"Unknown error" in Magnetostatic solution

JamesTeslaJamesTesla Member
edited November 13 in Structures


I try to simulate magnetic field around train rail thru which flows current. I set all parameters in a way that they are set in this tutorial:

After clicking "Solve" the simulation runs to 10% and than fails on "unknown error" and there is really no more available specification what has gone wrong... I hate this, when any program doesn't show me details, that he internally must have... That's why I'm asking you, more experienced ones, if you have any idea what to change to make it work. I do attach the files in .wbpz

I would try to use Maxwell but it seems it is not included in Academic version. I also tried same simulation with round wire only, same problem. Also attached. I tried Fluent simulation in my installation of Ansys and that worked... So... any tips? :-)


  • Hello,

    I'm moving this to the Structures category for better product alignment. We should be answering your question soon.

    Thank you.


  • wrbulatwrbulat Forum Coordinator


    I tried running your drat.wbpz. Looking at the solver output you may have received these messages:

    One problem that the MAPDL solver overcame is that legacy SOLID117 elements do not use the (default) distributed solver. But the fatal error was gauging failed. Edge-flux formulation emag elements have no companion contact element. The domain must be a single part (this will result in shared nodes at interfaces between bodies). You would need to use the "Form New Part" feature of Design Modeler to accomplish that:

    SOLID117 is a legacy emag element type that is no longer documented in the solver Help, however it is still by default used in Mechanical for emag Systems. Mechanical may be instructed to instead use the new technology SOLID236 by defining the following variable in the variable manager:

    SOLID236/237 support the distributed solver.

    In drat.wbpz, I see a conductor embedded in an "air box". The "air box" extends beyond the ends of the conductor. This might solve, but strictly speaking, it represents something that cannot exist in physical nature. Current cannot "appear out of nowhere" and enter the end of the conductor. The current is what we call "non-solenoidal".

    We usually cut the "air box" so that its surfaces are flush with the conductor ends, and make those surfaces flux parallel. This is a self-consistent set of conditions: the current is normal to and the magnetic flux tangential to such a surface. You probably will not be allowed to define electrical conditions on the ends of the conductor when they are covered by "air" after forming a single part.

    The SOLID236/237 element types have a myriad of features when accessed in the MAPDL solver environment. Very few of these features are natively exposed in Mechanical.

    I hope this helps!

    Kind regards,


Sign In or Register to comment.