Estimation of the Core Losses of Induction Machines Based on the Corrected Epstein Test Method

A revised method based on the standard Epstein test is proposed to improve the accuracy of the estimation of the core losses of induction machines (IMs). Since the magnetic field strength patterns at the four corners of the standard test samples are non-uniform, high-order excess losses at the corners would deteriorate the accuracy of the model for core losses. To address this issue, test samples were designed which have an appropriate geometry for the measurement of core losses under the test conditions. The excess losses could be eliminated by introducing an air-gap magnetic reluctance at the four corners so that the core-loss data from the tests consist of only hysteresis and eddy current parts. In experimental validations, three 5-HP IMs were bench-tested for the verification of the core-loss estimation. The use of commercial computer-aided engineering software packages allowed for the estimated core losses to be recorded and compared. The results indicate that high-order excess losses could account for 7% of the total core losses.