative deviation ones, and the former is better than the later. As to the coexisted deviation system, the results of two models are the worst during in three deviation systems. The predicted results of MIVM are better than HA in the positive deviation systems and the coexisted deviation ones, but in the negative deviation systems the MIVM as good as the HA model. The predicted effects of MIVM are worse in strong deviation systems. When one calculates the parameters B_ij and B_ji of the negative deviation systems, the regions in which there are three pairs of roots have been given .Usually of them, the pair root of B_ij B(ji)> 1.0 is the best. The results of coexisted deviation systems are worst, so and one can divide the systems into two sections at the component of x = 0.5, and calculate the parameters B_ij and B_ji. Using the two pairs of the parameters one can predict the enthalpy of formation, and the results are better than before. The reason is that the partial molar enthalpy of coexisted deviation system has changed from positive to negative. Due tothe different of molecular interaction which can be manifested by the enthalpy of formation. The way of subsection can reflect the true molecular interaction.Then based on the results in binary liquid alloys, the enthalpy of formation of 11 ternary liquid alloys and 5 multi-component systems have been predict and compare by MIVM and HA model. The results show that the predicted values of MIVM and HA model are good, but in ternary systems the MIVM is worse than HA, and in 5 multi-component systems the MIVM is better. If in short of the infinite dilution mixing enthalpy ,using the experiment values of enthalpy to fitting the parameters B_ij and B_ji;if no any the experiment values systems ,using Miedema model to calculate the mixing enthalpy and the infinite dilution mixing enthalpy, then fitting the parameters B_ij and B_ji , predict and compare. MIVM and Miedema models will bring certain errors during prediction, which are more obvious in ternary systems, in multi-components systems maybe due to all kinds of errors counterbalance that results in the predicted effect are good. To coexisted deviation systems, the predicted effect of subsection is better than do not.Finally, MIVM is of a clearer physical meaning ,predict only need the value of infinite dilution mixing enthalpy of correlation binary component ,which has certain predominancefor liquid alloys .If the distinction of physical or chemical character of the components is obvious, the applicability of MIVM is worse. HA needs integrate binary systems experiment values , the parameters Wh is an average value, the more big change of mixing enthalpy ,the worse the predicted effects ,which is related to the physical base.