Reaction force not correct.

[Roboticyang]

I create a magnet and a race track coil. The rad.FldEnrFrc function returned the forces are different in both magnitude sand trends using mag_obj and coil_obj as source objects, respectively. I am trying to compare this with my load cell measurement. I think the magnet object version of returned forces are closer to the measurements. And by checking the models in ParaView and comparing with that in ANSYS, the models seem correct to me. The materials are assigned correctly. I compared the B fields of magnet results between Radia and ANSYS.

If needed I can try upload my source code for reviewing.