Temperature influence and measurement uncertainty on NdFeB magnetic characteristics

Hard magnetic materials require specific technical data as accurate as possible to be used in designing and manufacturing devices with permanent magnets integrated into advanced systems. However, determining the specific permanent magnet characteristics is accompanied by a series of measurement errors that depend on the measurement method (in closed or open magnetic circuit, with continuous or pulsed magnetization), the geometry of the tested samples and their mechanical and thermal manufacturing. Among extrinsic factors of influence, the temperature remains one of the most important measurement error sources. For this reason, all measurement regulations specify the value of the testing temperature. The decrease in measurement uncertainty consists of corrections that are made on the magnetic characteristics (eddy current correction, demagnetization factor correction, etc.). A way to reduce the random errors in experimental determinations is the uses of several identical samples and consecutive measurements on the same sample. The resulting value strings are processed with statistical methods (e.g., Dixon test, Grubbs test, Cochran test, etc.) to eliminate the extreme values and estimate the values with a certain confidence interval. The present paper deals with the obtaining the characteristic parameters and the demagnetization curves for NdFeB type N38 cylindrical samples with the hysteresisgraph method, for the temperature range 25°C-120°C, by performing 10 successive measurements on the same sample for every chosen temperature value. Statistical data processing allows the temperature influence and measurement uncertainty on NdFeB magnetic characteristics to be established, namely the coercivity, the magnetic remanence, the maximum magnetic energy density and the coefficients of variation with the temperature of these parameters.