Multiscale modelling of the magnetic Barkhausen noise energy cycles

International audienceThe magnetic Barkhausen noise (MBN) control is popular for materials characterization and as a magnetic Non-Destructive Testing & Evaluation (NDT & E) method. MBN comes from the erratic and unpredictable magnetic domains motion during the magnetization process. Its correlation to the micro-structural properties is evident. MBN is usually studied through time independent indicators, like the MBNenergy, which is obtained by integrating the square of the MBN voltage signal with respect to the time axis. By plotting the MBN energy as a function of the tangent excitation field H, a hysteresis cycle can be observed. After renormalization, the comparison with the classic induction vs excitation B(H) hysteresis loop provides interesting observations. Similar shapes can be observed if the domain wall contribution is preponderant in the magnetization process. On the contrary, strong differences appear if the magnetization rotation mechanism is stronger. There is no available standard for the exploitation of MBN control devices. Usual procedures rely on rejection thresholds based on empirical relations. In this domain, simulation tools able to refine these thresholds and improve the understanding of the physical behavior are highly desired. In this study, a simulation method combining a multiscale model for the anhysteretic behavior and the Jiles-Atherton theory is proposed to simulate precisely the MBNenergy hysteresis cycles. The use of the multiscale model allows separating the contributions of domain wall motion and magnetization rotation mechanisms. The satisfying simulation results validate the approach and constitute a major step toward a comprehensive simulation tool dedicated to MBN