Ampère's Law
Ampère's Law: Theoretical Foundations
Ampère's Law
Professor, how is Ampère's Law different from the Biot-Savart Law?
They express the same physics in different forms. The line integral of the magnetic field along a closed path is equal to the total current enclosed.
$$ \oint_C \mathbf{H} \cdot d\mathbf{l} = I_{enc} $$
Differential form (static magnetic field version of Maxwell's 4th equation):
$$ \nabla \times \mathbf{H} = \mathbf{J} $$
Gauss's Law ($\nabla \cdot \mathbf{D} = \rho$) is the electric field counterpart of Ampère's Law.
For problems with high symmetry, can we solve directly using Ampère's Law?
Current Distribution Ampèrian Loop $H$ Infinite straight wire Concentric circle $I/(2\pi r)$ Inside solenoid Rectangle $nI$ Toroidal coil Central circle $NI/(2\pi r)$
Summary
- $\nabla \times \mathbf{H} = \mathbf{J}$ — Maxwell's 4th equation (static magnetic field)
- Governing equation for FEM static magnetic field analysis — Used in combination with vector potential
Coffee Break Trivia
Ampère himself was a "theorist who didn't experiment"
André-Marie Ampère, the discoverer of Ampère's Law, was a rare type of physicist who hardly conducted experiments himself. Shortly after Ørsted reported in 1820 that a magnetic field is generated around an electric current, Ampère constructed and formulated the theory in just a few weeks. Not from experiment to theory, but grasping the essence mathematically immediately upon hearing others' experimental reports—this super-fast abstraction ability gave birth to the concise single line: ∮H·dl = I_enc.
Professor, how is Ampère's Law different from the Biot-Savart Law?
They express the same physics in different forms. The line integral of the magnetic field along a closed path is equal to the total current enclosed.
Differential form (static magnetic field version of Maxwell's 4th equation):
Gauss's Law ($\nabla \cdot \mathbf{D} = \rho$) is the electric field counterpart of Ampère's Law.
For problems with high symmetry, can we solve directly using Ampère's Law?
| Current Distribution | Ampèrian Loop | $H$ |
|---|---|---|
| Infinite straight wire | Concentric circle | $I/(2\pi r)$ |
| Inside solenoid | Rectangle | $nI$ |
| Toroidal coil | Central circle | $NI/(2\pi r)$ |
- $\nabla \times \mathbf{H} = \mathbf{J}$ — Maxwell's 4th equation (static magnetic field)
- Governing equation for FEM static magnetic field analysis — Used in combination with vector potential
Ampère himself was a "theorist who didn't experiment"
André-Marie Ampère, the discoverer of Ampère's Law, was a rare type of physicist who hardly conducted experiments himself. Shortly after Ørsted reported in 1820 that a magnetic field is generated around an electric current, Ampère constructed and formulated the theory in just a few weeks. Not from experiment to theory, but grasping the essence mathematically immediately upon hearing others' experimental reports—this super-fast abstraction ability gave birth to the concise single line: ∮H·dl = I_enc.
Computational Methods for Ampère's Law
FEM Formulation
Substitute $\mathbf{B} = \mu\mathbf{H}$ and $\mathbf{B} = \nabla \times \mathbf{A}$ into $\nabla \times \mathbf{H} = \mathbf{J}$:
$\nu = 1/\mu$: Magnetic reluctivity. For nonlinear materials (iron cores), $\nu = \nu(|\mathbf{B}|)$ and Newton-Raphson iteration is required.
2D vs. 3D
| Dimension | Unknowns | Element | Computational Cost |
|---|---|---|---|
| 2D | $A_z$ (scalar) | Nodal element | Small |
| 2D axisymmetric | $rA_\phi$ (scalar) | Nodal element | Small |
| 3D | $\mathbf{A}$ (vector) | Edge element | Large |
Summary
- 2D is a scalar problem — Good computational efficiency
- 3D uses edge elements — Gauge condition handling is required
- Nonlinear materials → Newton-Raphson iteration
Ampèrian integration path—"Where to cut" can change computational cost by 100 times
When numerically implementing Ampère's Law, the computational cost changes drastically depending on how the Ampèrian loop is set. For circular or rectangular loops utilizing symmetry, the closed integral can be solved analytically, but for coils with complex shapes, discretization along an arbitrary integration path is necessary. In practice, the technique of "identifying symmetry planes to reduce the model to 1/4" is standard, and experienced engineers develop a habit of instantly looking for symmetry axes when seeing a shape. Saving computational resources starts with method selection.