In EMSolution Basics, the functions and features of EMSolution were introduced with some experience. In relation to this, the significance and problems of the method in finite element magnetic field analysis are also described.

As mentioned in the introduction, EMSolution has been developed primarily to provide versatile and fast analysis software for large-scale electromagnetic field analysis. Therefore, we rely on other commercial software for mesh generation and post-processing. Considering the large amount of labor required for modeling and mesh generation in actual analysis, the development of modelers and other software that are compatible with electromagnetic field analysis is highly desirable. In recent years, we have begun de-velopment of a modeler, and we would like to make it easier to use by incorporating the opinions of as many people as possible.

One of the problems in magnetic field analysis is the problem of anisotropy and nonlinearity of magnetic materials. In particular, problems involving hysteresis are expected to become more and more necessary in the future in order to improve the efficiency of equipment. However, we feel that it is not a simple matter to put this analysis into practical use. Considering the complexity and diversity of magnetic characteristics, we believe that it is necessary to develop a comprehensive analysis method that includes measurement of magnetic characteristics, design and fabrication of actual equipment, and not just program development. Recently, we pioneered the implementation of a hysteresis analysis method using PlayModel. ^[33] As mentioned earlier, we believe that we should make this technology usable for design in cooperation with measurement.

Another problem is analysis involving microstructures. For example, it is impossible to model a laminated iron core with many steel plates stacked on top of each other using the finite element method. To analyze the magnetic field of such a system, the homogenization method can be applied. It has been confirmed that very good results can be obtained compared to measurements at about commercial frequencies^[34]. However, it might not be applicable to the high-frequency region where skin effects appear, so further improvement is needed.

In recent years, analyses have become larger and larger, and in the future, a single computer will no longer be able to handle them. Therefore, we have recently started working on parallel computation using MPI with multiple nodes, and we hope to be able to analyze 10 times the number of elements (e.g., 30 million elements) in the same amount of time as in the past.

(Note) We completed the development of parallel computation using MPI. The result was released as HPC-EMSolution.

Finally, although I have tried to be as general as possible, please forgive me if I am biased in some cases, especially with respect to citations.