Do you have an example illustrating a procedure that can be used to account for the bidirectional coupling that exists in a magnetohydrodynamic pump? Eddy currents are induced in the electrically conductive fluid due to both the changing field and velocity effects. The eddy currents alter the field and the associated Lorentz forces move the fluid.

The attached example is a 2D planar model of a pressure driven flow of an electrically conductive fluid passing through a duct. A permanent magnet is positioned over the duct. Velocities from the CFD results are used to define the real constants in a conventional velocity effects magnetic model (each element has a unique real constant ID).

The presentation provides more details. The input file contains the APDL that creates the model and executes the coupled field solution in a DO LOOP with physics environments.

Although this isn't a pump so much as it is magnetic stirring due to conductive fluid flowing under a "DC" field source, it should be possible to adapt this method to magnetohydrodynamic pumps.

Note that at release 9.0, the multifield solver allows magnetic forces to be passed to CFD, but does not provide a way for CFD-calculated velocities to be recognized by the electromagnetic solution. In the attached input file, the velocities WERE successfully passed to the electromagnetic solution through real constant tables.

Show Form
No comments yet. Be the first to add a comment!