Modeling of electromagnetically driven flows occurring in Hall-Heroult cells

The horizontal turbulent motion of the cryolite bath and the metal pad in the electrolytic cells used to produce aluminum is caused by electromagnetic forces arising from interaction between large currents and strong magnetic fields. The movement of the bath is worth examining since higher velocities of the bath impair the current efficiency, and melt the "ledge" which protects the side wall of a Hall cell. Mathematical models have been developed to optimize cell performance under different operating conditions and busbar designs by predicting the flow profiles. However, experimental confirmation of effects predicted by these models has been hindered by the difficulty of measuring velocities in the cryolite bath and by the expenses of carrying out experiments in an operating cell. Hence a physical model, on a one-tenth scale, of a Hall cell was employed to generate reliable data on fluid velocities and magnetic fields, which can be used to test mathematical models. Moreover, this model should yield an insight into the way the electromagnetically driven flows in Hall cells would change under different abnormal conditions (e.g. cold anode, "muck", missing anode, and presence of ferromagnetic materials near the cell) and under various bus bar designs.; Wood's alloy simulated the cryolite bath of a Hall cell and circulated horizontally under the influence of electromagnetic forces from the passage of currents at an anodic current density of 1 amp/cm{dollar}sp2{dollar}. The electromagnetic probe reliably measured the horizontal velocities of Wood's alloy and the Hall-effect probe determined the magnetic fields. Magnetic fields and velocities were computed using existing mathematical models.; Results show that a cold anode, muck, or missing anode increase the circulation of Wood's alloy. The "quarter riser" and "end riser" designs increased the velocities of Wood's alloy compared to an intermediate design. Nonuniform current distribution on the cathode bottom results in an asymmetric fluid flow profile. The pattern of fluid flow can be explained from the torques of the electromagnetic forces. Predicted magnetic fields matched well with the measured ones indicating that the mathematical models for field calculations are reasonably correct. However, the predicted velocities showed only an approximate match with the measured values.