A new adaptive automated feedback system for Barkhausen signal measurement

https://doi.org/10.1016/j.sna.2005.11.026Get rights and content

Abstract

This paper describes a new Barkhausen signal measurement system that has recently been developed for use with soft magnetic materials. The key feature of this system is that it uses an automated digital feedback technique capable of fully controlling flux density waveforms. The system is capable of controlling the flux density waveforms not only under sinusoidal conditions, but also under triangular, trapezoidal, and even arbitrary conditions at flux densities of up to 1.7 T. Measurements have been made at magnetizing frequencies 25–75 Hz, and Barkhausen effect signals were detected over the frequency range 3.5–100 kHz.

Introduction

The magnetic Barkhausen effect, which appears as abrupt changes in the magnetization, occurs when a ferromagnetic material is subjected to an externally varying magnetic field [1]. The origin of the effect has been known to be primarily due to the discontinuous domain walls motion through the material caused by imperfections of the material. The imperfections such as inclusions, dislocations, grain boundaries, and voids form pinning sites, which impede domain wall motion until an external energy is provided enough to overcome the local energy barriers created by the pinning sites. The domain wall then suddenly jumps to the next available metastable state.

The Barkhausen signal can be detected using a coil encircling a sample while it is being magnetized. The abrupt change in magnetization induces an electrical voltage in the coil. The Barkhausen effect is very sensitive to the changes in the microstructure and stress of a material. Owing to this sensitivity, the Barkhausen measurements can be used for the non-destructive evaluation of a ferromagnetic material in the application areas such as assessing the quality of heat treatment, measuring surface treatment depth, and detecting grinding burns.

Hartman et al. built a system for measurement of ac Barkhausen noise in electrical steels [2]. Their system could measure Barkhausen noise rapidly and accurately at power frequencies not exceeding flux density of 1.2 T for non-oriented and up to 1.4 T for grain-oriented with an analogue feedback circuit.

Zhu et al. published a paper on multifunctional magnetic Barkhausen emission measurement system [3]. Their system allows flexible control of applied fields and multivariate analysis of Barkhausen signal emissions where Barkhausen emission measurements were made using a 2 Hz triangular excitation signal generated by the arbitrary waveform generator card.

The new Barkhausen measurement system that is introduced in this paper makes use of a computerised control technique where flux density is controlled with the aid of digital feedback for a wide range of materials for non-destructive evaluation. It can generate magnetic flux densities of 1.7 T with increments of 0.1 T with a form factor of 1.11 ± 1% for sinusoidal waveforms. The Barkhausen noise signals were analysed using algorithms embedded in LabVIEW software.

Hysteresis curves and Barkhausen signals for the materials studied here have been modelled using a mean field hysteresis modelling software package combined with a stochastic Barkhausen effect model calculation. These have allowed the both the measured magnetization versus magnetic field (M, H) curves and the Barkhausen voltage versus time to be quantified and understood in terms of the physical parameters of the models, which describe the various components of the magnetization process. Theoretical calculations of both hysteresis and Barkhausen effect have been made using the models and these have been compared with experimental results.

Section snippets

Experimental details

The block diagram of the Barkhausen noise measurement system used for control of the flux density waveform is shown in Fig. 1.

Output waveforms to control the applied flux density were generated through a digital output card, National Instruments PCI-6711(E) with 12-bit resolution with an update rate of 1 MS/s. This output waveform was fed through a power amplifier (PA) that acts like a lowpass filter [4]. An isolating transformer (IT) was used to remove any dc component in the magnetizing

The adaptive digital feedback technique

The digital feedback algorithm consists of five main parts (Fig. 2).

The first thread is the measuring thread (Fig. 2A), in which acquired signals are used to calculate the flux density, B (T), applied magnetic field, H (A/m), power loss per unit mass P (W/kg). This also feeds Fig. 2E which filters, analyse and displays the Barkhausen noise signal.

The second thread (Fig. 2B) is responsible only for digital feedback. The same acquired data were used as for the measuring thread. However,

Experimental results

Two samples were tested; sample 1 (S1) was conventional grain-oriented (CGO) 3% silicon iron steel of 300 mm length, 30 mm width, and 0.27 mm thickness. Sample 2 (S2) was thin film amorphous alloy (Metglas 2605SC) of 290 mm length, 25 mm width, and 0.021 mm thickness. Measurements were taken for flux density waveforms of sinusoidal, triangular and trapezoidal conditions at magnetizing frequencies of 25, 50 and 75 Hz for each condition.

It can be seen that the system is able to control the flux density

Hysteresis and Barkhausen effect modelling

In order to model the Barkhausen effect it is necessary first to model the hysteresis and then superpose the calculated Barkhausen voltage signal on this, because the amplitude of the Barkhausen effect depends on the differential permeability. The hysteresis curves of the materials were modelled using the methods developed previously [6]. Values of the basic hysteresis parameters were determined by fitting the calculated hysteresis curves to the measured data at the lowest frequency (Table 1).

Conclusion

The developed system for measuring Barkhausen noise with adaptive digital feedback control is a fast, user-friendly tool for non-destructive evaluation of materials. Barkhausen noise could be analysed in a wide range of materials at different magnetizing frequencies as well as flux densities of arbitrary waveforms and amplitudes.

Harshad Patel was born in Cardiff, United Kingdom, in 1980, and received his B.Eng. (Hons) from the School of Engineering, Cardiff University, United Kingdom in 2002. As part of his undergraduate degree he worked in industry with Corus Ltd. He is currently in his final year of Ph.D. at Wolfson Centre for Magnetics, Cardiff University. His research topic is Barkhausen Noise.

References (9)

  • K. Hartmann et al.

    A system for measurement of ac Barkhausen noise in electrical steels

    J. Magn. Magn. Mater.

    (2003)
  • D.C. Jiles

    Dynamics of domain magnetization and the Barkhausen effect

    Czech. J. Phys.

    (2000)
  • B. Zhu et al.

    Multifunctional magnetic Barkhausen emission measurement system

    IEEE Trans. Magn.

    (2001)
  • Amcron

    DC-300A series II power amplifier

    Reference Manual

    (1996)
There are more references available in the full text version of this article.

Cited by (27)

  • High-spatial-resolution magnetic Barkhausen noise sensor with shielded receiver

    2020, Sensors and Actuators, A: Physical
    Citation Excerpt :

    Therefore, in the design of MBN sensors, the design of an MBN receiver with high spatial resolution and high stability is an important subject. For characteristics and optimization design of U-shaped yoke MBN sensors, research mainly includes the following aspects: (1) Improving the stability of the test results by controlling the test conditions [11–23]; (2) improving the sensor sensitivity by selecting the appropriate coil and magnetic core material and improving the structural design [24–29]; and (3) improving the sensitivity of the sensor by optimizing the frequency-response characteristics of the receiver [30–32]. The representative research includes the study by Stupakov et al. [11], who studied the subject of ensuring repeatability of MBN detection.

  • Measurement methods

    2020, Barkhausen Noise for Non-destructive Testing and Materials Characterization in Low Carbon Steels
  • An alternative method to remove excitation field interference from magnetic Barkhause noise

    2014, Journal of Magnetism and Magnetic Materials
    Citation Excerpt :

    This dynamic range corresponds roughly to 12 bit. According to our experience, such resolution is usually sufficient for measuring the MBN and it is frequently used in experiments [17–26]. Hence, we can say that for both types of sensing coils and over the whole excitation frequency range investigated, we were able to measure the MBN with sufficient accuracy.

  • System for controllable magnetic measurement with direct field determination

    2012, Journal of Magnetism and Magnetic Materials
    Citation Excerpt :

    This also leads to Hmax symmetrization for the measured Hall fields with a good accuracy. The typical Vgen shift needed for the symmetrization is approximately 5 mV, which is comparable to the power amplifier offset and less than 1% of Vgen amplitude [13,20]. For the present, this symmetrization subroutine is used separately, after the waveform correction and just before the measurement block.

  • Governing conditions of repeatable Barkhausen noise response

    2009, Journal of Magnetism and Magnetic Materials
    Citation Excerpt :

    The difference between the RMS value and spectral energy in the 2–70 kHz range demonstrates the level of remaining noise outside this frequency region. According to Parseval's theorem, the total spectral energy in the whole range up to the Nyquist frequency (power spectrum) equals the RMS value of the signal, which is always fulfilled experimentally [12,22]. The influence of the BN sensor lift-off with respect to the sample surface on the measurement results is presented in Fig. 5.

View all citing articles on Scopus

Harshad Patel was born in Cardiff, United Kingdom, in 1980, and received his B.Eng. (Hons) from the School of Engineering, Cardiff University, United Kingdom in 2002. As part of his undergraduate degree he worked in industry with Corus Ltd. He is currently in his final year of Ph.D. at Wolfson Centre for Magnetics, Cardiff University. His research topic is Barkhausen Noise.

Stan Zurek was born in Radomsko, Poland on 5th March 1975. In 2000 he received the M.Sc. degree in Czestochowa University of Technology, Czestochowa, Poland. In 2005 he received the Ph.D. degree from School of Engineering, Cardiff University, Cardiff, United Kingdom. In the same year he joined the Wolfson Centre for Magnetics, School of Engineering, Cardiff University, Cardiff, United Kingdom, where he is currently a Research Associate. His main interests are rotational magnetisation, measurements of magnetic properties, magnetic sensors and digital techniques in the measurements.

Turgut Meydan is a Reader at Cardiff University.

David Jiles has a Ph.D. in Applied Physics and a D.Sc. in Physics and has spent his professional life working in magnetics. He has authored over four hundred scientific papers, two books: Introduction to Magnetism and Magnetic Materials and Introduction to the Electronic Properties of Materials. He is a Fellow of the American Physical Society, Fellow of the Institute of Electrical and Electronic Engineers and Fellow of the Magnetics Society in the United States. He is also Fellow of the Institution of Electrical Engineers, Fellow of the Institute of Physics and Fellow of the Institute of Mathematics and its Applications in the United Kingdom. Until 2005 he was Anson Marston Distinguished Professor at Iowa State University. He is currently Professor of Magnetics and Director of the Wolfson Centre for Magnetics at Cardiff University.

Lu Li was born in China. He completed his Ph.D. at Iowa State University in 2004. His Ph.D. was in modelling of the magneto-mechanical effect.

1

Permanent address: Ames Laboratory, Iowa State University, Ames, IA 50011, USA.

View full text