research36 min studyLesson 307 in Physics

Coulomb collision operators

Understand the grazing-collision Landau operator, conservation across species, entropy production, and the limits of fixed-bath relaxation models.

On this page

What you’ll understand

  • Explain the weak-coupling and small-angle assumptions behind Coulomb diffusion.
  • Distinguish per-species number conservation from total momentum and energy conservation.
  • Audit a simple bath operator against the invariants of the full collision operator.

Coulomb collisions in a weakly coupled plasma consist mainly of many small deflections. Their cumulative effect can be represented as friction and diffusion in velocity space rather than as rare hard-sphere impacts. The resulting operator must preserve the correct invariants while producing entropy under its statistical assumptions. These constraints are practical diagnostics: a convenient relaxation term can be useful for test particles while being unsuitable as a closed plasma's full collision operator.

Grazing encounters produce a velocity-space flux

The Coulomb force is long ranged, so a broad range of impact parameters contributes to cumulative scattering. Screening limits large impact parameters, while close encounters or quantum diffraction set an appropriate small-scale cutoff. Their ratio appears logarithmically. A large plasma parameter and a sizeable Coulomb logarithm support the weak-coupling expansion. Strongly coupled matter, large-angle-dominated scattering, and relativistic regimes require more than the elementary nonrelativistic Landau approximation.

For one species colliding with itself, the Landau form is a divergence of an integral velocity-space flux. Its tensor projects perpendicular to relative velocity. Friction and diffusion are linked by the evolving distribution; they are not arbitrary independent knobs. The kernel's geometry and the symmetric treatment of both particles make conservation possible. Retaining only test-particle drag against a frozen background discards that symmetry and transfers momentum and energy to an unmodeled bath.

C[f]=Γ∂∂vi∫d3v′ Uij(u)[f′∂f∂vj−f∂f′∂vj′],Uij=u2δij−uiuju3C[f]=\Gamma\frac{\partial}{\partial v_i}\int d^3v^{\prime}\,U_{ij}(\mathbf u)\left[f^{\prime}\frac{\partial f}{\partial v_j}-f\frac{\partial f^{\prime}}{\partial v_j^{\prime}}\right],\quad U_{ij}=\frac{u^2\delta_{ij}-u_iu_j}{u^3}
Here u = v − v′, f′ = f(v′), repeated indices are summed, and Γ > 0 contains the Coulomb prefactor and logarithm. With f in s³/m⁶, Γ has units m⁶/s⁴ and C has units f/s. This expression is the identical-species nonrelativistic weak-coupling operator.↔ Scroll wide equations horizontally.

Conservation is local but shared across species

The velocity divergence makes particle-number conservation follow from vanishing flux at large velocity. Momentum conservation also uses exchange symmetry between the two colliding populations. For different species, one population's collision term can change its momentum and energy; the changes in the partner population are equal and opposite. Each species' number is separately conserved by elastic collisions, while momentum and kinetic energy are conserved only after summing the appropriate pair or all species.

The full multispecies operator relaxes toward Maxwellians with a common bulk velocity and temperature, with separate densities allowed. Equal temperatures do not mean equal thermal speeds because masses differ. An isolated single-species self-collision operator cannot remove a uniform drift or change the total kinetic energy merely to reach a preferred laboratory Maxwellian. A model that forces every species toward zero drift independently may therefore violate total momentum even if it keeps the density exactly constant.

∫Cs d3v=0,∑s∫msvCs d3v=0,∑s∫msv22Cs d3v=0\int C_s\,d^3v=0,\qquad \sum_s\int m_s\mathbf v C_s\,d^3v=0,\qquad \sum_s\int\frac{m_sv^2}{2}C_s\,d^3v=0
Cs includes all elastic collision partners for species s. The three relations state per-species number conservation and total momentum and kinetic-energy conservation. Velocity-space boundary fluxes must vanish; reactions, external baths, and sources are excluded.↔ Scroll wide equations horizontally.

Entropy and simplified models require an accounting boundary

Under the Landau operator's assumptions, the summed collisional entropy production is nonnegative. The logarithm can be made dimensionless with an arbitrary fixed reference phase-space density; number conservation makes that choice irrelevant to entropy production. Equality characterizes the compatible Maxwellian equilibrium for a connected colliding mixture. This statistical irreversibility differs from ideal Vlasov transport, which rearranges fine structure without itself increasing the fine-grained entropy.

A linear Ornstein–Uhlenbeck bath operator is an instructive approximation, not the full Coulomb operator. It conserves test-particle number while relaxing mean velocity and variance toward prescribed bath values. Unless the bath response is included, the test particles alone do not conserve momentum or energy. Such a model may represent a dilute population in a large reservoir. For a closed plasma or precision transport calculation, conserving field-particle corrections and the correct velocity dependence can be essential.

Cbath[F]=ν∂v[(v−Ub)F+vb2∂vF],u˙=−ν(u−Ub),σ˙2=−2ν(σ2−vb2)C_{\mathrm{bath}}[F]=\nu\partial_v\left[(v-U_b)F+v_b^2\partial_vF\right],\quad \dot u=-\nu(u-U_b),\quad \dot\sigma^2=-2\nu(\sigma^2-v_b^2)
This is a one-dimensional fixed-bath illustration with ν in s⁻¹, Ub in m/s, and vb² a prescribed variance in m²/s². F integrates to test density. The moment formulas require decaying tails and constant bath parameters; they are not general Coulomb relaxation rates.↔ Scroll wide equations horizontally.

Audit a fixed-bath relaxation

For the displayed bath model take ν = 2 s⁻¹, Ub = 1 m/s, vb² = 1 m²/s², initial mean 5 m/s, and initial variance 9 m²/s². Find mean and variance at t = (ln 2)/2 seconds and identify what is conserved.

  1. The mean deviation obeys u−Ub=(5−1)e−νtu-U_b=(5-1)e^{-\nu t}. At the specified time exp(−νt) = 1/2, so u = 1 + 4/2 = 3 m/s.

  2. The variance deviation decays twice as fast: σ2−vb2=(9−1)e−2νt=8/4=2 m2/s2\sigma^2-v_b^2=(9-1)e^{-2\nu t}=8/4=2\,\mathrm{m^2/s^2}. The variance is therefore 3 m²/s².

  3. Integrating the velocity divergence preserves test-particle number. The changed mean and second moment show that test-particle momentum and energy are not conserved; the specified bath supplies their accounting partner.

The result

The mean is 3 m/s and variance is 3 m²/s². Particle number is conserved, while momentum and energy are exchanged with the implicit bath.

Explore the model

Interactive lab

Interactive lab · heat

Diffusion damps fine scales more quickly

Reset and set BOTH Spatial mode n to 1 and Diffusivity κ to 0.5; predict decay exponent 0.5t. Then Reset and set BOTH n to 2 and κ to 0.5; predict exponent 2t. Play from zero in each run. The exact periodic heat mode illustrates scale-sensitive smoothing; it is not a Landau operator, has no friction or bath, and does not test plasma momentum or energy conservation.

Diffusion damps fine scales more quicklyThe exact periodic heat-equation mode is u(x,t) = exp(−κn²t) sin(nx), on [0,2π]. Playback advances physical time. The initial sine is shown separately, and the unnormalized L² norm is √π exp(−κn²t). This analytic instrument does not run a finite-difference solver or test discretization stability. Current readings: Current amplitude: 1; Spatial L² norm: 1.772; Decay rate · κn²: 3.6 /s. All parameters are adjustable with the labeled controls below.Periodic heat equation mode 3 at time 0: the amplitude is 1.0π/2π3π/22π−1−0.500.51x · periodic intervalTemperature perturbation uExact mode · n = 3
Current temperature modeInitial sineCurrent amplitude
Current amplitude
1
Spatial L² norm
1.772
Decay rate · κn²
3.6 /s
t = 0 s

The exact periodic heat-equation mode is u(x,t) = exp(−κn²t) sin(nx), on [0,2π]. Playback advances physical time. The initial sine is shown separately, and the unnormalized L² norm is √π exp(−κn²t). This analytic instrument does not run a finite-difference solver or test discretization stability. Parameter changes pause playback and recompute the model at the current elapsed time. Reset restores both parameters and returns time to zero. Playback starts paused. Adjust either parameter, then choose Play or Step.

Look forA diffusion mechanism can act strongly on fine structure, but its conserved quantities depend on the complete operator.

Check your understanding

For elastic collisions between electrons and ions, which statement is required of the full paired operators?

Try it yourself

At fixed density, temperature, masses, and charges, a leading-log collision frequency is proportional to ln Λ. Compare systems with bmax/bmin = e¹² and e¹⁵. Find the frequency ratio and the inverse-frequency relaxation-time ratio.

Take a moment to make a prediction or work it out on paper. Open a hint when you need one.

Give me a hint

Evaluate the logarithms first and keep the quantities other than the cutoff ratio fixed.

Compare with the solution

The two Coulomb logarithms are 12 and 15. The leading-log collision-frequency ratio is ν2/ν1=15/12=1.25\nu_2/\nu_1=15/12=1.25. The corresponding timescale ratio is τ2/τ1=12/15=0.80\tau_2/\tau_1=12/15=0.80. This comparison concerns the logarithmic factor only; changing density or temperature would also alter the nonlogarithmic prefactor. Both logarithms are large enough for a leading-log illustration, but precise cutoff choices remain model dependent.

Take it with you

  • The Landau operator links friction and diffusion through weak small-angle Coulomb scattering.
  • Elastic collisions conserve each species’ number and the mixture’s total momentum and energy.
  • A fixed-bath operator can be useful while deliberately exchanging invariants with an external reservoir.

Read further

Original sources and deeper explanations to follow at your own pace.

Optional progress, saved in this browser when available. Revisit any lesson at any time.

Lesson not marked complete.