Aluminum investment casting with magnetic field control of the mould filling process

From a fluid mechanical point of view, the filling of the mould during aluminum investment casting processes comprises in essence a flow in a U-bend. It is characterized by a high value of the pouring velocity in the early stage, which decreases in the course of the process. This initial high velocity poses distinct problems such as entrapment of bubbles or debris. At present, different types of filters are used for the removal of inclusions, a desired secondary effect of which is a reduction of the melt velocity.
A non-invasive contact-less solution to control the flow of liquid aluminum is the application of a static (DC) magnetic field. Numerical calculations were performed to simulate the filling process and in particular the influence of the field. The free surface problem, which occurs in the riser of the casting unit, was taken into account by a volume-of-fluid method. 3D transient calculations employing the commercial finite-element code FIDAP were carried out for a simplified model system as well as for the real aluminum casting unit. The term for the electromagnetic force was implemented via a user defined subroutine while an additional equation for the electrical potential was solved. End effects owing to the limited size of the magnet poles were taken into account.
In parallel to the simulations, model experiments were performed using the eutectic alloy InGaSn (Tmelt = 10oC). The casting unit was modeled by a perspex pattern, and ultrasonic Doppler veloci­metry was applied for detailed acquisition of velocity data. Such measurements constitute a profound basis for the validation of the numerical simulations. On account of the excellent agreement, an up- scaling towards the realistic aluminium casting process is justified.
Finally, realistic tests with liquid aluminium were performed at an industrial installation. The primarily aimed at influence of the magnetic field, i.e. the attenuation of velocity peaks in the beginning of the process, could be clearly demonstrated. In a second set of experiments, the strength of the DC field was adapted to the process. At start-up, the maximum braking force was applied for a fixed time, followed by a reduction with increasing filling-level of the casting unit. This schedule provided a distinct damping of the peak velocities without the drawback of a significant prolongation in filling time. A remarkable diminishment of defects in the casting product could be achieved.