Inert Anodes For Aluminum Smelting And Corrosion Rates Prediction

Inert anode would mean a technological revolution of the traditional Hall-Heroult process. If successfully developed and applied, inert anode technology could have significant energy, cost, productivity and environmental benefits for the aluminum industry. Over the last few decades, inert anode technology has been one of the research frontiers in the realm of aluminum smelting in developed country. The core of the inert anode technology is the corrosion and corrosion control of anode materials. However, the accurate prediction of corrosion rates holds the key to corrosion control. The corrosion rates of inert anodes are not only related to composition and microstructure of materials, but also related to operating parameters. Aimed at the core of inert anode technology, the primary thrusts of the thesis were to develop an inert anode and evaluate its corrosion behavior systematically. The main results are as follows:Firstly, several metal oxides (ceramic) were prepared based on extensive review and discussion of the open published literature. Proper ceramic composition—NiFe₂O₄ was screened from solubilities performance tests and measured systematically. The effects of low temperature, low cryolitic ratio (CR), and high Al₂O₃ concentration in the performance of minimizing solubilities of NiFe₂O₄ ceramic were determined.Based on the solubilities performance, NiFe₂O₄ cermet materials with excellent microstructure were produced successfully by the powder metallurgy technique. The technically feasible manufacturing process used for the fabrication of inert anodes was brought forward by adjusting sintering speed. In addition, hot pressing of 5%Ni—NiFe₂O₄ cermet was also evaluated for densification of inert anodes. Some basic properties were also measured for the cermet anode materials, for example, thermal shock resistance and electrical conductivity. Such measurements are vitally important to the design, placement, and operation of inert anodes in Hall cells.The partial transient liquid phase (PTLP) bonding of NiFe₂O₄ cermet materials was proposed. The establishment of joining techniques is one of the most important subjects for inert anode technology. The technique isvery important to the voltage drop and durabilities. The partial transient liquid phase (PTLP) bonding of cermet materials were performed successfully at high temperature with Cu-containing interlayer.The corrosion and the behavior of 5%Cu—NiFe2O4 and 5%Ni — NiFe2O4 cermet in cryolite — alumina melts were investigated elementarily. The results showed that 5%Ni — NiFe2O4 cermet had relative low corrosion rates due to high relative densities. The corrosion behavior of 5%Ni—NiFe2O4 cermet anodes in cryolite—alumina melts were investigated systematically as a function of some operating parameters. The corrosion mechanisms of NiFe2O4 cermet anodes were proposed and the corrosion behaviors were explained successfully. The effects of temperature, cryolitic ratio, and A12O3 concentration on the corrosion rates were determined again.Grey realational analysis was proposed to evaluate the effect of operating parameters to corrosion rates of inert anodes. The reliable distinguishment of effect factors was a fundamental requirement for the effective control of corrosion. In inert anode system, the effect factors are coupling and cannot be varied independently. The classical statistical methods were not efficient enough to solve such complex problems and find out the main factors. The grey relational grade ranks the correlation extent of effect factors in the corrosion processes.The corrosion prediction of inert anodes based on mechanism analysises was very difficult because of the complexities of corrosion. The conventional mathematical models (power function model, gray theory model, multiple regression model, et al.) were established with experimental data. The results showed that the conventional mathematical modes had unfeasible corrosion data processing.The artificial neural network (ANN) model was established with experimental data considering the nonlinear relationship of operating parameters and corrosion rates. A corrosion prediction system was developed based on Visual C++ ADO Programming. Furthermore, genetic algorithm coding in real number was engaged to optimize the operating parameters and target parameters. The excellent transplantability and modularization make it convenient for corrosionprediction of other anode materials. The successful development of the system is an innovative creation in aluminum industry. It not only indicated the reliability of functional simulation, but also laid a firm foundation for optimal operation of innovative electrolysis cells. The successful prediction of corrosion rates greatly promoted the self-reliant innovation capability in the inert anode technology.