Neumann formula

The name Carl Neumann is a thread that weaves through disparate fields of physics and mathematics, connecting the tangible interaction of electric circuits to the abstract machinery of quantum mechanics. While his contributions are varied, they are united by deep, underlying principles of reciprocity, iteration, and interconnection. This article aims to illuminate these connections by exploring several key concepts bearing his name, demonstrating how a single mathematical framework can solve a wide array of problems. In the following chapters, we will first delve into the "Principles and Mechanisms," where we will derive the elegant Neumann formula for mutual inductance, understand the iterative power of the Neumann series, and explore his work on the special functions that form the language of physics. Subsequently, under "Applications and Interdisciplinary Connections," we will see these tools in action, from designing electrical components and analyzing knotted wires to performing complex computations and forming the basis of perturbation theory.

Imagine you are standing in a hall of mirrors. Your reflection appears in the mirror in front of you. But it also appears in the mirror behind you, which reflects the reflection in the first mirror, and so on, creating an infinite cascade of images. This idea of an effect causing an effect, which in turn causes another, is not just a visual trick. It is a deep principle that echoes through many corners of physics and mathematics. As we explore the work of the great 19th-century mathematician Carl Neumann, we will find this idea of interconnectedness and iteration appearing in surprisingly different contexts, from the invisible dance of electric currents to the very structure of the functions that describe our universe.

Let's begin with something tangible: two simple loops of wire, separated by some distance in space. If you drive an electric current through the first loop, a magnetic field springs into existence around it. Some of this magnetic field will inevitably pass through the area of the second loop. We call this "captured" field the ​​magnetic flux​​. The remarkable thing is that a changing flux will induce a current in the second loop, as if by magic. This is the principle of electromagnetic induction, the engine behind electric generators and transformers.

The ​​mutual inductance​​, denoted by the letter $M$, is a measure of this coupling. It tells us how much magnetic flux is captured by loop 2 for a given amount of current in loop 1. A big $M$ means the loops are "talking" to each other very effectively. So how do we calculate it?

One way is to calculate the entire magnetic field $\mathbf{B}$ produced by the first loop everywhere in space, and then integrate the part of it that passes through the surface $S_2$ of the second loop: $\Phi_2 = \int_{S_2} \mathbf{B}_1 \cdot d\mathbf{S}$. This is often a monstrous task. There must be a more elegant way.

And there is. Physics often hides its beauty one layer deeper than we first look. Instead of the magnetic field $\mathbf{B}$, let’s consider its parent, the ​​magnetic vector potential​​ $\mathbf{A}$, a field from which $\mathbf{B}$ is born via the relation $\mathbf{B} = \nabla \times \mathbf{A}$. A beautiful theorem by George Stokes tells us that the total flux of the curl of a vector field through a surface is equal to the circulation of that field around the boundary of the surface. For us, this means the complicated surface integral for flux can be replaced by a line integral around the wire loop itself!

$\Phi_2 = \int_{S_2} (\nabla \times \mathbf{A}_1) \cdot d\mathbf{S} = \oint_{C_2} \mathbf{A}_1 \cdot d\mathbf{l}_2$

This is a huge simplification. We no longer need to know the field everywhere, only on the path of the second wire. Now, we use the fact that the vector potential at a point $\mathbf{r}_2$ due to a current $I_1$ in a loop $C_1$ is given by an integral along that first loop:

$\mathbf{A}_1(\mathbf{r}_2) = \frac{\mu_0 I_1}{4\pi} \oint_{C_1} \frac{d\mathbf{l}_1}{|\mathbf{r}_2 - \mathbf{r}_1|}$

Putting these two pieces together and dividing by the current $I_1$ gives the mutual inductance. What emerges is one of the most elegant formulas in electromagnetism, the ​​Neumann formula​​ for mutual inductance:

$M = \frac{\mu_0}{4\pi} \oint_{C_1} \oint_{C_2} \frac{d\mathbf{l}_1 \cdot d\mathbf{l}_2}{|\mathbf{r}_1 - \mathbf{r}_2|}$

Look at this formula. It is perfectly symmetrical. It treats loop 1 and loop 2 in exactly the same way. This means that the inductance of loop 2 on loop 1 is identical to the inductance of loop 1 on loop 2 ($M_{12} = M_{21}$). This is a profound statement of ​​reciprocity​​. The way the wires "talk" to each other is a two-way street, a mutual conversation. This formula holds the key to calculating the interaction between any two circuits, no matter how complex their shapes—from simple circles to tangled messes like a trefoil knot. The geometry contains the physics.

Neumann's genius was not confined to electromagnetism. He gave us a powerful way of thinking about problems that refer to themselves. Consider the growth of a population. The number of births today, $B(t)$, depends on the number of individuals born in the past who are now old enough to have offspring. So, $B(t)$ depends on earlier values of $B$. How can you solve an equation where the answer appears on both sides?

Neumann's approach is beautifully intuitive. Let's frame the problem, as seen in population dynamics, like this:

$\text{Total Effect} = \text{Initial Cause} + \text{Effect of the Total Effect}$

In the language of integral equations, this might look like $B(t) = B_0(t) + \int K(t,s) B(s) ds$, where $B_0(t)$ is the "initial cause" (e.g., births from an initial population) and the integral represents the "effect of the total effect" (births from all subsequent generations).

Trying to solve this directly is like trying to see your own eyes without a mirror. So, we iterate. We start with an approximation: the total effect is just the initial cause. $B \approx B_0$. This is a rough first guess. Now, let's find the effect of this first guess and add it on. $B \approx B_0 + \text{Effect}(B_0)$. This is better. Now we take the effect of this new, better approximation and add that on as well. This is like the hall of mirrors. Each step adds a new reflection, a new echo. The total solution is the sum of all the echoes.

This leads to the ​​Neumann series​​. If our equation is of the form $f = g + \mathcal{K}f$, where $\mathcal{K}$ is an integral operator, the solution is:

$f = g + \mathcal{K}g + \mathcal{K}(\mathcal{K}g) + \mathcal{K}(\mathcal{K}(\mathcal{K}g)) + \dots = \sum_{n=0}^{\infty} \mathcal{K}^n g$

Each term in the series is a successive layer of cause and effect. In the population example, $g$ is the first generation, $\mathcal{K}g$ is the second generation (children of the first), $\mathcal{K}^2g$ is the third generation (grandchildren), and so on. This powerful method turns an impenetrable, self-referential equation into an infinite sequence of straightforward calculations. It is a fundamental tool used everywhere, from calculating particle scattering in quantum mechanics to rendering realistic lighting in computer graphics.

Finally, we find Neumann's name attached to the very functions that form the vocabulary of physics. When we solve problems involving spheres or cylinders, the simple sines and cosines of introductory physics are not enough. We need a richer language: the ​​special functions​​, like the Bessel functions and Legendre polynomials.

Consider the vibrations of a circular drumhead. Its modes of vibration are not described by simple sine waves, but by ​​Bessel functions​​, $J_n(z)$. These functions have many wonderful properties, one of which is the ​​Neumann addition formula​​:

$J_k(u+v) = \sum_{n=-\infty}^{\infty} J_n(u) J_{k-n}(v)$

This formula is the Bessel function's version of the angle-addition identity for sine, $\sin(u+v) = \sin(u)\cos(v) + \cos(u)\sin(v)$. It tells us how to "shift" the argument of the function. It's a rule of grammar in this hidden language. With this rule, seemingly impossible sums can collapse into simple expressions. For example, a nasty-looking sum like $\sum_{n=-\infty}^{\infty} (-1)^n J_n(x) J_{3-n}(x)$ can be shown, using a variation of this formula, to be equal to $-J_3(0)$, which is simply zero. The complexity vanishes, thanks to the underlying structure revealed by the identity.

Similarly, in problems with spherical symmetry, like the electric field around a charged sphere or the quantum mechanics of the hydrogen atom, we encounter ​​Legendre polynomials​​, $P_n(x)$. For every differential equation of a certain type, there are two independent families of solutions. The Legendre polynomials are the well-behaved family. Their less-behaved siblings are the ​​Legendre functions of the second kind​​, $Q_n(z)$. How are they related? Once again, a ​​Neumann formula​​ provides the bridge:

$Q_n(z) = \frac{1}{2} \int_{-1}^{1} \frac{P_n(t)}{z-t} dt$

This integral formula constructs the second, more mysterious solution ($Q_n$) from the first, well-known one ($P_n$). It is a generating machine. By providing this link, Neumann gave us a powerful tool to understand the complete set of solutions and to uncover the rich web of relationships that exist between them.

From the palpable force between two wires, to the abstract process of iterative problem-solving, to the grammar of the special functions that describe waves and fields, we see the signature of a unified mathematical framework. The recurrence of Neumann's name is no coincidence. It points to the existence of deep, recurring principles in nature—reciprocity, iteration, and interconnection—and the enduring power of mathematics to reveal them.

We have spent some time getting to know a rather beautiful piece of mathematical physics: the Neumann formula for mutual inductance. We have seen where it comes from and the principles that govern it. But to truly appreciate a tool, you must see it in action. It is one thing to have a finely crafted hammer; it is another to build a house with it. So, our task now is to explore the "houses" that Neumann's formula—and other powerful ideas bearing his name—have helped to build. We will see that this is not just a formula for electrical engineers, but a gateway to understanding geometry, computation, and even the abstract machinery of modern physics.

Let us begin in the most familiar territory: the world of circuits, currents, and magnetism. The most straightforward application, the kind you find in textbooks, is calculating the magnetic handshake between two simple, coaxial loops of wire. Using Neumann's formula, we can derive an exact, if somewhat formidable, expression for their mutual inductance. It turns out to involve those elegant and mysterious functions known as elliptic integrals. This is no mere mathematical curiosity; this is the very heart of how transformers, induction motors, and wireless charging systems work. The ability to precisely calculate the coupling between coils is the foundation of much of our electrical world.

But the real world is rarely as neat as two perfect, coaxial circles. What about the intricate pathways on a printed circuit board, or the complex arrangement of wires in an electric motor? The Neumann formula retains its power. We can adapt it from elegant, closed loops to the more practical case of finite, straight segments of wire [@problemid:588487]. By integrating along the length of two wires, we can determine their magnetic cross-talk, a critical factor in designing high-speed electronics where unwanted interference can be a major problem.

This is where the physicist's art of approximation often proves more insightful than a brute-force calculation. What if the two coils are very far apart? You might imagine that the intricate details of their shape become less important, and you would be right. By applying the Neumann formula to two wire arcs separated by a large distance, we can use an expansion to find a much simpler, leading-order result. The complex integral melts away, revealing a simple truth: the mutual inductance weakens with distance, proportional to the square of the wires' radius and inversely to the separation. This is a classic example of how physicists find simple, powerful laws hiding within complex equations.

Now, for a truly mind-bending application. What if you take a single wire and, instead of a simple loop, tie it in a knot? Does a trefoil knot have a different self-inductance than a simple circle of the same length? You bet it does! The generalized Neumann formula for self-inductance allows us to explore this beautiful intersection of electromagnetism and topology. The inductance becomes a measure of the wire's "knottedness"—how many times its own magnetic field links through itself. This isn't just a party trick; the study of knotted magnetic fields is crucial in plasma physics, where fusion researchers try to contain searingly hot, knotted ropes of plasma. It even appears in biology, where the long strands of DNA are often tangled and supercoiled, and their geometric configuration affects their biological function.

As the geometry of our wires becomes more complex—like our trefoil knot or the components in a real-world device—the double integral in Neumann's formula becomes hopelessly difficult to solve with pen and paper. This is where we turn to our powerful partner: the computer. But telling a computer to "solve an integral" is an art in itself.

One of the most ingenious methods is to play a game of chance. The Monte Carlo method approaches the integral by "randomly sampling" pairs of points on the two loops and averaging their contributions. It is as if you are estimating the average depth of a lake by throwing in a thousand stones at random locations and measuring where they land. What is remarkable is that with enough "throws," this method can give a surprisingly accurate answer for even the most convoluted shapes. Furthermore, physicists and engineers have developed clever tricks, like using "control variates" (an approximate, solvable version of the problem), to guide the random sampling and get a good answer much faster.

Another, more direct approach is to chop the loops into a large number of tiny, straight segments and sum up the interactions between them—a method known as numerical quadrature. This is like approximating the area of a curve by drawing a lot of little rectangles under it. A beautiful trick in this game is called Richardson extrapolation. By calculating the inductance twice, once with a coarse grid ($N$ segments) and once with a finer grid ($2N$ segments), you can combine the two answers to cleverly cancel out the leading source of error. It is a mathematical sleight of hand that squeezes a much more accurate result out of your initial, imperfect calculations. These computational methods form a vital bridge, allowing us to apply the elegant physics of Neumann's formula to the messy, complex reality of modern engineering design.

The name Carl Neumann is attached to such a wealth of powerful ideas that to stop at inductance would be to see only one facet of a brilliant diamond. His work echoes through the halls of pure mathematics, and these "other" Neumann formulas and series often find their way back into physics, providing essential tools for entirely different problems.

Consider the "special functions" of mathematical physics—functions like sines and cosines, but tailored for more complex symmetries. For problems with spherical symmetry, like calculating the gravitational field of a planet or the electric field around an atom, we use Legendre functions. It turns out there is a ​​Neumann's formula for Legendre functions​​, which gives the value of a definite integral of their product. This identity is a crucial piece of the mathematical toolkit for solving differential equations in spherical coordinates.

If we move to problems with cylindrical symmetry—the vibration of a drumhead, the flow of water in a pipe, or the propagation of light in an optical fiber—we encounter Bessel functions. And here, too, we find a ​​Neumann's Integral Representation​​. This remarkable identity expresses the product of two Bessel functions as an integral of another Bessel function. It is a fundamental tool for analyzing waves and vibrations in cylindrical systems.

Perhaps the most abstract and far-reaching of these ideas is the ​​Neumann series​​. It has nothing to do with inductance or integrals, but everything to do with one of the most fundamental operations in mathematics: inversion. Suppose you have a transformation, or "operator," that is very close to doing nothing at all (the identity operator). The Neumann series gives you a recipe for finding the inverse of that transformation by writing it as an infinite sum.

At first glance, this might seem hopelessly abstract. But this idea is the mathematical engine behind one of the most powerful techniques in all of physics: ​​perturbation theory​​. Imagine you want to calculate the energy levels of an atom sitting in a weak electric field. Solving this problem exactly is impossible. But we can solve the problem of the atom in isolation. The electric field is a small "perturbation." We can write the operator describing the full system as (Ideal System + Small Perturbation). The Neumann series (or a conceptual cousin) gives us a systematic way to calculate the corrections to the energy levels, term by term, as a power series in the strength of the perturbing field. From calculating the behavior of molecules to predicting the interactions of subatomic particles in quantum field theory, this method of "approaching the truth through successive corrections" is absolutely central to modern physics, and the Neumann series is its mathematical foundation.

From the tangible magnetism between two wires to the abstract machinery of quantum mechanics, the ideas of Neumann provide a stunning tour of the deep and beautiful unity of physics and mathematics. They are not just isolated formulas to be memorized, but powerful ways of thinking that, once understood, unlock a deeper view of the world's intricate workings.