Principles of the Magnetic Methods in Geophysics

Chapter 1 Magnetic Field in a Nonmagnetic Medium

Alex A. Kaufman, Richard O. Hansen, Robert L.K. Kleinberg

Abstract

Publisher Summary

This chapter focuses on the magnetic field in a nonmagnetic medium. Numerous experiments performed at the beginning of the19th century demonstrated that constant currents interact with each other that mean mechanical forces act at every element of the circuit. Certainly, this is one of the amazing phenomena of the nature and would have been very difficult to predict. In fact, it is almost impossible to expect that the motion of electrons inside of wire may cause a force on moving charges somewhere else, for instance, in another wire with current, and for this reason the phenomenon of this interaction was discovered by chance. By analogy with the attraction field caused by masses, it is proper to assume that constant (time-invariant) currents create a field and because of the existence of this field and of the existence of this field, other current elements experience the action of the force F. Such a field is called the magnetic field, and it can be introduced from Ampere’s law.

1.1 Interaction of constant currents and Ampere's law

Numerous experiments performed at the beginning of the19th century demonstrated that constant currents interact with each other; that is mechanical forces act at every element of the circuit. Certainly, this is one of the amazing phenomena of the nature and would have been very difficult to predict. In fact, it is almost impossible to expect that the motion of electrons inside of wire may cause a force on moving charges somewhere else, for instance, in another wire with current, and for this reason the phenomenon of this interaction was discovered by chance. It turns out that this force of interaction between currents in two circuits depends on the magnitude of these currents, the direction of charge movement, the shape and dimensions of circuits, as well as the their mutual position with respect to each other. The list of factors clearly shows that the mathematical formulation of the interaction of currents should be much more complicated task than that for masses or electric charges. In spite of this fact, Ampere was able to find a relatively simple expression for the force of the interaction of so-called elementary currents:

     (1.1)

where I1 and I2 are magnitudes of the currents in the linear elements dl(p) and dl(q), respectively, and their direction coincides with that of the current density; Lqp the distance between these elements and is directed from the point q to the point p, which can be located at the center of these elements; finally μ0 is a constant equal to

and is often called the magnetic permeability of free space. Certainly, this is confusing definition, since free space does not have any magnetic properties. We will use the S.I. system of units where the distance is measured in meters and force in newtons. Of course, with a change of the system of units the value of μ0 varies too. In applying Ampere's law (Equation (1.1)), it is essential that the separation between current elements must be much greater than their length; that is

Correspondingly, points: p and q can be located anywhere inside their elements. It is easy to see some similarity of Ampere's law and Newton's law of attraction; they describe a force between either elementary currents or elementary masses. Let us illustrate Equation (1.1) by three examples shown in Fig. 1.1. Suppose that elements dl(p) and dl(q) are in parallel with each other. Then, as follows from definition of the cross product, the force dF(p) is directed toward the element dl(q), and two current elements attract each other (Fig. 1.1(a)). If two current elements have opposite directions, the force dF(p) tries to increase the distance between elements, and therefore they repeal each other (Fig. 1.1(b)). If the elements dl(p) and dl(q) are perpendicular to each other, as is shown in Fig. 1.1(c), then in accordance with Equation (1.1) the magnitude of the force acting at the element dl(p) equals

Fig. 1.1 (a) Parallel current elements. (b) Anti-parallel current elements. (c) Current elements perpendicular to each other. (d) Interaction of closed current circuits.

and it is parallel to the element dl(q). At the same time, the force dF(q) at the point q is equal to zero, that is Newton's third law becomes invalid. This contradiction results from the fact that Equation (1.1) describes an interaction between current elements instead of closed current circuits. In other words, this equation is written for unrealistic case, since we cannot create a constant current in an open circuit, but as all experiments show, Equation (1.1) gives the correct result for closed current lines. For instance, applying the principle of superposition, the force of interaction between two arbitrary and closed currents (Fig. 1.1(d)) is defined as

     (1.2)

The internal integral in Equation (1.2) characterizes the force acting at some point of the current line L1, for instance, point p1 and caused by all elements of the current line L2. Thus, the force F represents a sum (integral) of forces applied at different points of the same circuit and, as is well known, its action causes in general a deformation, translation and a rotation of the current line L1. It is obvious that in the case of closed circuits the interaction between them obeys Newton's third law.

The relationship between the force F and currents (Equation (1.2)) is called Ampere's law for closed circuits with constant currents, and it is impossible to overestimate its importance, since it is the theoretical foundation of many devices measuring magnetic field as well as electromotors, transforming electric energy into mechanical energy, and it has numerous applications in physics and technology. Finally, it is proper to notice the following:

(a) The force of interaction is independent of properties of the medium which surrounds the currents.

(b) The Ampere's law was formulated for currents which are independent of time. It turns out that this law allows us to calculate the force of interaction even in the case of alternating currents as long as displacement currents can be neglected.

(c) It is natural to be surprised and impressed that Ampere found Equation (1.1) since in reality he had only experimental data describing interaction for closed current circuits.

1.2 Magnetic field of constant currents

By analogy with the attraction field caused by masses, it is proper to assume that constant (time-invariant) currents create a field, and due to the existence of this field other current elements experience the action of the force F. Such a field is called the magnetic field, and it can be introduced from Ampere's law. In fact, we can write Equation (1.1) as

     (1.3)

Here

     (1.4)

Equation (1.4) establishes the relationship between the elementary current and the magnetic field caused by this element, and it is called Biot–Savart law. In accordance with Equation (1.4), the magnitude of the magnetic field dB is

     (1.5)

where (Lqp, dl) is the angle between the vectors Lqp and dl, and the vector dB is perpendicular to these vectors as is shown in Fig. 1.2(a). The unit vector, b0, characterizing the direction of the field, is defined by

Fig. 1.2 (a) Illustration of Equation (1.4). (b) Field due to the surface currents.

We may say that the magnetic field exists at any point regardless of presence or absence of a current at this point. In S.I. units, the magnetic field is measured in teslas and it is related to other units such as gauss and gamma in the following way:

1.2.1 General form of Biot–Savart law

Now we generalize Equation (1.4) assuming that along with linear currents there are also volume and surface currents. First let us represent the product I dl as

     (1.6)

since the vector of the current density j and the vector dl have the same direction and the elementary volume is equal to dV=dS dl. Thus, in place of Equation (1.4), we can write

     (1.7)

and this expression describes the magnetic field due to an elementary volume with the current density j(q).

If the current is concentrated in a relatively thin layer with thickness dh, which is small with respect to the distance to the observation points, it is often convenient to replace this layer by a current sheet. As is seen from Fig. 1.2(b), the product I dl can be modified in the following way:

     (1.8)

Here dS is the surface element, and

     (1.9)

is the surface density of currents. Correspondingly, for the magnetic field caused by the elementary surface current, we have

     (1.10)

Now applying the principle of superposition for all three types of currents and making use of Equations (1.4), (1.7) and (1.10), we obtain the general form of the Biot–Savart law:

     (1.11)

In order to understand better this relationship between the magnetic field and currents (Biot–Savart law), it is appropriate to add the following:

1. Equation (1.11) allows us to calculate the magnetic field everywhere except points with linear and surface currents.

2. Unlike volume distribution of currents, linear and surface analogies are only mathematical models of real distribution of currents, which are usually introduced to simplify calculations of the field and study its behavior. For this reason, the equation

     (1.12)

in essence comprises all possible cases of the current distribution and can be always used to determine the field B.

3. In accordance with Biot–Savart law, the current is the sole generator of the constant magnetic field and the distribution of this generator is characterized by the magnitude and direction of the current density vector. As is well known, the vector lines of j(q) are always closed. This means that a time-invariant magnetic field is caused by generators of the vortex type and correspondingly we are dealing with a vortex field, unlike, for example, the gravitational field.

4. As was pointed out earlier, all the experiments that allowed Ampere to derive Equation (1.1) were carried out with closed circuits. At the same time, Equation (1.1), as well as Equations (1.4) and (1.7), is written for the element dl, where a constant current cannot exist if this element does not constitute a part of a closed circuit. In other words, Equations (1.1) and (1.4) cannot be proved by experiment, but interaction between closed circuits takes place as if the magnetic field B, caused by the current element I dl, is described by Equation (1.4). Let us illustrate this ambiguity in the following way. Suppose that the magnetic field dB due to the elementary current I dl is

where ϕ is an arbitrary continuous function. Then, the magnetic field caused by the constant current in the closed circuit is

As is well known from vector analysis, the circulation of a gradient is equal to zero and therefore

Thus, the ambiguity in the expression of the magnetic field due to an elementary current vanishes, when the interaction or the magnetic field of closed circuits is considered, and the magnetic field B is uniquely defined by Equations (1.11) and (1.12).

5. In accordance with Equation (1.11), the magnetic field caused by a given distribution of currents depends only on the coordinates of the observation point p; that is, it is independent of the presence of other currents. In this light, it is important to emphasize that the right-hand side of Equation (1.11) does not contain any terms that characterize physical properties of the medium where these currents are located. Therefore, the field B at the point p, generated by the given distribution of currents, remains the same if free space is replaced by a nonuniform medium. For instance, if the given current circuit is placed inside of a magnetic material like iron (Chapter 2), the field B caused by this current is the same as if it were in free space. Of course, as is well known and it will be discussed later, the presence of such medium results in a change of the total magnetic field B, but this means that inside of the magnetic medium, as well on its surface, along with the given current there are other currents which also produce a magnetic field. This conclusion directly follows from Equation (1.11), which states that any change of the magnetic field B can occur only due to a change of the current distribution.

6. It is convenient to distinguish two types of currents, namely: conduction and magnetization currents (Chapter 2):

     (1.13)

where jc and jm are vectors of the current density which characterize the distribution of the conduction and magnetization currents. Thus, in place of Equation (1.12), derived for conduction currents, we can write

     (1.14)

This is important generalization of Biot–Savart law, which establishes the relationship between the magnetic field and currents in any medium. Later we will take into account the influence of currents in a magnetic medium but for now it is assumed that such medium is absent and only conduction currents are considered.

7. From Ampere's and Biot–Savart laws, we have for the force with which the magnetic field acts on the elementary current j dV:

     (1.15)

As is well known from Coulomb's law, the force of the electric field acting on elementary charge with the density δ(p) is equal to

     (1.16)

From comparison of Equations (1.15) and (1.16), we can conclude that there is analogy between vectors B and E. In fact, these two vectors determine the force acting on the corresponding generator of the field. In this sense, the vector B, describing the magnetic field, is similar to the vector E, which characterizes the electric field. There is another common feature of these fields, namely, each of them is caused by generators of one type only which have an obvious meaning: either charges or currents.

8. Equation (1.14) allows us to determine the magnetic field provided that the distribution of currents is known. In other words, using Biot–Savart law we can solve the forward problem. At the same time, if part of the currents is not given, Equation (1.14) becomes useless and we have to solve a boundary-value problem.

9. Earlier we emphasized that in general any constant magnetic field is caused by a combination of conduction and magnetization currents. The first one represents a motion of free charges, while magnetization current is a physical concept which allows one to take into account motion of charges within atoms. In this chapter, we focus on the field generated by the conduction currents only, but later investigate the influence of magnetization currents.

10. Biot–Savart law can be applied for calculating time-varying magnetic fields as soon as an influence of displacement currents is negligible.

1.3 The vector potential of the magnetic field

Although calculation of the magnetic field, making use of the Biot–Savart law, is not a very complicated procedure, it is still reasonable to find a simpler way of determining the field. With this purpose in mind, by analogy with the scalar potential of the gravitational and electric fields, we introduce a new function which is more simply related to the currents than the magnetic field. Moreover, there is another reason to consider this function, namely, it allows us to derive a system of equations for the field B and simplifies the formulation of boundary-value problem, when currents can be known only if the magnetic field is already determined. Certainly, in such cases the Biot–Savart law cannot be applied and it is very useful to introduce this function. As we know, the magnetic field caused by conduction currents with density j is

     (1.17)

can be represented as

     (1.18)

are gradients when either point q or p changes, respectively.

Its substitution into Equation (1.17) gives

     (1.19)

since the relative position of vectors forming the cross product is changed. Now we will make use of the equality

     (1.20)

which follows from the vector identity

Applying Equation (1.20), we can rewrite Equation (1.19) as

     (1.21)

The current density j is a function of the point q and does not depend on the location of the observation point p. Therefore, the integrand of the second integral is zero