An approach to machine modeling by using finite reluctances is proposed. The finite reluctance approach is well suited for fast modeling and dynamic simulation of electrical machines in general and of fractional-slot surface-mounted permanent magnet machines in particular. The latter usually have very large air gap and slots, with relevant leakage and large tangential components of flux density in air gap, slot, and magnets as well. Modeling tools as winding functions are not accurate enough in this case, and thus time-stepping finite elements are used, which are slow and with limited time-step resolution (i.e. low simulation frequency). The approach used in this paper uses magnetic networks built on the principle of the cell theory. Network reluctances and magnet MMFs are defined by using unique rules. This paper shows that, besides faster dynamic simulations, the finite reluctance approach allows the graphical representation of the magnetic vector and scalar fields inside the machine, in a way very similar to the more cumbersome finite element method, but with a much smaller computational effort. The results from both the methods are presented and compared for a case study of permanent magnet linear synchronous motor.
