Font Size: a A A

Mathematical Modeling Of Multiple Physical Fields And Its Application In Aluminum Reduction Cells

Posted on:2009-08-20Degree:DoctorType:Dissertation
Country:ChinaCandidate:W LiuFull Text:PDF
GTID:1101360278954185Subject:Non-ferrous metallurgy
Abstract/Summary:PDF Full Text Request
Aluminum reduction cell is the core equipment for industrial electrowinning of aluminum. Its developments and advancements represent up-to-date improvements of aluminum reduction technologies. In aluminum reduction cells multiple physical fields take place, develop and interact with each other inside the cell and affect economic indexes e.g. energy consumption, current efficiency and pot life time during production. Therefore to build up the liable, versatile and easy-to-use multiple physical coupled simulation models are of great significance to develop advanced reduction cells with higher current efficiency, higher labor productivity, lower energy cost and longer pot life.The coupled simulations of aluminum reduction cells are very complicated. That tens of media materials with different shapes distribute inside and outside the cell makes some preparation work for numerical calculation so complex including grid meshing, boundary conditions applying and etc. Solution procedures or difficulties for individual field or coupled fields are quite different. Some problems such as coupled relations, contact phenomena, magnetic open boundary and convergence are much difficult to deal with.The paper focuses on five physical fields in the aluminum reduction cell, which are the electric, magnetic, thermal, flow and stress fields respectively. The integrated simulation models for the multiple physical fields were set up based on less boundary condition, reasonable partitions of field domains and shared model data and result data between different physical fields. These models were employed in simulation researches on large-scale, super-scale and drained cells.The main achievements are summarized as follows:(1) From the point of view to decrease method errors, contact equations were incorporated into the electro-magnetic model to eliminate calculation errors of the horizontal current and vertical magnetic component. The multi-cell model including inside inductors and outside busbar circuits and the electro-to-magnetic transfer model for busbar circuits were built up. Results indicate that using the contact model makes electric fluxes distributed smoothly and horizontal current decreased greatly. With the use of contact equations calculation error of vertical magnetic component was decreased to a less extent for the case with the uniform distributions of the electric field, but by 0.6 mT for the case with the nonuniform distributions of the electric field. The changeable cross-section busbar model makes calculation error decreased by 0.3-0.4 mT than the constant one no matter how the electric field distributes.(2) The single-domain electrolyte-aluminum two phase flow model was built up based on Volume of Fluid (VOF) interface tracking and user-defined interpolation functions for discrete electromagnetic forces to simulate the static flow filed. The model was observed to achieve convergence with the maximum residual level less than 1×10-4 after 2 000-3 000 iterations. The electro-magneto-flow model was applied on the 350 kA prebake cell of one plant located at Henan and well validated for feasibility and accuracy.(3) As the modern prebake anode cell with side risers is concerned, impacts of inside conductors and outside busbars on distributions of the electric, magnetic and flow field were systematically studied. Characteristics of the magnetic distribution and methods to design busbar were summarized. The general relation was obtained that the distribution of one field coordinates with those of other fields. Namely if electric field distributes more smoothly, so does the magnetic field. The electrolyte-aluminum interface distortion also becomes much flatter.(4) Surface fitting function based on orthogonal least square method was established to deal with thousands of discrete data of magnetic vertical component from the magnetic model. This function can make a breakthrough of the limits of the constant or linear expression in the wave stability analysis and improve accuracy. Magnetic results under a range of aspect ratio of pots were calculated and imported into the MHD model to find out how instability varies with the aspect ratio. The results show that magnetic fields distribute similarly for the cells with different aspect ratios and reduce with increasing aspect ratio, which therefore cause the pots more stable. Instability analysis also show that total magnetic field after appended by busbar circuits makes the pots become less stable. (5) The transient heating-up thermal and stress models during bake-out and start-up periods were developed. They were used on 300 kA pot at one pilot smelter in Henan. It is found that the designed coke bed laying method with non-uniform resistivity reduces temperature difference by 8.0%-30.0% and heating-up rate by 4.5%-12.5% respectively in the lining, and decreases the risk of early failure of the cell effectively. Suppose all materials are linear elastic, pot displacement and thermal stress are the largest with semi-graphitic blocks as cathodes and the smallest with graphitized blocks as cathodes respectively at the end of the bake-out. Howerer after 30 d thermal and sodium expansion stage displacement and stress become largest with anthracitic blocks and smallest with graphitized blocks respectively.(6) As required by the drained reduction cell and the super-scale reduction cell fundamental design issues are to keep thermal balance and design optimized magnetic field. One reasonable technical and structure scheme with good thermal balance for drained cells, and one structure and busbar design configuration with vertical magnetic component under 3.5 mT for super-scale cells were found out finally.
Keywords/Search Tags:aluminum reduction, multiple physical fields, busbar design, magneto-hydro-dynamic instability, numerical simulation
PDF Full Text Request
Related items