New feature extraction for applied stress detection on ferromagnetic material using magnetic Barkhausen noise

This paper reports on the new feature extraction for the determination of applied stress using magnetic Barkhausen noise (MBN). Low frequency triangular waveform excitation is used to produce alternative magnetic field on case-carburized and tempered En36 steel specimens. Under applied elastic stress, both of the compressive and tensile type, the MBN signal is measured and the profile of MBN signal is obtained. Effective damping in Barkhausen noise for asymmetric shape is studied with compressive and tensile applied stress characterization. Then the skewness of the MBN profile is presented as a new feature for stress detection. Non-linear behavior and sensitivity of the new feature are compared with root mean square (RMS) peak value and peak position. Domain wall energy and eddy current damping are discussed as the reason of the phenomenon. (C) 2015 Elsevier Ltd. All rights reserved.