advanced27 min studyLesson 356 in Physics

Organize the hydrodynamic limit by a small parameter

Separate solvability, moment matching, and the order of dissipative corrections.

On this page

What you’ll understand

  • Define spatial and temporal kinetic-to-macroscopic scale ratios.
  • Explain how Chapman–Enskog solvability uses collision invariants.
  • Identify when a first-order constitutive closure is controlled.

Hydrodynamics is more than a smooth plot of particle density. It is a regime in which collisions relax fast kinetic distortions before macroscopic fields vary appreciably. The Chapman–Enskog construction organizes that separation of scales into an expansion about a local Maxwellian. Zeroth order gives ideal-fluid fluxes and first order gives viscosity and heat conduction. The expansion also tells us where its own formulas cease to be justified, particularly near walls, sharp fronts, and strongly nonthermal distributions.

Scale separation is local

The Knudsen number compares a molecular mean free path with a macroscopic variation length. The relevant length may be a channel width, a temperature-gradient length, or a wavelength; choosing the largest apparatus dimension can conceal a much steeper local variation. A temporal condition also compares the collision time with the field-evolution time. Small spatial Knudsen number alone does not control a rapidly driven high-frequency disturbance or an initial layer that still contains strong kinetic distortions.

After nondimensionalization, the streaming terms are of order one while collisions carry an inverse small parameter. Write the distribution as a leading term plus successive corrections. The leading collision equation sets the zeroth-order shape to a local Maxwellian, assuming ordinary elastic collision connectivity. Small parameter refers to a scale hierarchy, not automatically to low flow velocity: Mach number and Knudsen number are separate quantities, though some asymptotic regimes couple them.

Kn=ℓL≪1,ωτ≪1,∂tf+v⋅∇xf=1εC[f],f=M+εf(1)+⋯\mathrm{Kn}=\frac{\ell}{L}\ll1,\qquad \omega\tau\ll1,\qquad \partial_tf+\mathbf v\cdot\nabla_xf=\frac1{\varepsilon}\mathcal C[f],\qquad f=M+\varepsilon f^{(1)}+\cdots
ℓ is mean free path, L a chosen gradient length, τ collision time, and ω an angular driving rate. In the last two expressions variables and the operator have been nondimensionalized and ε is of order Kn.↔ Scroll wide equations horizontally.

The collision null space fixes solvability

Linearize the collision operator about the local Maxwellian. At first order, streaming of that Maxwellian drives a correction through the linearized operator. Its invariant moments must vanish for a solution to exist, because collisions cannot create the protected moments. These solvability conditions are the leading conservation equations. In Chapman–Enskog theory, time derivatives are expanded consistently and lower-order balance laws replace them when forming the next correction; one does not invert the collision operator blindly on its null space.

Choose each correction to have zero density, momentum, and energy moments. This matching convention ensures the hydrodynamic fields remain defined by the full distribution rather than being counted again in the correction. The nonzero higher moments of the first correction then supply viscous stress and heat flux. A finite moment truncation and a Chapman–Enskog gradient expansion are related closure tools, but they are not interchangeable derivations and need not have the same stability properties.

Lf(1)=D0M,∫f(1){1,v,v2} d3v=0\mathcal Lf^{(1)}=\mathcal D_0M,\qquad \int f^{(1)}\{1,\mathbf v,v^2\}\,d^3v=0
L is the linearized dimensionless collision operator and D₀ is streaming using leading-order time evolution. The zero moments are a matching convention; solvability separately requires the source’s invariant moments to vanish.↔ Scroll wide equations horizontally.

First order is a claim with a domain

If gradients and nonhydrodynamic corrections remain small, inserting the first correction into moment fluxes gives the Navier–Stokes–Fourier constitutive laws. This is a formal asymptotic construction, not a universal error theorem for every initial or boundary-value problem. Near a wall, particles arriving from the boundary have an angular distribution set by accommodation and reflection. A kinetic boundary layer of order a mean free path can coexist with a well-controlled bulk continuum limit.

The heat lab tests one consequence of a diffusive closure. Reset and set mode to 1 and diffusivity to 0.5, then Step once; predict decay exponent 0.25. Reset and set mode to 4 and diffusivity to 0.5, then Step once; predict exponent 4. Faster spatial variation damps faster within that equation. The exact single-mode display contains no mean free path or collision time and cannot establish that the high-mode continuum approximation remains valid.

Pij=pδij−2ηSij+O(Kn2),q=−κ∇T+O(Kn2)P_{ij}=p\delta_{ij}-2\eta S_{ij}+O(\mathrm{Kn}^2),\qquad \mathbf q=-\kappa\nabla T+O(\mathrm{Kn}^2)
Sij=½(∂iuj+∂jui)−δij∇·u/3 is the traceless strain rate in s⁻¹. η has units Pa·s and κ W/(m·K). Remainders refer to consistent dimensionless asymptotic ordering of the fluxes.↔ Scroll wide equations horizontally.

Audit the expansion parameter

A gas has characteristic molecular speed 400 m/s and relaxation time 2 μs. A temperature pattern varies over 4 cm and is driven at angular rate 500 s⁻¹. Estimate mean free path, Knudsen number, temporal ratio, and nominal first/second correction sizes.

  1. Using the specified characteristic-speed estimate, the mean free path is ℓ=vτ=400×2×10⁻⁶=8×10⁻⁴ m, or 0.8 mm. This defines the convention for the estimate; different thermal-speed choices change order-one factors.

  2. The gradient length is 0.04 m, giving Kn=0.0008/0.04=0.02. The temporal ratio is ωτ=500×2×10⁻⁶=0.001, so both supplied scale ratios are small.

  3. With order-one dimensionless coefficients and well-prepared bulk data, a first correction of order 0.02 and a next correction of order 0.0004 are plausible formal estimates. They are not certified error bounds and do not resolve wall layers or unusually large coefficients.

The result

The estimates are ℓ=0.8 mm, Kn=0.02, and ωτ=0.001. A bulk first-order closure is plausible under the additional regularity and near-equilibrium assumptions.

Explore the model

Interactive lab

Interactive lab · heat

Separate fast damping from validity

Reset and set n=1, κ=0.5, then Step once; predict amplitude exp(−0.25). Reset and set n=4, κ=0.5, then Step once; predict exp(−4).

Separate fast damping from validityThe 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 forThe analytic heat mode exposes the κn² rate. It cannot test kinetic scale separation, numerical stability, or the validity of a high-wavenumber fluid closure.

Check your understanding

Why impose zero invariant moments on the first distribution correction?

Try it yourself

Keep the same molecular speed and collision time but reduce the gradient length to 1 mm and drive at angular rate 10⁶ s⁻¹. Evaluate the two scale ratios and assess a first-order bulk closure.

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

Give me a hint

Recompute both ratios; do not infer validity from the earlier apparatus length.

Compare with the solution

The mean free path remains 0.8 mm, so Kn=0.8. The temporal ratio is ωτ=2. Neither is small. A first-order Chapman–Enskog closure is uncontrolled for this forcing and spatial scale; more terms are not automatically a remedy, and a kinetic treatment with appropriate initial and boundary data is needed.

Take it with you

  • Knudsen and temporal ratios describe distinct scale requirements.
  • Collision invariants impose solvability and moment matching.
  • First-order hydrodynamics is a controlled regime, not a universal kinetic solver.

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.