Time dependent electric, magnetic and hydrodynamic interaction in aluminium electrolysis cells

The waves in commercial cells for electrolytic aluminium production originate at the interface between the liquid aluminium and electrolyte, but their effect can spread into the surrounding busbar network as electric current perturbation, and the total magnetic field acquires a time dependent component. The presented model for the wave development accounts for the nonuniform electric current distribution at the cathode and the whole network of the surrounding busbars. The magnetic field is computed for the continuous current in the fluid zones, all busbars and the ferromagnetic construction elements. When the electric current and the associated magnetic field are computed according to the actual electrical circuit and updated for all times, the instability growth rate is significantly affected. The presented numerical model for the wave and electromagnetic interaction demonstrates how different physical coupling factors are affecting the wave development in the electrolysis cells. These small amplitude self-sustained interface oscillations are damped in the presence of intense turbulent viscosity created by the horizontal circulation velocity field. Additionally, the horizontal circulation vortices create a pressure gradient contributing to the deformation of the interface. Instructive examples for the 500 kA demonstration cell are presented.

Predictive reliability, prognostics and risk assessment for power modules

This paper describes a framework that is being developed for the prediction and analysis of electronics power module reliability both for qualification testing and in-service lifetime prediction. Physics of failure (PoF) reliability methodology using multi-physics high-fidelity and reduced order computer modelling, as well as numerical optimization techniques, are integrated in a dedicated computer modelling environment to meet the needs of the power module designers and manufacturers as well as end-users for both design and maintenance purposes. An example of lifetime prediction for a power module solder interconnect structure is described. Another example is the lifetime prediction of a power module for a railway traction control application. Also in the paper a combined physics of failure and data trending prognostic methodology for the health monitoring of power modules is discussed.