Induction heating past the Curie point

This example demonstrates the fully coupled capability of the ThermNet and MagNet solvers. The problem consists of a workpiece and a driving coil. Eddy current losses in the workpiece raise its temperature past the Curie point, which in turn affects the distribution of the eddy currents. Without a fully coupled solver and temperature dependent material properties the change in the eddy current distribution as the temperature changes would not be modeled correctly. This example also illustrates the economies that are made in meshing, since for the thermal problem only the workpiece itself needs to be meshed. In the magnetic problem the coil and surrounding air must also be meshed, and moreover the workpiece must be refined to accurately model the skin depth. Using separate meshes for the two solutions reduces the size of the thermal problem, since its meshing requirements are much less stringent in this case.

Results

Click image to see full size version

The workpiece is composed of a material whose magnetic properties vary as a function of temperature. The graph shows the B-H curve at different temperatures. ThermNet interpolates these curves at temperatures between those specified.

Click image to see full size version

The temperature at the base of the workpiece eventually increases beyond the Curie point, which impacts the magnetic properties. The B-H curve in the hot region corresponds to the red trace in the graph of B-H curves.

Click image to see full size version

Here the peak field is driving the material into saturation and this effect is compounded by the effect of the increasing temperature on the material properties. The result is that the skin depth increases as the workpiece heats up. This effect can only be modeled if the eddy current distribution is recalculated at each transient time step, as is done in ThermNet.