| In order to study the thermodynamic and distribution behavior of vanadium in converters of PZH Steel, it is necessary to investigate the component activities in CaO-MgO-Al2O3-V2O3-SiO2 slag system which is obtained based on the composition of vanadium slag. The concentration of vanadium can be calculated from the activities of SiO2, MgO, A12O3, V2O3 in CaO-MgO-Al2O3-V2O3-SiO2 slag when the system get to equilibrium between slags and iron, and reducing agent like C, SiC, Mg and Al etc. are adopted. Furthermore, the distribution ratio of V between slag and metal phase is conducive to know the distribution tendency of vanadium in converters which can be calculated with the activities data above. This process has directive significance for promoting the reduction degree of vanadium oxide, improving the recovery of vanadium and optimizing the process level of vanadium smelting.The binary activities data of FeO-V2O3, CaO-V2O3, Al2O3-V2O3, Cr2O3-V2O3, SiO2-V2O3 and MgO-V2O3 are obtained from the binary phase diagram with the formulas raised by Zhou Guozhi. The corresponding parameters Bij, Bji and Kij are fitted by the molecular interaction volume model (MIVM) and its pseudo-multicomponent approach. This will make preparations for predicting the thermodynamic properties of multicomponent slag system by using MIVM and its pseudo-multicomponent approach.The activities of Fe in Fe-Cr-P system are predicted by MIVM and UIPF and compared with the experimental data. The average relative errors of calculated by MIVM and UIPF are 5.7%and 38%, respectively. The average relative errors of activities of Si in Cu-Fe-Si system calculated by MIVM and Wagner are 59%and 65%, respectively, compared with the experimental data. The activities of each component in Fe-Mn-Si system predicted by MIVM are in good agreement with experimental data. On this basis, MIVM and UIPF are applied to predict the thermodynamic properties of Fe-Cr-V, Fe-Si-V systems containing V, which lack experimental data. What’s more, the effects of component content on activities are also discussed in this work.This work has used Coexistence theory and MIVM to predict the activities of four different multicomponent oxide systems and the average sum of relative errors are 12% and 17%, respectively, compared with the experimental data. The results show that the MIVM is superior to Coexistence theory model which confirm the accuracy and feasibility of MIVM further. On this basis, we predict the thermodynamic properties of multi-slags containing V2O3 by MIVM.The activities of V2O3 in the CaO-SiO2-V2O3 systems are predicted by MIVM and the average relative errors are 33%compared with the experimental data. And the average relative errors of activities of SiO2 in the CaO-MgO-Al2O3-SiO2 system are 13% predicted by MIVM. Then the activities of each component in MgO-V2O3-SiO2 and CaO-MgO-V2O3-SiO2 slag systems are predicted, and the effects of basicity on the component activities are discussed further. The accuracy of the thermodynamic properties of multicomponent system predicted by MIVM are verified and the calculations has a close association with its sub-binary system. The accuracy of multicomponent system depends on the accuracy of fitting the sub-binary system directly. That is to say that the higher accuracy when fitting the sub-binary system used, the better accuracy of the multicomponent system will have.In order to build models for predicting the activities of each component in CaO-MgO-Al2O3-V2O3-SiO2 quintuple slag system, MIVM and its pseudo-multicomponent approach are applied in this work to predict the activities of each component in slags containing V2O3. The average relative errors of calculations obtained by MIVM are 44%,respectively, compared with experiments. What more, the distribution ratio of V between slag and metal, Lv, are further studied and analyzed when using C and SiC as reducing agent. The results show that the reducing capability of SiC is stronger than C under high temperature. |
PDF Full Text Request