[zethani]

I often need to simulate fully developed MHD flow of liquid metals in rectangular ducts under the presence of an applied static magnetic field. This problem can be solved in CFX using the electromagnetic model with electric field setting on electric potential and magnetic field setting on user defined (usually I specify a unidirectional constant field). Fully developed condition is obtained by simulating a tiny slice of the duct (3/4 nodes along the axis) where the "inlet" and "outlet" surfaces are connected with a translational periodicity bounday condition. A fixed mass flow rate is specified as interface model to give a momentum source.

Now, I have observed that sometimes during the simulation electrical currents appear in the streamwise/axial direction, which are clearly unphysical since I am trying to achieve a fully developed condition. The main problem is that these currents do not tend toward zero as the simulation progresses but, conversely, their magnitude can increase leading to completely unphysical results.

I have dealt with this problem for quite some time now but I have been unable to pinpoint what is the actual cause of this behavior. I believe that there is nothing obviously wrong in my input file, since the simulation runs just fine for several thousands iterations with no problem at all.

I guess that more than a question, I am just wondering if anyone in this forum has ever dealt with the same problem and what is his/her experience in addressing/solving it.

Thank you for sharing

zethani

[Opaque]

What version of the software are you currently using?

Since you mention flow in a rectangular duct with a fixed magnetic field specified, you are modeling the "Hartmann problem" or a variation of it, correct?

You may need to contact ANSYS CFX support, and ask for validation of the Hartmann problem.

Are those currents present near the walls? or in the core of the flow?