Magnetic Mirror
Intuition
Squeeze magnetic field lines together at two ends of a tube and you've built a bottle for charged particles. A particle spiraling toward the pinch feels the field strengthen; its gyration steals energy from its forward motion until — if the pinch is strong enough — it stops, turns, and bounces back. Two mirrors face-to-face trap particles indefinitely. Nature builds them too: Earth's Van Allen belts are mirror machines.
The field
The course models a mirror through its flux function in cylindrical coordinates \((r, \theta, z)\):
which evaluates to
\(|\mathbf{B}|\) is weakest at the midplane (\(z = 0\)) and grows toward the mirror points. Field lines are contours of \(\psi\) — and a well-magnetized particle's guiding center stays on its contour.
Why particles reflect: the adiabatic invariant
The magnetic moment
is an adiabatic invariant: nearly constant when the field changes slowly over a gyration. Since \(B\) does no work, \(v^2 = v_\parallel^2 + v_\perp^2\) is also constant. As the particle rides into stronger \(B\), constant \(\mu\) forces \(v_\perp^2\) to grow — so \(v_\parallel^2\) must shrink. If \(B\) reaches \(B_\text{reflect} = B_\text{min}\, v^2/v_{\perp,0}^2\) before the throat, the particle reflects.
The loss cone
Particles with too much parallel velocity escape. With mirror ratio \(R_m = B_\text{max}/B_\text{min}\), the trapping condition on the midplane pitch angle \(\theta\) is
Velocities inside the loss cone (\(\theta < \theta_\text{trap}\), \(\sin\theta_\text{trap} = R_m^{-1/2}\)) stream straight through. This leak is why mirror fusion machines struggled — collisions constantly scatter particles into the cone.
Numerical experiment
Integrate orbits with the 3-D leapfrog solver in the field above (worked example):
- Launch with various \(v_\parallel/v_\perp\): watch trapped orbits bounce and passing orbits escape
- Verify \(\mu\) stays constant while \(v_\parallel\) oscillates
- Verify the guiding center hugs a \(\psi\)-contour
Common mistakes
- Thinking the magnetic force decelerates the particle. \(\mathbf{B}\) does no work; it redirects kinetic energy from parallel to perpendicular. Total speed never changes.
- Treating \(\mu\) as exactly conserved. It's adiabatic — conserved to exponential accuracy while fields vary slowly; fast field changes (or sharp field gradients) break it.
- Forgetting the loss cone. A mirror is a leaky bottle by construction; only pitch angles outside the cone are held.
Related concepts
- Guiding center & drifts — framework
- Fermi acceleration — what moving mirrors do
- Adiabatic invariants in PC368 — the theory treatment
- Magnetic mirror orbit example
Knowledge graph position
Prerequisites: Guiding center, Leapfrog. Leads to: Fermi acceleration.
Quiz
Q1 (computational). A mirror has \(R_m = 4\). What fraction of an isotropic particle population is trapped?
Answer
Loss cone half-angle: \(\sin\theta_c = 1/2 \Rightarrow \theta_c = 30°\). Solid-angle fraction lost (both cones): \(1 - \cos\theta_c = 1 - \frac{\sqrt3}{2} \approx 0.134\) each side ⇒ trapped fraction \(= \cos\theta_c \approx 87\%\).
Q2 (conceptual). A trapped particle's \(v_\perp\) is largest at which point of its bounce?
Answer
At the mirror (turning) points, where \(B\) is largest: \(\mu = mv_\perp^2/2B\) const forces \(v_\perp^2 \propto B\). There \(v_\parallel = 0\) — all energy is perpendicular.
Q3 (MCQ). Earth's radiation belts trap particles because the geomagnetic field:
- (a) is uniform (b) strengthens toward the poles (c) vanishes at the equator (d) is purely radial
Answer
(b). Field lines converge near the magnetic poles — natural mirror points; particles bounce pole-to-pole.