IEEE TRANSACTIONS ON MAGNETICS, VOL. 46, NO. 2, FEBRUARY 2010 513 Magnetic Barkhausen Noise for Characterization of Recovery and Recrystallization Kizkitza Gurruchaga, Ane Martínez-de-Guerenu, Miguel Soto, and Fernando Arizti CEIT and Tecnun (University of Navarra), 20018 San Sebastián, Spain The capacity of magnetic Barkhausen noise (MBN) measurements to characterize recovery and the onset and evolution of recrystal- lization processes occurring during the annealing of cold rolled low carbon steel is analyzed. Cold rolled low carbon steel samples were isothermally annealed at laboratory under different conditions in order to promote various degrees of recovery or recrystallization. The effect of recovery and recrystallization processes on the MBN envelope, the amplitude of the peak of the MBN envelope, the time integral of the MBN envelope and the MBN energy is discussed and related to the microstructural changes produced by these softening processes. The obtained results prove that several parameters derived from the MBN are able to follow the progress of recovery and recrystallization. Index Terms—Low carbon steel, magnetic Barkhausen noise, nondestructive testing, recovery, recrystallization. I. INTRODUCTION M AGNETIC BARKHAUSEN NOISE (MBN) results from the discontinuous movement of magnetic domain walls (DWs) during magnetization of a polycrystalline fer- romagnetic material when the DWs overcome local pinning sites. It is well known that microstructural features, such as dislocation density and the different arrangements of disloca- tions, distribution and size of grains, grain boundaries, second phase precipitates or even applied or residual stresses act as local pinning sites that strongly hinder the movement of DWs and hence influence the MBN. The MBN is sensitive to the microstructure as these microstructural features affect both the pinning strength and the mean free path of the displacement of DWs during magnetization. Several earlier studies have successfully analyzed the influ- ence of variations in dislocation density [1]–[3] and average grain size [4], [5] on MBN. Therefore, MBN is expected to be affected by both recovery and recrystallization, due to the changes that these processes produce during annealing treat- ments on the cold rolled steel microstructure. Recovery involves the rearrangement of dislocations into low energy configura- tions and the annihilation of dislocations, while recrystallization implies the nucleation and growth of new defect free grains [6]. Earlier studies on a cold rolled low carbon steel showed that some parameters derived from magnetic hysteresis loops can be useful to characterize recovery and to detect the onset of recrys- tallization [7], [8]. Especially, the coercive field is able to follow the progress of recovery processes [7], [9], [10] since it is proportional to the square root of the dislocation density [11]. Moreover, in some cases, when the variation of the effective size of the microstructure during recrystallization is very small, can also be employed to characterize the progress of recrystal- lization [7]. Following these studies, this paper researches into Manuscript received June 18, 2009; revised July 22, 2009; accepted July 23, 2009. Current version published January 20, 2010. Corresponding author: A. Martínez-de-Guerenu (e-mail: amartinez@ceit.es). Digital Object Identifier 10.1109/TMAG.2009.2029069 the use of some other parameters derived from MBN measure- ments for the nondestructive characterization of recovery and recrystallization during the annealing of this cold rolled steel. II. EXPERIMENTAL PROCEDURE An industrially produced extra low carbon steel, cold rolled to a final thickness of 0.53 mm through a reduction of 76% and a hot band grain size of 12 m was used in this study. The chemical composition of the steel in wt % was: 0.03C—0. 38Mn—0.0004S—0.11Si—0.037P—0.035Al—0.004N. Sam- ples (200 mm long 14 mm wide) cut perpendicular to the rolling direction of the cold rolled sheet, were isothermally annealed at laboratory at low temperatures, 300, 400, and 500 C, in order to promote recovery and avoid interaction with recrystallization, and at a higher temperature, 575 C, in order to produce samples with various degrees of recrystallization [7]. The MBN measurements were made using a system designed and constructed at the authors’ laboratory [12]. The Barkhausen emission was detected with a small 50 turn encircling search coil surrounding the sample. The tangential magnetic field strength was measured by a Honeywell solid-state SS495A1 Hall sensor placed at the surface of the samples next to the pickup coil. Near saturation MBN was measured applying sinusoidal magnetic field strengths of about 4.1 kA/m at 0.1 Hz. The MBN was filtered with a band-pass filter with cut off frequencies of 1 kHz and 25 kHz. The samples were demagnetized at the test frequency prior to each test by applying a sinusoidal signal whose amplitude diminished gradually, in several cycles, to values close to zero. The MBN activity was characterized by the MBN envelope or root-mean-square RMS profile of MBN , repre- sented as a function of the measured ; the amplitude of the peak of the ; the time integral of the , which represents the area below the envelope over a semi-period of magnetizing cycle; and by the Barkhausen noise energy , which is computed by integrating the time dependence of the squared raw MBN signal over a semi-period of magnetizing cycle [3]. 0018-9464/$26.00 © 2010 IEEE