Skip to content

Guiding Center & Drift Theory

Intuition

A charged particle in a magnetic field is a fast circle glued to a slow wanderer. Squint past the rapid gyration and you see the circle's center drifting smoothly — that center is the guiding center, and drift theory is the physics of where it goes. It converts an impossible-to-follow spiral into four clean drift formulas.

The decomposition

\[\mathbf{x}(t) = \underbrace{\mathbf{X}_\text{gc}(t)}_{\text{guiding center (slow)}} + \underbrace{\boldsymbol{\rho}(t)}_{\text{gyration (fast)}}\]

Validity conditions: the Larmor radius is small compared to field-variation scales (\(\rho_L \ll L_B\)), and the gyration period is short compared to field time scales (\(2\pi/\omega_c \ll \tau_\text{field}\)). Averaging over the gyration then yields:

\[\boxed{\,\mathbf{v}_{\perp,\text{gc}} = \mathbf{v}_E + \mathbf{v}_{\nabla B} + \mathbf{v}_c + \mathbf{v}_P\,}\]

The four drifts

Drift Formula Cause Charge-dependent?
E×B \(\mathbf{v}_E = \dfrac{\mathbf{E}\times\mathbf{B}}{B^2}\) electric field No
Grad-B \(\mathbf{v}_{\nabla B} = -\dfrac{m v_{\perp}^2}{2qB^3}\nabla B\times\mathbf{B}\) \(\lvert\mathbf{B}\rvert\) gradient Yes
Curvature \(\mathbf{v}_c = \dfrac{m v_{\parallel}^2}{qB^2}\dfrac{\hat{\mathbf{R}}}{R}\times\mathbf{B}\) curved field lines Yes
Polarization \(\mathbf{v}_P \approx \dfrac{m}{qB^2}\dfrac{d\mathbf{E}_\perp}{dt}\) time-varying \(\mathbf{E}\) Yes (and mass-dep.)

Charge dependence matters enormously: charge-dependent drifts separate ions from electrons → currents and charge separation → new electric fields → new E×B drifts. That chain is the engine of tokamak vertical drift, of the polarization current in waves, and of much plasma misbehavior.

Physical pictures

  • E×B: during the half-gyration where the particle moves with \(\mathbf{E}\) it speeds up (bigger Larmor radius); against, it slows (smaller). Fat half + thin half = net sideways march — independent of charge and mass (both flip together).
  • Grad-B: larger circles where \(B\) is weak, tighter where strong; the orbit doesn't close and inches sideways.
  • Curvature: a particle streaming along a bent field line feels a centrifugal push \(mv_\parallel^2/R\) outward; crossed with \(\mathbf{B}\), that force drifts it.
  • Polarization: when \(\mathbf{E}\) changes, the E×B drift must change too; the inertia-limited catch-up appears as a drift along \(d\mathbf{E}/dt\) — heavy ions lag more, carrying most of the polarization current.

Numerical verification (the course's approach)

Rather than trust the averaging, integrate the full Lorentz force with the 3-D leapfrog solver and watch each drift emerge:

  1. Uniform E, B → helical cycloid drifting at exactly \(E/B\)
  2. Line charge + axial B → azimuthal E×B orbit around the wire
  3. Toroidal field (\(B \propto 1/r\)) → vertical grad-B + curvature drift, the reason a pure toroidal trap fails (tokamaks add poloidal field to short it out)
  4. Magnetic mirror → trapping and reflection
  5. Wave fields → drift-theory regime \(\alpha \ll 1\) through stochastic \(\alpha \gtrsim 1\), where \(\alpha = mk^2\phi/qB^2\)

Common mistakes

  • Applying drift formulas where gyration isn't fast. If \(\rho_L \sim L_B\) the decomposition fails; only the full orbit integration is trustworthy (that's why the course teaches both).
  • Expecting E×B to create current. It moves ions and electrons together — flow, not current. The charge-dependent drifts make the currents.
  • Dropping the parallel dynamics. Drifts describe motion ⊥ to \(\mathbf{B}\); along the field, particles stream freely (or mirror — see magnetic mirror).

Knowledge graph position

Prerequisites: Lorentz force, Leapfrog, cyclotron motion. Leads to: Magnetic mirror, Fermi acceleration.

Quiz

Q1 (computational). \(E = 100\) V/m ⊥ \(B = 0.1\) T. Find the E×B drift speed for an electron and for a proton.

Answer

\(v_E = E/B = 1000\) m/s — identical for both. Charge and mass cancel in \(\mathbf{E}\times\mathbf{B}/B^2\).

Q2 (conceptual). In a purely toroidal field, why do particles escape even though field lines close on themselves?

Answer

\(B \propto 1/r\) gives both grad-B and curvature drifts, which are vertical and charge-dependent. Charges separate vertically, the resulting E field drives an outward E×B drift for everyone — the plasma expels itself.

Q3 (MCQ). Which drift survives even when \(q \to -q\) and \(m \to m_e\)?

  • (a) grad-B (b) curvature (c) E×B (d) polarization
Answer

(c). E×B is charge- and mass-independent — it is really the velocity of the frame in which \(\mathbf{E}_\perp\) vanishes.