← Part II: Magnetostatics & Induction
Jackson Chapter 5

Magnetostatics

The Biot-Savart law, Ampere's law, magnetic vector potential, and the magnetostatic boundary value problem at the graduate level.

1.1 The Biot-Savart Law

The Biot-Savart law is the magnetostatic analog of Coulomb's law. For a steady current density$\mathbf{J}(\mathbf{r}')$, the magnetic field at point $\mathbf{r}$ is:

$$\boxed{\mathbf{B}(\mathbf{r}) = \frac{\mu_0}{4\pi}\int \frac{\mathbf{J}(\mathbf{r}')\times(\mathbf{r}-\mathbf{r}')}{|\mathbf{r}-\mathbf{r}'|^3}\,d^3r'}$$

For a filamentary wire carrying current $I$, replace $\mathbf{J}\,d^3r' \to I\,d\boldsymbol{\ell}'$.

The Biot-Savart law automatically satisfies $\nabla\cdot\mathbf{B}=0$ (no magnetic monopoles), since the integrand can be written as the curl of a vector function. This is the fundamental property distinguishing magnetostatics from electrostatics.

Magnetic Field of a Current Loop (Dipole Pattern)ImB ∝ 1/r³(far field)
Figure. Magnetic field lines (cyan) around a current loop (gold). The pattern is that of a magnetic dipole, with field lines emerging from the north and looping back to the south.

Derivation 1: B-field of a Circular Current Loop on Axis

We compute the on-axis field of a circular loop of radius $R$ carrying current $I$ using the Biot-Savart law.

Step 1: Set up geometry

Place the loop in the $xy$-plane centered at the origin. A current element at angle $\phi'$ is $d\boldsymbol{\ell}' = R\,d\phi'(-\sin\phi'\,\hat{x}+\cos\phi'\,\hat{y})$. The field point is on the axis: $\mathbf{r} = z\hat{z}$.

Step 2: Compute the separation vector

The vector from source to field point is $\mathbf{r}-\mathbf{r}' = -R\cos\phi'\,\hat{x}-R\sin\phi'\,\hat{y}+z\hat{z}$ with magnitude $|\mathbf{r}-\mathbf{r}'| = \sqrt{R^2+z^2}$.

Step 3: Evaluate the cross product

Computing $d\boldsymbol{\ell}'\times(\mathbf{r}-\mathbf{r}')$, the $x$ and $y$ components integrate to zero by symmetry. The $z$-component gives $R^2 d\phi'$.

Step 4: Integrate over the loop

$$B_z = \frac{\mu_0 I}{4\pi}\int_0^{2\pi}\frac{R^2\,d\phi'}{(R^2+z^2)^{3/2}} = \frac{\mu_0 I R^2}{2(R^2+z^2)^{3/2}}$$

Step 5: Limiting cases

At the center ($z=0$): $B_z = \mu_0 I/(2R)$. Far from the loop ($z \gg R$): $B_z \approx \mu_0 m/(2\pi z^3)$ where $m = I\pi R^2$ is the magnetic dipole moment, confirming the dipole approximation.

1.2 Ampere's Law

Ampere's law in integral form relates the circulation of the magnetic field to the enclosed current:

$$\boxed{\oint \mathbf{B}\cdot d\boldsymbol{\ell} = \mu_0 I_{\text{enc}}}$$

In differential form: $\nabla\times\mathbf{B} = \mu_0\mathbf{J}$. This is valid only for steady currents ($\nabla\cdot\mathbf{J}=0$).

Ampere's law is most useful when the current distribution possesses sufficient symmetry (cylindrical, planar, or toroidal) to determine $\mathbf{B}$ directly. Otherwise, one must resort to the Biot-Savart law or the vector potential.

Derivation 2: Ampere's Law from the Biot-Savart Law

We derive the differential form of Ampere's law by taking the curl of the Biot-Savart integral.

Step 1: Write B in terms of a vector identity

The Biot-Savart integrand involves $\mathbf{J}\times\hat{\mathscr{r}}/\mathscr{r}^2$. Using $\hat{\mathscr{r}}/\mathscr{r}^2 = -\nabla(1/\mathscr{r})$:

$$\mathbf{B} = \frac{\mu_0}{4\pi}\int \mathbf{J}(\mathbf{r}')\times\nabla\!\left(\frac{1}{|\mathbf{r}-\mathbf{r}'|}\right)d^3r' = \frac{\mu_0}{4\pi}\nabla\times\int\frac{\mathbf{J}(\mathbf{r}')}{|\mathbf{r}-\mathbf{r}'|}\,d^3r'$$

Step 2: Take the curl of B

Apply $\nabla\times(\nabla\times\mathbf{A}) = \nabla(\nabla\cdot\mathbf{A})-\nabla^2\mathbf{A}$ to the integral expression. The Laplacian of $1/|\mathbf{r}-\mathbf{r}'|$ gives $-4\pi\delta^3(\mathbf{r}-\mathbf{r}')$.

Step 3: The divergence term vanishes for steady currents

The $\nabla(\nabla\cdot\mathbf{A})$ term involves $\nabla'\cdot\mathbf{J}$, which vanishes by the continuity equation for steady currents. Therefore:

$$\nabla\times\mathbf{B} = \frac{\mu_0}{4\pi}\int \mathbf{J}(\mathbf{r}')\,4\pi\delta^3(\mathbf{r}-\mathbf{r}')\,d^3r'$$

Step 4: Evaluate using the delta function

$$\boxed{\nabla\times\mathbf{B} = \mu_0\mathbf{J}}$$

This is Ampere's law in differential form.

Step 5: Integral form via Stokes' theorem

Integrating over a surface $\mathcal{S}$ bounded by loop $\mathcal{C}$ and applying Stokes' theorem:

$$\oint_\mathcal{C}\mathbf{B}\cdot d\boldsymbol{\ell} = \int_\mathcal{S}(\nabla\times\mathbf{B})\cdot d\mathbf{a} = \mu_0\int_\mathcal{S}\mathbf{J}\cdot d\mathbf{a} = \mu_0 I_{\text{enc}}$$

1.3 The Magnetic Vector Potential

Since $\nabla\cdot\mathbf{B}=0$, we can always write $\mathbf{B}=\nabla\times\mathbf{A}$ for some vector potential$\mathbf{A}$. The vector potential is not unique: the gauge transformation $\mathbf{A}\to\mathbf{A}+\nabla\Lambda$ leaves $\mathbf{B}$ unchanged.

$$\boxed{\mathbf{A}(\mathbf{r}) = \frac{\mu_0}{4\pi}\int \frac{\mathbf{J}(\mathbf{r}')}{|\mathbf{r}-\mathbf{r}'|}\,d^3r'}$$

In the Coulomb gauge ($\nabla\cdot\mathbf{A}=0$), $\mathbf{A}$ satisfies $\nabla^2\mathbf{A} = -\mu_0\mathbf{J}$, a vector Poisson equation.

Derivation 3: Poisson Equation for the Vector Potential

Starting from Ampere's law and the definition of $\mathbf{A}$, we derive the governing equation for the vector potential.

Step 1: Substitute B = curl A into Ampere's law

$$\nabla\times(\nabla\times\mathbf{A}) = \mu_0\mathbf{J}$$

Step 2: Apply the vector identity

$$\nabla(\nabla\cdot\mathbf{A}) - \nabla^2\mathbf{A} = \mu_0\mathbf{J}$$

Step 3: Choose the Coulomb gauge

Set $\nabla\cdot\mathbf{A} = 0$. This is always achievable: if $\nabla\cdot\mathbf{A}' \neq 0$, perform a gauge transformation with $\nabla^2\Lambda = -\nabla\cdot\mathbf{A}'$.

Step 4: Obtain the vector Poisson equation

$$\boxed{\nabla^2\mathbf{A} = -\mu_0\mathbf{J}}$$

Each Cartesian component satisfies a scalar Poisson equation.

Step 5: Solution by analogy with electrostatics

By analogy with $\nabla^2\Phi = -\rho/\epsilon_0$ having solution $\Phi = (1/4\pi\epsilon_0)\int\rho/|\mathbf{r}-\mathbf{r}'|\,d^3r'$, we immediately write the solution above.

1.4 Magnetic Multipole Expansion

Expanding $1/|\mathbf{r}-\mathbf{r}'|$ for $r \gg r'$ in the vector potential gives the magnetic multipole expansion. The magnetic monopole term vanishes identically ($\nabla\cdot\mathbf{B}=0$), leaving the dipole as the leading term:

$$\mathbf{A}_{\text{dip}}(\mathbf{r}) = \frac{\mu_0}{4\pi}\frac{\mathbf{m}\times\hat{r}}{r^2}, \qquad \mathbf{m} = \frac{1}{2}\int \mathbf{r}'\times\mathbf{J}(\mathbf{r}')\,d^3r'$$

Derivation 4: Magnetic Dipole Moment of a Current Loop

We show that a planar current loop of area $a$ has magnetic moment $\mathbf{m} = I\mathbf{a}$.

Step 1: Start from the general definition

For a filamentary loop, $\mathbf{m} = \frac{1}{2}\oint \mathbf{r}'\times(I\,d\boldsymbol{\ell}') = \frac{I}{2}\oint\mathbf{r}'\times d\boldsymbol{\ell}'$.

Step 2: Apply the vector identity for loop integrals

The integral $\frac{1}{2}\oint\mathbf{r}'\times d\boldsymbol{\ell}'$ equals the vector area $\mathbf{a}$ of the loop (by Stokes' theorem applied to the identity function).

Step 3: For a planar circular loop

With $\mathbf{r}' = R(\cos\phi'\hat{x}+\sin\phi'\hat{y})$ and $d\boldsymbol{\ell}' = R(-\sin\phi'\hat{x}+\cos\phi'\hat{y})d\phi'$:

$$\mathbf{r}'\times d\boldsymbol{\ell}' = R^2 d\phi'\,\hat{z}$$

Step 4: Integrate

$$\mathbf{m} = \frac{I}{2}\int_0^{2\pi}R^2\,d\phi'\,\hat{z} = I\pi R^2\hat{z} = IA\hat{z}$$

Step 5: General result

$$\boxed{\mathbf{m} = NIA\hat{n}}$$

For $N$ turns, the direction $\hat{n}$ follows the right-hand rule. This result holds for any planar loop shape, not just circular.

1.5 Magnetostatic Boundary Conditions

At an interface carrying surface current density $\mathbf{K}$, the magnetic field satisfies:

$$B_{\perp}^{\text{above}} - B_{\perp}^{\text{below}} = 0, \qquad \mathbf{B}_{\parallel}^{\text{above}} - \mathbf{B}_{\parallel}^{\text{below}} = \mu_0(\mathbf{K}\times\hat{n})$$

The normal component of $\mathbf{B}$ is continuous (from $\nabla\cdot\mathbf{B}=0$), while the tangential component has a discontinuity proportional to the surface current (from Ampere's law applied to a thin pillbox/loop).

Derivation 5: Boundary Conditions from Maxwell's Equations

We derive both boundary conditions using pillbox and Amperian loop arguments.

Step 1: Normal component from Gauss's law for B

Apply $\nabla\cdot\mathbf{B}=0$ to a thin pillbox of area $A$ and vanishing thickness straddling the interface:

$$\oint\mathbf{B}\cdot d\mathbf{a} = B_\perp^{\text{above}}A - B_\perp^{\text{below}}A = 0$$

Step 2: Normal component is continuous

$$B_\perp^{\text{above}} = B_\perp^{\text{below}}$$

Step 3: Tangential component from Ampere's law

Apply $\oint\mathbf{B}\cdot d\boldsymbol{\ell} = \mu_0 I_{\text{enc}}$ to a narrow rectangular loop of width $\ell$ perpendicular to $\mathbf{K}$:

$$B_\parallel^{\text{above}}\ell - B_\parallel^{\text{below}}\ell = \mu_0 K \ell$$

Step 4: Tangential discontinuity

$$B_\parallel^{\text{above}} - B_\parallel^{\text{below}} = \mu_0 K$$

Step 5: Compact vector form

In full vector notation: $$\boxed{\hat{n}\times(\mathbf{B}^{\text{above}} - \mathbf{B}^{\text{below}}) = \mu_0\mathbf{K}}$$

This is the magnetic analog of the surface charge boundary condition in electrostatics. The vector potential $\mathbf{A}$ itself is continuous across any interface.

Historical Notes

Oersted (1820): Hans Christian Oersted discovered that electric currents deflect compass needles, establishing the connection between electricity and magnetism. This was the first experimental evidence that electric currents produce magnetic fields.

Biot and Savart (1820): Jean-Baptiste Biot and Felix Savart quantified Oersted's discovery, establishing that the force on a magnetic pole near a long wire falls off as $1/r$ and is perpendicular to the wire. Laplace later cast this in differential form.

Ampere (1820-1825): Andre-Marie Ampere showed that parallel wires carrying currents in the same direction attract, and formulated the circuital law bearing his name. He proposed that all magnetism originates from microscopic current loops, anticipating the modern understanding of magnetic materials.

Maxwell (1865): James Clerk Maxwell recognized the need to add the displacement current term to Ampere's law for consistency with charge conservation, completing the set of equations that unified electricity and magnetism.

Applications

MRI Magnets

Superconducting solenoids producing fields of 1.5-7 T are designed using the exact Biot-Savart solutions for finite solenoids. Shimming coils correct field inhomogeneities to parts per million.

Helmholtz Coils

Two coaxial loops separated by their radius produce a remarkably uniform field in the central region. The design condition follows from requiring $d^2 B_z/dz^2 = 0$ at the midpoint.

Magnetic Shielding

High-permeability shells redirect external fields around a protected region. The boundary conditions at the shell surfaces determine the shielding factor, which scales as $\mu_r$.

Tokamak Design

Toroidal and poloidal field coils in fusion reactors are designed using Biot-Savart calculations. The vector potential is essential for computing magnetic flux surfaces that confine plasma.

Simulation: Magnetostatics

This simulation computes the B-field of a current loop via the Biot-Savart law, verifies Ampere's law for an infinite wire, visualizes the vector potential and B-field of a solenoid, and plots magnetic dipole field lines.

Magnetostatics: Biot-Savart, Ampere's Law & Vector Potential

Python
script.py146 lines

Click Run to execute the Python code

Code will be executed with Python 3 on the server

Video Lectures: Electricity & Magnetism

Sources of Magnetic Fields (UCLA Physics)

Electricity and Magnetism Series - First Lecture

Electricity and Magnetism Series - 2nd Lecture

Electricity and Magnetism Series - Lecture 3

Rate this chapter: