Electromagnetic field analysis softwareEMSolution

EMSolution Basics 4-6
Handling of motion

EMSolution handles motion in three ways as described in EMSolution Basics 2.

 

The first is using the 2-potential method, in which the magnetic field/current source can be moved independently of the mesh. However, the source magnetic field must be obtained by the Biot-Savart law, which is not applicable when a magnetic material or conductor moves together with the current source.

 

The second is the sliding method. In this method, a regularly divided sliding surface is set up and the potentials of the fixed and moving parts are connected. At each step, the position of the sliding surfaces and the finite element mesh changes, and the connections between the elements change, thus changing the connections in the overall matrix. However, the finite element mesh does not need to be repartitioned, and the change is fast. At the sliding surface, elements are not conformed at the shape function level, so element partitioning tangent to the sliding surface is somewhat limited. However, it is a useful and effective method for practical purposes.

 

The third method simulates motion by deforming the mesh of the air region around the moving part in time. Two meshes are prepared at both ends, and the nodal positions at each time point are obtained by interpolation. In this case, the connections in the overall matrix do not change in time, but the element geometry changes, so the element matrix changes, so the overall matrix must be updated. The problem with this approach is that the deformation of the element during the move is large. It becomes difficult to apply, for example, when moving to a point where the distance from the fixed part be-comes small. However, since the analysis can be performed without recreating the elements, it is expected to be faster than the method of recreating the mesh of the deformed part, especially in three-dimensional analysis.

 

The method described here is basically a method that does not re-quire mesh resegmentation. However, at the present time, both methods have limitations in their applicability, and we believe that they need to be improved in the future.