This paper presents a method for the determination of the mean path length of the 25 cm Epstein frame. The proposed method employs two Epstein frames of different dimensions. A standard 25 cm and a revised 17.5 cm Epstein frame were built. The corner areas of the two frames were kept identical. By subtracting the absolute power loss (in watts) measured in strips of the same material by means of the two frames and combining it with the difference in size of the two frames, the exact mean path length of the Epstein frame can be accurately obtained for any magnetizing condition including the case of controlled distorted (nonsinusoidal) magnetization. Results indicate that material permeability and anisotropy are the most important factors that affect the mean path length, whereas the peak flux density and magnetizing frequency also play an important role

This paper presents the rotational power loss P, curves measured for nonoriented, conventional grain-oriented and high-permeability grain-oriented electrical steel under controlled loci of flux density B, or magnetic field H. The P is measured at 10, 50, and 250 Hz with the radius of circular loci of B being controlled from 0.5 to 2.0 T and the radius of circular loci of H from 50 to 20 kA/m. The well-known phenomenon of decreasing P at higher magnetization is confirmed in the results. The maximum P under controlled circular H can be up to 2.5 times higher than under controlled circular B and up to six times higher than the value of alternating loss for anisotropic electrical steels. In thicker laminations at higher frequencies, the peaks in P curves vanish due to increased contribution of eddy currents