Main characteristics of analysis methods in EMSolution
Standardized analysis methods such as the $A-\phi$ method, the edge element finite element method, and the ICCG method were incorporated from the beginning to reduce analysis volume and increase speed, enabling large-scale analysis.
Tree gauges by tree structures and $\phi$=0 gauges can be imposed [13]. Recently, it has become clear that calculations are faster when no gauge is imposed (gauge indefinite), but these gauges reduce the number of unknowns and reduce the com-putational capacity. Since 2000, gauge indefinite has been more commonly used.
The two-potential ($A-A_r$) method, which uses reduced magnetic potential, is employed. $A_r$ can be replaced by the magnetic scalar potential $\Omega_r$.
In nonlinear analysis, the convergence of nonlinear iterative calculations using the ICCG and Newton-Raphson methods is adjusted to improve the calculation speed of nonlinear cal-culations.
For transient analysis, the Crank-Nicolson $\phi$ method is used.
The nodal force method is uniquely developed and employed to calculate electromagnetic forces, including their distribution. This eliminates the need for cumbersome specification of integration surfaces as in the Maxwell stress method. Lorentz force in non-magnetic materials can also be calculated.
Two-dimensional analysis is analyzed as a three-dimensional problem. That is, it is analyzed as a one-layer finite element mesh. This allows for a unified interpretation of the two- and three-dimensional problem. Although there is some computational waste because the problem is treated as a three-dimensional problem, the degrees of freedom and other factors are the same, making it equivalent to a normal two-dimensional analysis.