The application of ferromagnetic steel products is pervasive in society, with important applications arising in electrical steel, oil and gas pipelines, transportation infrastructure, naval structures, aircraft landing gear, and automotive components. Magnetic properties of electrical steel materials play a key role in electrical motors and transformers, with a direct impact on energy efficiency. Measurement of response to magnetization has implications for non-destructive inspection methods, such as magnetic flux leakage, magnetic Barkhausen noise, and metal magnetic memory method. Examples include flaw detection, characterization of material properties, and identification of stress state in steel. An understanding of the magnetic response of steel materials can be facilitated by the use of magnetic objects (MOs). MOs are defined as regions of relatively independent magnetic behavior, typically about the size of a grain, to which fundamental magnetic energy considerations may be applied. This Tutorial outlines mechanisms by which MOs may be applied for modeling magnetic response in steel and presents examples of their application. MOs incorporate material physical properties such as microstructure, grain size, crystallographic texture, the presence of dislocations and impurity elements, and the presence of residual stress and stress load on the component. They can also accommodate a description of the evolution of magnetic domain structure under magnetizing conditions. As the MO model incorporates fundamental physics principles, it allows estimates of physical parameters that can be used to provide insights into the connections between magnetic properties and material properties, including hardness, embrittlement, and the presence of applied and residual stress. Practical applications include non-destructive characterization of the stress state of steel and an improved understanding of magnetic processes in electrical steel. Examples where such models may be applied include magnetic Barkhausen noise and magnetic memory method for the characterization of steel materials. This Tutorial summarizes recent advances in the MO model and its applications, providing the foundation for its further development. Magnetic objects have the potential to provide fundamental explanations and could form the basis for magnetic measurements and magnetization processes, including magnetic flux leakage, magnetic Barkhausen noise, and magnetic hysteresis.

Ferromagnetic iron-based alloys used to form steel products are ubiquitous in our society, with critical applications arising in electrical steels, oil and gas pipelines, naval structures, aircraft landing gear, and automotive components.1 Magnetic properties of these materials can be critical for such applications as electric motors and transformers, where magnetic properties have direct implications for energy efficiency. Inspection of ferromagnetic components using magnetic flux leakage (MFL)2 and pulsed eddy current (PEC)3 also require the characterization of magnetic properties such as permeability and anisotropy. This enables compensation for variations of these parameters on the sizing of flaws or measurement of wall thickness, respectively. The presence of applied stress due to loading or the presence of residual stress due to manufacturing processes or plastic deformation can also lead to changes in magnetic properties, which directly affect the accuracy of inspection results obtained using MFL and magnetic Barkhausen noise.2 

The interplay between microstructure and material stress state, and its effect on the formation of domains in ferromagnetic materials, is important for the application of magnetic materials used for electrical motors4 and transformers,5 inspection of oil and gas pipelines,2 naval structures,6 and aircraft landing gear.7 This interplay depends on length scales over which ferromagnetic domains form and are defined by the minimization of local magnetic energy elements of which formation of demagnetizing fields plays the largest role.1 Domain structure forms and evolves under the applied field according to the underlying microstructure, crystallographic texture, presence of inclusions, dislocations, and material stress. Under the action of an applied magnetic field or stress, domain configurations are modified to accommodate the energy of the domain configuration.1 

Consideration of basic ferromagnetic energy terms at the micromagnetic scale and their extension to larger length scales has been implemented in multiscale modeling to examine coupled magnetoelastic behavior,8,9 including hysteresis effects10 and magnetic Barkhausen noise (MBN).11 While single crystals are considered as the building blocks for larger length scales,8,11 detailed domain structure effects have not been incorporated.

Domain wall motion under tensile stress conditions has recently been investigated using magneto-optical Kerr microscopy (MOKE), which was used to show the correlation between measured domain wall velocity and MBN signal.12 This work12 and others13 have also reported the observation of domain structure refinement and its reorientation in the direction of applied tensile stress, with reverse processes occurring under compressive stress.14,15 Examination of domain wall displacement under stress conditions using MOKE combined with magneto-optical indicator film has been used to examine the effects of cyclic stress and stability of domain wall displacement in grain oriented electrical steel.16 Recent work by Ding et al.17 has also examined domain wall dynamics under applied stress conditions, indicating that skewness of the MBN profile provides a robust feature for applied stress characterization, and have considered results in light of domain wall energy, characteristic relaxation time and distances between pinning edges.

In order to begin to quantitatively accommodate the various properties of ferromagnetic steel that affect the formation and evolution of domain structures under magnetizing conditions, a Tutorial that provides a physics based approach is presented here using the basic concept of the magnetic object (MO), first introduced by Bertotti.18,19 The magnetic object is considered as a separate region of correlated magnetic domain structure, which in the first stage of application, is relatively independent of other neighboring domain structures. The MO is regarded as characteristically existing within single crystallographic iron grains. The MO concept has been further developed by minimizing the contributions to the total energy of the domain structure in order to explain the effects of elastic stress on magnetic Barkhausen noise signals.1,20 In this context, the MO model has been modified from that originally proposed18,19 by first considering a flux-closed MO under zero field conditions,1,20,21 which can form a magnetic moment under the action of applied stress.22 The MO model is ideal in this case as it starts in the low field flux closed state and considers the exchange of magnetoelastic energy, under elastic stress conditions, with surface magnetostatic energy. This exchange results in the formation of magnetic moments within grains, which under stress gradient conditions, combine to generate a net magnetization in steel, the leakage fields of which may be detected by a magnetic field sensor.22 

The basic magnetic object model has been used to provide an explanation for a number of observed magnetic effects in polycrystalline steel. Multiple MOs have been used to explain anisotropy in magnetic Barkhausen noise (MBN) measurements under single and dual easy axis conditions,23 the effects of tensile stress on MBN anisotropy,20 and measurement of tensor components of magnetostriction.24 MOs have helped elucidate the reduction in MBN with increased temper embrittlement, as due to migration of impurity elements and carbon toward grain boundaries where they no longer directly interact with the domain structure within the MO.25 The MO model has also been extended to examine how MBN changes under applied uniaxial elastic tensile stress.25,26 Recently, MOs have been used to explain the production of self magnetic flux leakage signals under stress concentration conditions under zero applied field.22 This Tutorial presents the current state of the MO model along with prospects for its further development and application.

The origin of ferromagnetism in iron and its alloys are the exchange energy ( E e x ) between unpaired 3d electrons on neighboring atoms.27, E e x, arising between neighboring atoms at lattice sites i and j, is given as27,28
(1)
where J i j is the exchange integral for two electrons with spins S i and S j, located at r i and r j, respectively, and ϕ is the angle between the spins. A parallel spin (ϕ = 0) configuration in ferromagnetism implies a positive value for J, which results in a minimum exchange energy as follows from Eq. (1). Since Eq. (1) depends on the distance between spins, | r i r j |, atomic separation also affects the exchange energy. In the case of Fe, the exchange force initially increases with the distance between atoms, or in other words with increasing strain. The variation in exchange energy with atomic separation is described by the estimated Bethe–Slater curve plotted as exchange energy as a function of the ratio of interatomic distance, a, to the radius r of the 3d electron shell in various materials shown in Fig. 1.27 
FIG. 1.

Bethe–Slater curve describing variations in exchange energy for increasing the ratio of interatomic distance, a, to radius r of the 3d electron shell in various materials. Transition across the axis from positive to negative exchange energy results in ferromagnetism.

FIG. 1.

Bethe–Slater curve describing variations in exchange energy for increasing the ratio of interatomic distance, a, to radius r of the 3d electron shell in various materials. Transition across the axis from positive to negative exchange energy results in ferromagnetism.

Close modal

The reduced exchange energy associated with the separation of atoms (i.e., tensile strain) in Fe is counterbalanced by the elastic energy, which increases as a quadratic function of the strain in the crystal until equilibrium is reached at a finite strain.27 The application of stress to the material will affect the state of this energy balance and, therefore, can be used to define stress dependent energy, E σ.28 The relation between the stress and strain components for a cubic crystal is given by the relation between the elastic compliance tensor and the elastic constants for a cubic crystal.29 In the case of iron, E σ along the crystallographic [100] direction is relevant under low and medium field conditions in the elastic stress regime. In the plastic regime, continued application of stress only produces a slight increase in interatomic spacing as planes slip past each other and dislocations are introduced, resulting in only a small change in E σ.

Stress induced changes in magnetization reflect the underlying physical principle of the exchange of mechanical and magnetic energy. The phenomenon is also called inverse magnetostriction, the Villari effect or piezomagnetism.30 Jiles and Atherton,31 Jiles,32 Sablik et al.,33 Sablik,34 and Shi et al.35 developed a model to describe changes in magnetism with applied stress by proposing that stress can be considered equivalent to the application of a magnetic field.22 In addition to these phenomenological models34 and multiscale models,8–11 a more fundamental physics-based approach presented here is centered on the magnetic object (MO), which helps identify the effects on domain structure first within an ideal single Fe crystal incorporating various competing energy sources using energy minimization,18,19 followed by effects that may arise within a polycrystalline matrix, consisting of many such MOs. Expressions for the total energy of the local domain structure in the MO have been developed in order to elucidate the effects of stress on magnetic Barkhausen noise.1,20 In this regard, the MO model has been further developed by proposing a flux-closed magnetic object under zero field conditions.1,20,21 In the absence of an applied field, the MO has been used to show the formation of a magnetic moment under the action of applied stress.22 Under zero applied field conditions a flux-closed configuration, as shown in Fig. 2, minimizes the total energy of a domain configuration. Application of stress has a number of profound effects on domain structure, and in particular, depends on the relative orientation of stress with respect to the domain structure's easy axis (defined as being parallel to the 180° domains). This includes reorientation of the domain structure's easy axis into the direction of applied uniaxial tensile stress and away from the direction of applied compressive stress.20 An increase in uniaxial tensile stress leads to reorientation of the domain structure's easy axis, if not aligned with the tensile stress direction, followed by refinement of the domain structure, i.e., an increase in the number of 180° domain walls, while compressive stress has the opposite effect.20 Under tensile stress conditions, the introduction of a 180° domain walls is limited by the energy cost associated with the addition of 180° domain wall area. The energy cost also depends on grain size with smaller grains permitting fewer 180° domain walls to form with stress.1 Alternatively, the application of tensile stress will introduce energy to the MO that can go into the formation of a magnetic moment.22 The generation of many such moments can result in a bulk magnetization, thereby producing a direct physical connection between stress and induced magnetization.22 

FIG. 2.

Magnetic object with indicated height a, width b, and thickness T. Domain structure example is shown with two 180° domain walls and eight 90° domain walls in zero stress state with flux closure that minimizes total energy. Easy axis is defined by the crystallographic [100] direction along which 180° domains walls are aligned.

FIG. 2.

Magnetic object with indicated height a, width b, and thickness T. Domain structure example is shown with two 180° domain walls and eight 90° domain walls in zero stress state with flux closure that minimizes total energy. Easy axis is defined by the crystallographic [100] direction along which 180° domains walls are aligned.

Close modal

Sections IVVII will examine some particular examples of where the MO formalism can be applied. The effects of stress on ferromagnetic steel domain structure, with reference to the MO model, are examined and conditions of uniaxial and orthogonal biaxial applied stress conditions are considered. The formation of different domain structures due to elastic tensile and compressive stress conditions is put forward as the mechanism for observed asymmetric response between compressive and tensile stress to the magnetization of ferromagnetic steel structures undergoing applied stress conditions as reported by Craik and Wood.36 

In the absence of pinning, domain structures form to minimize the total energy ( E T ) within a ferromagnetic material. In Fe, the crystallographic anisotropy energy ( E K ) ensures that domains lie along one of the three [100] directions near zero field. Within an MO, therefore, E T may be expressed as the sum of exchange ( E e x ), crystallographic anisotropy ( E K ), magneto-elastic (Eσ), magneto-static ( E m s ), domain wall ( E γ ),1 applied field ( E H ),27 and pole ( E p )37 energies written as
(2)
The domain wall energy, E γ, is a consequence of introducing more than one domain to minimize the magnetostatic energy.27,28 E K is constant in the MO model in the low to medium field regime, since magnetization lies along one of the crystallographic [100] easy axis directions in iron. Eliminating the constant terms, the total energy of the idealized geometry of the MO, shown in Fig. 2, within which domains are separated by 180° and 90° domain walls (at 45° to [100] direction27), with energies per unit area of E γ 180 and E γ 90, respectively, becomes
(3)
where each of the remaining terms and their effect on domain structure will be considered separately below.
Domain walls form to minimize the magnetostatic energy given as27,28
(4)
which is proportional to the volume integral of the self inner-product of the magnetic flux density, overall space, by reducing the total magnetization of the MO. The magnetostatic energy of the MO under zero applied field is given as28 
(5)
where N is a form factor that relates the energy outside the object to the energy inside the object, V is the uncompensated domain volume of width d shown in Fig. 3, and M s is the domain saturation magnetization. The progressive introduction of 180° domain walls, which minimizes the magnetostatic energy, reduces the total energy of the domain structure but cannot proceed indefinitely, since the domain walls themselves introduce additional energy. For a single iron crystal, the crystallographic anisotropy energy allows the formation of 90° domain walls, which arise at 45° to the [100] directions.27 Under this condition and with no applied field, zero stress state, and no domain wall pinning, the lowest energy state of the MO is one of complete flux closure. A potential domain configuration shown in Fig. 2, of an equiaxed single iron crystal, contains n = 2, 180° domain walls. This will be taken as an example of the MO model, as it contains the basic elements needed to demonstrate properties of MOs including 90° and 180° domain wall elements. The 180° domain walls are orientated along the [100] direction and define the MO's easy axis of magnetization, while the 90° domain walls provide flux closure in the zero applied field case. The evolution of this domain structure under an applied field and the subsequent modification of additional energy terms, as described by Eq. (3), are considered next.
FIG. 3.

Growth of moment of width d due to an applied external magnetic field (H), showing the initial stages of the evolution of domain structure with increasing magnetization.

FIG. 3.

Growth of moment of width d due to an applied external magnetic field (H), showing the initial stages of the evolution of domain structure with increasing magnetization.

Close modal
The pole energy causes material magnetization when the material is subjected to an applied field,37 an example of which is shown in Fig. 3, and is due to the torque on an atomic dipole moment in the presence of an applied magnetic field.27,28,37 The pole energy for a single dipole within a magnetic field is given by27 
(6)
where m is the magnetic moment, H is the field acting at an angle θ to the moment, as shown in Fig. 3, and ϕ is the direction of the easy axis of the MO. The growth of the moment due to an increase in the external magnetic field is constrained by the crystallographic and exchange energies. To maintain minimum energy, domains most closely aligned with the applied field grow at the expense of neighboring domains by domain wall motion, as shown in Fig. 3. The new domain configuration causes flux lines to escape the boundaries of the MO. Here, the crystallographic anisotropy energy E k is again nearly zero, since the spins align with the crystallographic directions of the grain. However, the magnetostatic energy E m s is not zero since the MO now has a net magnetic moment. The magnetostatic energy E m s is taken to be of the same form as Eq. (5) and is given as
(7)
where N d is a form factor, T is the MO thickness shown in Fig. 2, M s is the saturation magnetization, and d is the width of the moment-producing region of the MO as shown in Fig. 3. The magnitude of the net moment of the MO is μ 0 E m s. The pole energy, which is responsible for magnetizing the MO, is derived from Eq. (6) and is given as
(8)
where ξ p is another constant, a, d , and T are the length, width and thickness, respectively, of the moment producing region of the MO as shown in Fig. 3, and θ is the angle between the applied field H and the MO magnetic moment direction (ϕ=0).
In the case of low stress (i.e., less than 20% of the yield stress), the magnetostatic energy and the pole energy are much larger than the magnetoelastic energy. The optimal width d of the moment-producing region may then be found by taking the derivative of the total energy and setting it to zero,38 
(9)
The form of the spacing is obtained as
(10)

The interaction of the MO with a low to medium applied field motivates the examination of the change in domain structure over the entire range of magnetization.

For increasing applied field, magnetic objects can be used to qualitatively describe the progression of the domain structure up to saturation. Figure 4 shows the flux density as a function of applied magnetic field (H) for the initial magnetization curve and resultant change in domain wall configuration as represented by the MO in Fig. 2. Progression of domain structure with increasing magnetization in iron starting at the demagnetized state (i) where domain structure is flux closed; (ii) movement of primarily 180° domain walls with consequent formation of a dipole moment aligned along the easy axis direction closest to the applied field; (iii) completion of 180° domain wall motion and the end of the most rapid change in magnetization with field; and (iv) completion of 90° domain wall motion; followed by (v) domain vector rotation against the crystallographic easy axis energy into the direction of the applied field.

FIG. 4.

Change in domain structure with applied external magnetic field (H). Evolution of domain structure with increasing magnetization in iron starting in (i) demagnetized state, where domain structure is flux closed, (ii) bulk magnetization mainly by movement of 180° domain walls for domains most closely aligned with the applied field, (iii) end of 180° domain wall motion, (iv) removal of 90° domain volume, and finally, (v) domain vector rotation against the crystallographic easy axis with saturation in the direction of the applied field.

FIG. 4.

Change in domain structure with applied external magnetic field (H). Evolution of domain structure with increasing magnetization in iron starting in (i) demagnetized state, where domain structure is flux closed, (ii) bulk magnetization mainly by movement of 180° domain walls for domains most closely aligned with the applied field, (iii) end of 180° domain wall motion, (iv) removal of 90° domain volume, and finally, (v) domain vector rotation against the crystallographic easy axis with saturation in the direction of the applied field.

Close modal

The movement of domain walls and therefore, growth of moments, is inhibited by the presence of pinning sites. In the MO model pinning is represented as variations in the energy landscape that can arise due to the pinning of domain walls by inclusions, impurities, and dislocations.27 Pinning of domain walls due to magnetostatic coupling across martensite twins has also been identified.39 The inset shown on the right side of Fig. 5 represents the movement of a domain wall, due to an applied field H, over a hypothetical energy landscape, which varies with position x. The transition from the pinning site at position 2 to the next highest pinning energy at position 3 represents a Barkhausen jump1 with a jump distance ℓP, while the domain wall at position 1 remains pinned. Ding et al.17 have considered pinning edge distances and their effect on the MBN signal skewness. Changes in pinning site density via temper embrittlement and the resulting effect of pinning on abrupt changes in domain structure as measured by Barkhausen noise, including the effects of stress, have recently been considered in light of the MO model.25,26,40

FIG. 5.

Motion of a domain wall over an energy landscape containing pinning sites, under action of an applied field, H. At (1) domain wall remains pinned while 2nd domain wall jumps from (2) to (3), resulting in an abrupt change in magnetization.

FIG. 5.

Motion of a domain wall over an energy landscape containing pinning sites, under action of an applied field, H. At (1) domain wall remains pinned while 2nd domain wall jumps from (2) to (3), resulting in an abrupt change in magnetization.

Close modal

In the case, that stress is applied to the MO the energy balance of the domain structure is modified. The interaction of domain structure and stress is considered for various situations below. Here, the change in magnetoelastic energy can be compensated by a reorientation of the domain structure if it is not already aligned within 45° of the principal stress direction.20 This is followed by either a refinement of the domain structure (an increase in the number of 180° domain walls),1,20 or an increase in magnetostatic energy by the formation of a magnetic moment.22 Each of these cases will be considered separately below.

How the energy of a flux closed MO, consisting of n 180° domain walls changes with applied uniaxial stress is considered in this section. In particular, how domain configurations are modified when tensile stress is applied perpendicular to an already existing easy axis. Figure 2 shows such a starting configuration for n = 2 180° domain walls. The application of elastic uniaxial tensile stress, σ, at angle θ between −45° and +45° with respect to the saturation magnetization along a crystallographic easy axis for a single domain of volume (V) is given as28,
(11)
where λ 100 = 2.07 × 10 5 is the saturation strain due to magnetostriction along the [100] direction in Fe. With reference to Fig. 2, since the 180° and 90° domains are oriented at right angles with respect to each other, applying the trigonometric identity sin ( θ ) = cos ( θ + 90 ° )and substituting into Eq. (11), for a multiple domain configuration the expression becomes1,20
(12)
where V180° is the volume of the 180° domains and V90° is the volume of the 90° domains. For a generalized MO domain structure with n 180° domain walls (assuming a ≥ b), i.e., the 180° domain volume is given as1,20
(13)
and the 90° domain volume as
(14)
Substituting (6) and (7) into (5) with identities 2 co s 2 θ = 1 + cos 2 θ and co s 2 θ + si n 2 θ = 1, the increase in magneto-elastic energy under applied tensile stress conditions can be obtained as1,20,21
(15)
Equation (15) has a minimum at θ = 0 ° and a maximum at θ = 90 °. It can be shown10 that for stresses applied at angles, θ′, for 135° < θ′ < 45° with respect to the easy axis, a reorientation of the domain configuration as shown in the transition from Fig. 6(a) and 6(b) for uniaxial tensile stress will occur. The change in energy ΔE from the angle of applied stress relative to the original easy axis direction, θ′ = θ + 90°, to the reoriented domain structure with angle θ may be obtained. Using the identity cos 2 θ = co s 2 θ si n 2 θ, applied to Eq. (12) for θ and θ = θ′ − 90° the difference in energy is given by20 
(16)
FIG. 6.

Reorientation of domain structure from (a) initial state with uniaxial tensile stress at angle 135° > θ′ > 90° with respect to easy axis to (b) with uniaxial tensile stress at angle θ = θ′ − 90°.

FIG. 6.

Reorientation of domain structure from (a) initial state with uniaxial tensile stress at angle 135° > θ′ > 90° with respect to easy axis to (b) with uniaxial tensile stress at angle θ = θ′ − 90°.

Close modal
If the number of 180° domain walls, between the two configurations remains unchanged, substituting in Eqs. (13) and (14) into Eq. (16) gives
(17)
resulting in a transition of the MO to a lower energy state.

In the case of uniaxial compressive stress, for angles −45° < θ < 45° with respect to the easy axis the opposite occurs [Fig. 6(b) would be reoriented to Fig. 6(a)] since the sign in Eq. (17) becomes positive, but compressive stress now provides the negative sign. Reorientation of domain structures due to uniaxial applied stress has been reported in the literature for imaged domains by Craik and Wood36 and others12–16,41,42 and by Barkhausen noise measurements.20,43–46

The effect of uniaxial tensile stress on the refinement of the domain structure, or an increase in the number of 180° domain walls, which demonstrates a grain size dependence, is considered next.

Application of uniaxial tensile stress at angles θ within 45° of the easy axis results in an increase in the 180° domain volume and a decrease in the 90° domain volume with a consequent increase in the number of 180° domain walls.46–48  Figure 7 shows this refinement of the domain structure from (a) an initial configuration with n = 2 180° domain walls to (b) with n = 3 180° domain walls. This result can be obtained from the MO model by considering the magnetoelastic and domain wall energies.1,20

FIG. 7.

Refinement of domain structure from initial state with n = 2 180° domain walls as shown in Fig. 2 to n = 3 180° domain walls under uniaxial tensile stress at angle θ to the easy axis.

FIG. 7.

Refinement of domain structure from initial state with n = 2 180° domain walls as shown in Fig. 2 to n = 3 180° domain walls under uniaxial tensile stress at angle θ to the easy axis.

Close modal
Domain wall energy for a variable number n of 180° domain walls within an MO is given by1,20
(18)
where γ 180 ° = 1.6 × 10−3 J/m2 in iron18 is the energy per unit area and A180° is the surface area of the 180° domain walls within an MO. Domain wall energy for 90° domain walls for cubic domain structures is given by22,
(19)
where γ 90 ° = 0.8 × 10−3 J/m2 is energy per unit area28 and A 90 ° is the surface area of 90° domain walls. Note that the 90° domain wall area is independent of n the number of 180° domain walls. An applied tensile stress increases the exchange energy. However, for typical stresses this will not modify the direction of the easy axis that is along one of the three crystallographic easy axes.27,32 In steel, the stress induced energy only equals the crystallographic energy at 4550 MPa,27 which is far beyond the yield strength of even the highest strength steels considered here. However, the 180° domain wall energy is modified by the applied stress and can be obtained by considering the difference in stress induced energy on either side of the domain wall as given by Eq. (11).27 As a function of angle θ relative to the easy axis this is obtained as20,
(20)
where δ is the 180° domain wall thickness, which is somewhat greater than two 90° domain walls or 80 nm.27 The stress dependence of the energy of the 90° domain wall may be considered negligible.27 In effect the refinement of the domain structure occurs, since it is energetically favorable to reduce the volume of 90° domains perpendicular to the applied tensile stress direction.
The MO energies that change with the addition of a 180° domain wall are magneto-elastic, Eσ, as given by Eq. (15), the domain wall energies E γ 180 ° and E γ 90 ° as described by Eqs. (18) and (19), respectively, and the increase of the 180° domain wall energy with tensile stress, Eq. (20). The total energy of the MO under uniaxial tensile stress, within ±45° of the easy axis, becomes
(21)
Change in number n, of 180° domain walls with stress is in integer steps. Therefore, it is reasonable to consider an n to n+1 transition. The associated change in energy is obtained by subtracting Eq. (21) from itself with n substituted by n+1, which is obtained as1,20
(22)
The threshold for the addition of an additional 180° domain wall arises when Δ E T = 0. Applying this to Eq. (22) allows a definition for the threshold stress at which a 180° domain wall can be added to the MO under tensile stress to be expressed as10 
(23)

The threshold stress for the addition of a 180° domain wall increases with the number of 180° domain walls present and also depends on the size and relative dimensions of the MO. In general, the threshold stress decreases with increasing MO size. The increase in the number of 180° domain walls with tensile stress is limited since the addition of each domain wall also costs energy.1,20 Note that under compressive stress conditions, the number of 180° domain walls that can be removed is finite, i.e., it is required that n ≥ 0, hence an asymmetry between tensile and compressive stress conditions with respect to the formation of domain structures arises. Observations of the refinement of domain structures are given by Dijkstra and Martius,47 Craik and Wood,36 Qiu et al.,43 Liu et al.,12 Perevertov et al.,13–15 and Zeng et al.16 Barkhausen measurements on pipeline steel also support the reorientation and refinement of domain structure with applied tensile stress in steel for the typical stresses and grain sizes observed there.20 

The grain size of cold-rolled and recrystallized steel is primarily in the range of 5–20 μm,49–51 while mild steel grains can reach 50 μm.22 A range of 5 to 50 μm is plotted in Fig. 8 for equiaxed grains (a − b = T). The equations also accommodate non-equiaxed grains that may arise under particular processing conditions.21,40 Note that for small grain size high-strength steels the extent of domain refinement is limited, and the stress energy must be balanced by other mechanisms to be considered below. Calculated threshold stresses are consistent with those observed to have an effect on MBN response in smaller grained pipeline steel.20 

FIG. 8.

Threshold stress for addition of 180° domain walls for various potential pipeline steel grain sizes, including 5 (black square), 10 (gray circle), 15 (brown triangle), 20 (red square), 25 (green circle), 30 (black circle), 40 (blue square), and 50 μm (green inverted triangle) equiaxed sizes. Solid curves are a best fit to the data of the threshold stress equation.

FIG. 8.

Threshold stress for addition of 180° domain walls for various potential pipeline steel grain sizes, including 5 (black square), 10 (gray circle), 15 (brown triangle), 20 (red square), 25 (green circle), 30 (black circle), 40 (blue square), and 50 μm (green inverted triangle) equiaxed sizes. Solid curves are a best fit to the data of the threshold stress equation.

Close modal
After reorientation has occurred and further domain refinement is limited by large threshold stresses in small grain materials, the balance of MO energy under stress conditions can be accomplished by the motion of domain walls that increase the MO's magnetostatic energy. In the MO model, this can result in the formation of a magnetic moment of width d as shown in Fig. 9. The increase in magnetostatic energy in the small moment limit d << a (the field produced by the free pole of one surface makes an insignificant contribution to the field produced by the pole on the other surface), is given by28 
(24)
where M S = 2.15 A / m is the saturation magnetization of the domain. Accompanying the increase in the moment width, d, is a reduction in the horizontal dimension of the flux-closed portion in Fig. 9 from width b to width (b–d). This modifies Eqs. (13)–(15) by substitution of b with (b–d). The moment of width d that forms is also subject to a change in energy by the applied stress as per Eq. (11), with Vol. = adT. Combined with Eq. (15), the total energy of the MO, as expressed by Eq. (3) for the configuration in Fig. 9 under the applied stress condition without an applied field becomes22 
(25)
where the increase in 180° domain wall energy due to stress, Eq. (20), has been considered negligible. The formation of the moment in Fig. 9 results in a decrease of the 90° domain volume and 90° domain wall area, and a small increase in the 180° domain wall area. Collecting like powers of d in Eq. (16) and using the trigonometric identity 2 cos 2θ = 1 + cos 2θ, terms containing the product a × d cancel, yielding a form of Eq. (25) as
(26)
FIG. 9.

Tensile stress (σ) applied at angle θ with respect to the easy axis shown in Fig. 2 results in an increase in 180° volume at the expense of 90° domain volume and thereby, increases saturation magnetization volume, M S, in the form of a moment of width d, aligned along easy axis closest to the direction of tensile stress.

FIG. 9.

Tensile stress (σ) applied at angle θ with respect to the easy axis shown in Fig. 2 results in an increase in 180° volume at the expense of 90° domain volume and thereby, increases saturation magnetization volume, M S, in the form of a moment of width d, aligned along easy axis closest to the direction of tensile stress.

Close modal
The derivative of Eq. (17) with respect to d, can be used to obtain the equilibrium value of moment width d that can be used to minimize E T as22 
(27)
Setting the derivative in Eq. (18) equal to zero and solving for d, with the constraint that θ is within 45°of the easy axis, gives22,
(28)
Since the net contribution from the 180° and 90° domain wall energy is negligible [see Eqs. (18) and (19) above], and 4.32 × 10 5 M s 2 3 λ 100 σ (since M S = 2.15 A / m and λ 100 = 2.07 × 10 5 for iron28) a good approximation for the moment width d can be given as22 
(29)
where the constant C = 3.11 × 10−5 MPa−1. Note that the moment width d is proportional to the applied elastic tensile stress and is also a function of the grain size as indicated by the width b, the domain structure, as characterized by the number n of 180° domain walls, and the angle θ of the stress with respect to the easy axis. If compressive stress is present, the moment is anticipated to form along the easy axis that is closest to the perpendicular with respect to the stress direction.

For grains in the range of sizes in steels, between 5 μm and 50 μm, moment width d as a function of stress for a cubic grain can be calculated using Eq. (29). At 250 MPa, the 5 μm size MO has two 180° domain walls (n = 2), while the 25 μm size has 5 (see Fig. 8). Equation (29) predicts an increasing moment width d with increasing elastic stress and MO (grain) size but is inversely proportional to the number n of 180° domain walls. For the 5 μm cubic grain with n = 2, stressed at 250 MPa and θ = 0° the moment width d is calculated as 0.02 μm, while for the 25 μm cubic grain with n = 5 the moment width is 0.04 μm.

The volume of the moment, V = adT, may be multiplied by the saturation magnetization, M S, giving the moment in the ith MO within a structure as22 
(30)
Substituting the approximate form of Eq. (29) into Eq. (30) an expression for the ith moment's dependence on stress within an MO is obtained as22 
(31)
Representing polycrystalline steel by using multiple MOs, the result will be a net magnetization in the direction of applied elastic tensile stress. This in turn will generate an effective internal field that will be prone to flux leakage depending on bulk stress, and local stress concentration due to defects or structure geometry, but one that is still susceptible to overall demagnetizing field effects. The total internal magnetization will be the vector sum of the moments over the volume of the sample, which may be expressed as
(32)
where T i is the thickness, b i is the width, a i is the height, θ i is the angle of the applied stress direction relative to the ith MO's easy axis, ni is the number of 180° domain walls in the MO and σ i is the elastic vector stress component at the ith MO. Note that the total moment m is now a vector whose direction is primarily determined by that of the existing tensile stress σ. The volume of the ith MO is given by V i = T i b i a i. For a uniaxial tensile or compressive stress, without stress concentrations, the stress is in a constant vector direction. Assuming an average distribution for θ i and average number n of 180° domain walls an expression for the total stress induced moment can be written as
(33)
where κ is a constant. For an average distribution of grain orientations, cos 2 θ i averaged between −45° and +45° is 2/π, and assuming an average of two 180° domain walls per MO, κ = 1/π. Note that the total moment m is a function of the volume of the sample under stress, which is the sum over all the magnetic objects. The sampled volume is therefore
(34)
Equation (33) can therefore, be rewritten as the local moment per unit volume or magnetization, M, given as
(35)

Equation (35) states that stress in the elastic regime induces a local magnetization per unit volume, therefore, acting as an effective field. Note, however, that the actual direction of magnetization is doubly degenerate, meaning that magnetization may form either along the [100] direction closest to the uniaxial tensile stress direction or 180° away from it. While this can facilitate magnetization processes in the presence of a magnetic field, even the Earth's field, the energy associated with the formation of a demagnetizing field in the presence of this degeneracy will inhibit macroscopic changes in magnetization under uniform stress conditions. Pinning of domain structures will also play a role in the generation of potential leakage fields, which can also be affected by the application of stress.52–54 

In the presence of a local stress gradient, Eq. (35) becomes
(36)
where is the gradient operator. In this case, a local stress gradient is predicted to generate a local leakage field.

The question arises whether domain refinement or formation of magnetic moments will occur under applied tensile elastic stress conditions. It was noted above that the addition of 180° domain walls occurs as an integer step, with threshold stresses for the addition of one wall being higher and occurring less often for small grain materials than larger grained ones. Therefore, the smaller grained materials are more likely to exhibit stress induced magnetization than refinement. In addition, with the example, comparing the 5 and 25 μm cubic grain sizes above, 125 5 μm cubic grains fit into the volume of one 25 μm cubic grain. Therefore, the effect of stress induced magnetization of small grain materials is again expected to be greater.

As presented in Sec. VI B, stress can generate magnetization and therefore, acts as an effective field. It would be useful, therefore, to make an estimate of the effective magnitude of this field for comparison with other possible external fields. In this section, the equivalent field is estimated for comparison with the Earth's field. Figure 3 shows the idealization of a magnetic field H applied at an angle θ with respect to the easy axis of an MO. The field applies a pressure P (force per unit area) on the 180° domain wall that is given by28 
(37)
and for the 90° domain wall, where θ″ is measured with respect to the [110] direction as18 
(38)
where θ″ = 45°θ. The work done to move a single 180° domain wall a distance d, and therefore, the increase in energy within the MO, is given by
(39)
and for the 90° domain wall as
(40)
where n≥1 and the small change in length of the 180° and 90° domain walls has been neglected, as well as the potential presence of any pinning. The total work done by the field is the sum of the work done on both 180° and 90° domain walls by the applied field
(41)
A comparison of Fig. 9 and Fig. 3 again emphasizes the equivalence of applied stress and field at the level of an MO. The equivalent field can be obtained by equating Eq. (24), the magnetostatic energy generated by the stress, with Eq. (41) giving the equivalent field due to stress as
(42)

Again, equiaxed grain sizes of 5 and 25 μm, may be compared using the moment width calculated in Sec. VI B, of 0.02 and 0.04 μm, respectively, under a uniaxial tensile stress of 250 MPa applied at θ = 0°. Using Eq. (42), this produces an equivalent energy change for the 5 μm MO of 9.4 × 10−16 J, requiring an equivalent field of H = 450 A/m, which is B = μ0H = 0.57 mT. The larger 25 μm cubic grain has the energy of 1.9 × 10−14 J, but only produces an equivalent field of 181 A/m, which is 40% of that obtained for the 5 μm size grain. The effective field of 0.57 mT for the 5 μm size grain is 11 times larger than the nominal 0.05 mT of Earth's field.

Detection of stress induced leakage fields from lengths of steel pipe has been reported in the literature41 and was first reported by Atherton et al.52,53 The application of stress to large structures will cause domain walls to become unpinned resulting in a change in magnetization due to stress cycling, nominally toward a lower energy state as pinning barriers are overcome. These observations may be explained by an examination of residual magnetic field under application of cyclic stress, which has shown repeatable domain wall displacements and that these are stable under cyclic stress and during relaxation time after release of stress.16 In contrast, the residual states of domain rotation are inconsistent,16 therefore suggesting that this is the mechanism that takes place during the first application of stress.

Under many structural steel conditions, orthogonal biaxial stresses may arise. Domain refinement under orthogonal biaxial stress conditions55 is considered here. With the addition of a second orthogonal stress, σ 2, in Eq. (12), with reference to Fig. 10, an additional angular dependent term is introduced at 90° to the first as
(43)
where σ 1, the primary stress, is orthogonal to σ2 and the identity co s 2 θ = si n 2 ( 90 ° + θ ) was used. Considering the particular case where σ 1 > σ 2, domain refinement may arise as well as additional magnetostatic energy. Considering refinement first, substituting Eqs. (13) and (14) into Eq. (43) gives
(44)
which can be simplified using the trigonometric identity 2 co s 2 θ = cos 2 θ + 1 as
(45)
FIG. 10.

Refinement of domain structure to n = 3 from n = 2, shown in Fig. 2, under orthogonal biaxial applied tensile stress with σ1 > σ2.

FIG. 10.

Refinement of domain structure to n = 3 from n = 2, shown in Fig. 2, under orthogonal biaxial applied tensile stress with σ1 > σ2.

Close modal
Combining the energy terms for the MO with reference to Fig. 10, including domain wall energies Eqs. (18) and (19), and Eq. (20) under orthogonal biaxial stress the total energy becomes
(46)
For an n to n+1 transition, substituting n+1 into (38) and subtracting, this becomes
(47)
which turns into Eq. (22) if σ 2 = 0. The addition of the orthogonal stress σ 2 of the same sign (tensile in this case) reduces the effect of σ 1 for an n to n+1 transition under the condition that δ << b, (δ is 80 nm, the thickness of a 180° domain wall.27 This is shown by the expression for the threshold stress obtained by setting Δ E T = 0 and σ 1 = σ T in Eq. (47) as
(48)
which reduces to Eq. (23) if σ 2 = 0. Note that the threshold stress σ T required to introduce a 180° domain wall is counteracted directly by the orthogonal stress σ 2 and indirectly by the increase in 180° domain wall energy, where the additional wall energy is given by the product δ σ 2 in the numerator on the right-hand side. For the case where the stresses are of opposite sign, for example, a compressive axial stress combined with an orthogonal tensile stress, the effective stress is enhanced, as might arise in a pressurized pipeline subject to a bending stress. When σ 1 = σ 2, Eq. (47) becomes
(49)
canceling the effect of stress σ 1, except for an increase in 180° domain wall energy, and no dependence on n, and therefore, no increase in the number of 180° domain walls. From Eq. (48), the application of an orthogonal tensile stress σ 2 to an already existing tensile stress condition σ 1 tends to reduce the effect of the original stress on the resulting domain structure, and thereby, increases the threshold stress to refine the existing domain structure. A compressive stress σ 2 would have the opposite effect, reducing the threshold stress, leading to greater refinement. This effect on domain structure under biaxial stress conditions is supported by magnetic Barkhausen noise measurements as reported by Vengrinovich et al.56 and biaxial stress effects in magnetic flux leakage measurements as reported by Crouch et al.57 
Under orthogonal biaxial stress conditions, the formation of a moment of width d is considered next. Combined with Eq. (45), the total energy of the MO as expressed by Eq. (3) for the configuration in Fig. 11 under applied stress condition with a b to (b–d) substitution gives
(50)
where the increase in 180° domain wall energy due to stress given by Eq. (20) has again been considered negligible. Equation (50) can be simplified as
(51)
which reduces to Eq. (25) when σ 2 = 0. Again, taking the derivative of E T with respect to d,
(52)
and setting ∂ET/∂d = 0, the width of the moment d that minimizes the energy becomes
(53)
Again, as for Eq. (28), since the net contribution from the 180° and 90° domain wall energy is negligible [see Eqs. (18) and (19)], and 4.32 × 10 5 M s 2 3 λ 100 σ (since MS = 2.15 A/m and λ 100 = 2.07 × 10−5 for iron28) a good approximation for the moment width for d can be given as
(54)
where the constant C = 3.11 × 10−5 MPa−1 and d is now proportional to the difference in orthogonal stresses. Equation (54) returns to Eq. (29) if σ 2 = 0. If σ 1 = σ 2, d becomes zero and no moment is formed. The local moment per unit volume or magnetization, M, can be obtained in the same manner as that used to obtain Eq. (35) as
(55)
FIG. 11.

Formation of a moment of width d, aligned along the easy axis closest to the direction of greatest tensile stress (σ1) applied at angle θ and stress (σ2) at angle θ +90° for the orthogonal biaxial stress case. Formation of moment requires σ1 > σ2, but where application of σ2 reduces growth of the moment.

FIG. 11.

Formation of a moment of width d, aligned along the easy axis closest to the direction of greatest tensile stress (σ1) applied at angle θ and stress (σ2) at angle θ +90° for the orthogonal biaxial stress case. Formation of moment requires σ1 > σ2, but where application of σ2 reduces growth of the moment.

Close modal

Therefore, the presence of an orthogonal stress, of the same sign as the primary stress, will reduce the effects of the primary stress on the domain structure, while if an orthogonal stress of opposite sign is introduced, the effects will be further enhanced.

The particular case of orthogonal stresses can be generalized to any biaxial stress case with the appropriate treatment of the equations presented here. The resulting expressions are, however, anticipated to be more complex than the simpler orthogonal case.

The application of MOs has been used to explain a wide variety of magnetic phenomena.1,12,20–22,24–26 However, as indicated here, the development is not complete. Future prospects for the MO from primarily a single object to the consideration of multiple objects and their interactions may result in the development of a more physics based hysteresis model that incorporates the effects of pinning, grain size, microstructure, texture, and elastic and plastic stress. Future work could involve the examination of the MO model in light of multiscale modeling.8–11 Interpretation of magneto-optical Kerr microscopy (MOKE) results under stress and variable grain size may further validate the MO model presented here.

This Tutorial compiles existing work on the magnetic object model in the presence of uniaxial or orthogonal stress conditions. The magnetic object model is used to show that stress in the elastic regime can modify the domain structure as well as act as a source of effective magnetization in polycrystalline steels. Application of elastic stress modifies the existing domain structure and forms magnetic moments within grains which can generate a net magnetization within steel that, as a consequence, may exhibit leakage flux under stress gradient conditions or produce a demagnetization of steel under cyclic stress conditions. The level of local magnetization is proportional to the elastic uniaxial tensile stress and is estimated to be ten times that of the Earth's field. When orthogonal stresses of the same type arise (both tensile or both compressive), elastic stress induced magnetization is reduced or, in the case of equal stresses, will be canceled. Model predictions are consistent with experimental results that investigate changes in domain structure with stress and with magnetization measurements.

This work was supported by the Natural Sciences and Engineering Research Council of Canada (No. RGPIN-2017-06365).

The authors have no conflicts to disclose.

The data that supports the findings of this study are available within the article.

1.
T. W.
Krause
and
A. A.
Samimi
, “
Micromagnetic techniques
,” in
ASM Handbook
, edited by
A.
Ahmad
and
L. J.
Bond
(
ASM International
,
2018
), Vol. 17, pp.
515
530
.
2.
T. W.
Krause
,
K.
Mandal
,
C.
Hauge
,
P.
Weyman
,
B.
Sijgers
, and
D. L.
Atherton
, “
Correlation between magnetic flux leakage and magnetic Barkhausen noise: Stress dependence in pipeline steel
,”
J. Magn. Magn. Mater.
169
,
207
219
(
1997
).
3.
D. P. R.
Desjardins
,
T. W.
Krause
, and
L.
Clapham
, “
Transient eddy current method for the characterization of magnetic permeability and conductivity
,”
Nondestruct. Test. Eval. Int.
80
,
65
70
(
2016
).
4.
A. A.
Samimi
,
T. W.
Krause
,
L.
Clapham
,
M.
Gallaugher
,
Y.
Ding
, and
R.
Chromik
, “
Manufacturing inspection of electrical steels using magnetic Barkhausen noise: Residual stress detection
,”
CINDE J.
37
,
26
30
(
2016
).
5.
T. W.
Krause
,
J. A.
Szpunar
,
M.
Birsan
, and
D. L.
Atherton
, “
Correlation of magnetic Barkhausen noise with core loss in oriented 3% Si-Fe steel laminates
,”
J. Appl. Phys.
79
(
5
),
3156
3167
(
1996
).
6.
A. A.
Samimi
,
T. W.
Krause
, and
L.
Clapham
, “
Stress-Response of magnetic Barkhausen noise in submarine hull steel: A comparative study
,”
J. Nondestruct. Eval.
35
(
32
),
7
(
2016
).
7.
R.
Morita
,
S.
Kobayashi
,
A. G.
Odeshi
,
J. A.
Szpunar
,
K.
Miura
, and
Y.
Kamada
, “
Effects of high-strain-rate deformation on magnetic hysteresis in high-tensile steels
,”
AIP Adv.
6
,
055903
(
2016
).
8.
N.
Buirona
,
S.
Heb
,
L.
Hirsinger
,
S.
Depeyre
,
M.
Bernadou
, and
R.
Billardon
, “
From micromagnetic to multiscale modeling of the coupled magnetoelastic behavior of ferromagnetic materials
,”
Physica B
306
,
33
37
(
2001
).
9.
L.
Daniel
,
O.
Hubert
,
N.
Buiron
, and
R.
Billardon
, “
Reversible magneto-elastic behavior: A multiscale approach
,”
J. Mech. Phys. Solids
56
,
1018
1042
(
2008
).
10.
L.
Daniel
,
M.
Rekik
, and
O.
Hubert
, “
A multiscale model for magneto-elastic behaviour including hysteresis effects
,”
Arch. Appl. Mech.
84
,
1307
1323
(
2014
).
11.
P.
Fagan
,
B.
Ducharne
,
L.
Daniel
, and
A.
Skarlatos
, “
Multiscale modelling of the magnetic Barkhausen noise energy cycles
,”
J. Magn. Magn. Mater.
517
,
167395
(
2021
).
12.
J.
Liu
,
G. Y.
Tian
,
B.
Gao
,
K.
Zeng
,
Y.
Zheng
, and
J.
Chen
, “
Micro-macro characteristics between domain wall motion and magnetic Barkhausen noise under tensile stress
,”
J. Magn. Magn. Mater.
493, 165719 (2020).
13.
O.
Perevertov
,
J.
Thielsch
, and
R.
Schäfer
, “
Influence of applied tensile stress on the hysteresis curves and magnetic domain structure of grain-oriented transverse Fe–3%Si steel
,”
J. Magn. Magn. Mater.
385
,
358
367
(
2015
).
14.
O.
Perevertov
and
R.
Schäfer
, “
Magnetic properties and magnetic domain structure of grain-oriented Fe-3% Si steel under compression
,”
Mater. Res. Express
3
(
9
),
096103
(
2016
).
15.
O.
Perevertov
and
R.
Schäfer
, “
Influence of applied compressive stress on the hysteresis curves and magnetic domain structure of grain-oriented transverse Fe–3%Si steel
,”
J. Phys. D: Appl. Phys.
45
,
135001
(
2012
);
J. Magn. Magn. Mater.
493,
165719
(
2020
).
16.
K.
Zeng
,
G.
Tian
,
J.
Liu
,
B.
Gao
, and
F.
Qiu
, “
Repeatability and stability study of residual magnetic field for domain wall characterization
,”
J. Magn. Magn. Mater.
485
,
391
400
(
2019
).
17.
S.
Ding
,
G. Y.
Tian
,
G.
Dobmann
, and
P.
Wang
, “
Analysis of domain wall dynamics based on applied stress measurements on Q235 steel using magnetic Barkhausen noise
,”
J. Magn. Magn. Mater.
421
,
225
229
(
2017
).
18.
G.
Bertotti
, “
Physical interpretation of eddy current losses in ferromagnetic materials. I. Theoretical considerations
,”
J. Appl. Phys.
57
(
6
),
2110
2117
(
1985
).
19.
G.
Bertotti
,
Hysteresis in Magnetism for Physicists, Materials Scientists and Engineers
(
Academic Press
,
San Diego
,
1998
).
20.
T. W.
Krause
,
L.
Clapham
,
A.
Pattantyus
, and
D. L.
Atherton
, “
Investigation of the stress-dependent magnetic easy axis in steel using magnetic Barkhausen noise
,”
J. Appl. Phys.
79
(
8
),
4242
4252
(
1996
).
21.
L.
Clapham
,
C.
Heald
,
T.
Krause
,
D. L.
Atherton
, and
P.
Clark
, “
Origin of a magnetic easy axis in pipeline steel
,”
J. Appl. Phys
86
(
3
),
1574
1580
(
1999
).
22.
M.
Kashefi
,
L.
Clapham
,
T. W.
Krause
,
P. R.
Underhill
, and
A. K.
Krause
, “
Stress induced self magnetic flux leakage at stress concentration zone
,”
IEEE Trans. Magn.
57
(
10
),
1
8
(
2021
).
23.
T. W.
Krause
,
K.
Mandal
, and
D. L.
Atherton
, “
Modelling of magnetic Barkhausen noise in single and dual easy axis systems in steel
,”
J. Magn. Magn. Mater.
195
,
193
205
(
1999
).
24.
Y.
Yu
,
L.
Clapham
,
D. L.
Atherton
, and
T. W.
Krause
, “
Tensor magnetostriction loops for a steel cube
,”
J. Magn. Magn. Mater.
171
,
94
102
(
1997
).
25.
T. W.
Krause
,
A.
Saleem
, and
P. R.
Underhill
, “
Towards a physics based model of magnetic Barkhausen noise in steel
,”
Adv. Mater. Lett.
11
(
6
),
20061526
(
2020
).
26.
M.
Kashefi
,
T. W.
Krause
,
P. R.
Underhill
,
A.
Saleem
, and
S. P.
Farrell
, “
Decoupling the effect of stress and microstructure on MBN response in cast Q1 N steel
,”
Mater. Sci. Technol.
37
,
1225
1235
(
2021
).
27.
B. D.
Cullity
,
Introduction to Magnetic Materials
(
Addison-Wesley
,
Reading, MA
,
1972
).
28.
S.
Chikazumi
and
S. H.
Charap
,
Physics of Magnetism
(
Krieger
,
1964
).
29.
G. E.
Dieter
,
Mechanical Metallurgy
, 2nd ed. (
McGraw Hill, Inc.
,
New York
,
1976
), p.
61
.
30.
S.
Bao
,
M.
Fu
,
Y.
Gu
, and
S.
Hu
, “
Evolution of the piezomagnetic field of ferromagnetic steels subjected to cyclic tensile stress with variable amplitudes
,”
Exp. Mech.
56
,
1017
1028
(
2016
).
31.
D. C.
Jiles
and
D. L.
Atherton
, “
Theory of ferromagnetic hysteresis
,”
J. Magn. Magn. Mater.
61
(
1–2
),
48
60
(
1986
).
32.
D. C.
Jiles
, “
Theory of the magnetomechanical effect
,”
J. Phys. D: Appl. Phys.
28
(
8
),
1537
1546
(
1995
). .
33.
M. J.
Sablik
,
S. W.
Rubin
,
L. A.
Riley
,
D. C.
Jiles
,
D. A.
Kaminski
, and
S. B.
Biner
, “
A model for hysteretic magnetic properties under the application of noncoaxial stress and field
,”
J. Appl. Phys.
74
(
1
),
480
488
(
1993
).
34.
M. J.
Sablik
, “
A model for asymmetry in magnetic property behavior under tensile and compressive stress in steel
,”
IEEE Trans. Magnet.
33
(
5
),
3958
3960
(
1997
).
35.
P.
Shi
,
K.
Jin
, and
X.
Zheng
, “
A general nonlinear magnetomechanical model for ferromagnetic materials under a constant weak magnetic field
,”
J. Appl. Phys.
119
,
145103
(
2016
).
36.
D. J.
Craik
and
M. J.
Wood
, “
Magnetization changes induced by stress in a constant applied field
,”
J. Phys. D.: Appl. Phys.
3
,
1009
(
1970
).
37.
A.
Hubert
and
R.
Schäfer
,
Magnetic Domains—The Analysis of Magnetic Microstructures
(
Springer-Verlag
,
Berlin
,
2000
).
38.
A. K.
Krause
, “
Nonlinear ferromagnetic effects on magnetic Barkhausen noise measurements of steels
,”
M.Sc. thesis
(
Queen’s University
,
2021
).
39.
J. N.
Armstrong
and
M. R.
Sullivan
, “
Role of magnetostatic interactions in micromagnetic structure of multiferroics
,”
J. Appl. Phys.
103
,
023905
(
2008
).
40.
A.
Saleem
,
P. R.
Underhill
,
S. P.
Farrell
, and
T. W.
Krause
, “
Magnetic Barkhausen noise measurements to assess temper embrittlement in HY-80 steels
,”
IEEE Trans. Magnet.
56
(
3
),
1
8
(
2020
).
41.
W. D.
Corner
and
J. J.
Mason
, “
The effect of stress on the domain structure of cube-textured silicon iron
,”
Proc. Phys. Soc.
81
,
925
(
1963
).
42.
R.
Carey
, “
‘The origin of the band structure domain pattern in cube-textured silicon-iron
,”
Proc. Phys. Society
80
,
934
(
1962
).
43.
F.
Qiu
,
M.
Jovičević-Klug
,
G.
Tian
,
G.
Wu
, and
J.
McCord
, “
Correlation of magnetic field and stress-induced magnetic domain reorientation with Barkhausen noise
,”
J. Magn. Magn. Mater.
523
,
167588
(
2021
).
44.
T. W.
Krause
,
J. M.
Makar
, and
D. L.
Atherton
, “
Investigation of the magnetic field and stress dependence of 180° domain wall motion in pipeline steel using magnetic Barkhausen noise
,”
J. Magn. Magn. Mater.
137
,
25
34
(
1994
).
45.
T. W.
Krause
,
A.
Pattantyus
, and
D. L.
Atherton
, “
Investigation of strain dependent magnetic Barkhausen noise in steel
,”
IEEE Trans. Magn.
31
,
3376
3378
(
1995
).
46.
T. W.
Krause
,
N.
Pulfer
,
P.
Weyman
, and
D. L.
Atherton
, “
Magnetic Barkhausen noise: Stress-dependent mechanisms in steel
,”
IEEE Trans. Magn.
32
,
4764
4766
(
1996
).
47.
L. J.
Dijkstra
and
U. M.
Martius
, “
Domain pattern in silicon-iron under stress
,”
Rev. Mod. Phys.
25
(
1
),
146
150
(
1953
).
48.
J. W.
Shilling
and
G. L.
Houze
, Jr.
, “
Magnetic properties and domain structure in grain-oriented 3% Si-Fe
,”
IEEE Trans. Magn. MAG
10
,
195
223
(
1974
).
49.
K. S.
Ghosh
and
D. K.
Mondal
, “
Mechanical and electrochemical behavior of a high strength low alloy steel of different grain sizes
,”
Metall. Mater. Trans. A
44
,
3608
3622
(
2013
).
50.
L.
Smart
and
L. J.
Bond
, “
Material property relationships for pipeline steels and the potential for application of NDE
,”
AIP Conf. Proc.
1706
, 160003 (
2016
).
51.
M. A.
Islam
and
Z. N.
Farhat
, “
Mechanical and electrochemical synergism of API X42 pipeline steel during erosion–corrosion
,”
J. Bio. Tribo. Corros.
1, 26 (
2015
).
52.
S. G. H.
Staples
,
C.
Vo
,
D. M. J.
Cowell
,
S.
Freear
, and
C.
Ives
, “
Solving the inverse problem of magnetisation–stress resolution
,”
J. Appl. Phys.
113
,
133905
(
2013
).
53.
D. L.
Atherton
and
A.
Teitsma
, “
Detection of anomalous stresses in gas pipelines by magnetometer survey
,”
J. Appl. Phys.
53
,
8130
(
1982
).
54.
D. L.
Atherton
and
D. C.
Jiles
, “
Effect of stress on magnetization
,”
NDT Int.
19
(
1
),
15
(
1986
).
55.
T. W.
Krause
,
A. A.
Samimi
,
A. K.
Krause
,
P. R.
Underhill
, and
L.
Clapham
, in “
Modelling the effect of stress on magnetic Barkhausen noise
,” in
ICBM12—12th International Conference on Barkhausen Noise and Micromagnetic Testing, IKTS
, Dresden, Germany,
September 25
26
(Stresstech,
2017
).
56.
V.
Vengrinovich
,
D.
Vintov
,
A.
Prudnikov
,
P.
Podugolnikov
, and
V.
Ryabtsev
, “
Magnetic Barkhausen effect in steel under biaxial strain/stress: Influence on stress measurement
,”
J. Nondestruct. Eval.
38
,
52
(
2019
).
57.
A. E.
Crouch
,
R. E.
Beissner
,
G. L.
Burkhardt
,
E. A.
Creek
, and
T. S.
Grant
, Magnetic flux leakage inspection of gas pipelines: The effects of biaxial stress, Topical Report, April 1993–September 1995, Southwest Research Institute.
Published under a nonexclusive license by AIP Publishing.