7.1 Boltzmann equation

Foundations of the Boltzmann Equation

The Boltzmann equation describes how the distribution of gas particles evolves over time due to their free motion and collisions. It forms the cornerstone of kinetic theory, connecting microscopic particle dynamics to macroscopic fluid behavior like pressure, viscosity, and heat conduction.

For non-equilibrium statistical mechanics, this equation is essential: it tells you how and why a gas that starts out of equilibrium relaxes toward a steady state, and it gives you quantitative predictions for transport phenomena along the way.

Kinetic Theory Basics

Kinetic theory treats a gas as a huge number of particles (atoms or molecules) in constant random motion. Rather than tracking each particle individually, you use statistical methods to describe their collective behavior.

• Particle interactions obey conservation laws: energy, momentum, and mass are all preserved in collisions.
• Macroscopic observables like pressure and temperature emerge from averaging over the microscopic motions of many particles.
• The theory works best for dilute gases, where particles spend most of their time flying freely between brief, well-separated collisions.

Phase Space Concepts

A single particle's state is fully specified by its position $(x, y, z)$ and momentum $(p_x, p_y, p_z)$, giving a 6-dimensional phase space. Each point in this space represents one possible state of a particle.

• Liouville's theorem states that phase space volume is conserved under Hamiltonian dynamics. Think of it as an incompressible fluid: the "cloud" of states can deform but not shrink or expand.
• Phase space provides the natural framework for tracking how the particle distribution evolves in both position and velocity simultaneously.

Distribution Function Fundamentals

The distribution function $f(\mathbf{r}, \mathbf{v}, t)$ gives the density of particles at position $\mathbf{r}$, with velocity $\mathbf{v}$, at time $t$. More precisely, $f(\mathbf{r}, \mathbf{v}, t) \, d^3r \, d^3v$ is the number of particles in a small phase space volume element around $(\mathbf{r}, \mathbf{v})$.

• Integrating $f$ over all velocities gives the local number density: $n(\mathbf{r}, t) = \int f(\mathbf{r}, \mathbf{v}, t) \, d^3v$
• Integrating over all of phase space yields the total particle number $N$.
• The entire Boltzmann equation is a statement about how $f$ changes in time.

Structure of the Boltzmann Equation

The Boltzmann equation tracks the time evolution of $f(\mathbf{r}, \mathbf{v}, t)$ by accounting for two effects: free streaming (particles moving and responding to forces) and collisions.

The full equation reads:

$\frac{\partial f}{\partial t} + \mathbf{v} \cdot \nabla_{\mathbf{r}} f + \frac{\mathbf{F}}{m} \cdot \nabla_{\mathbf{v}} f = C[f]$

The left-hand side describes what would happen without collisions; the right-hand side corrects for them.

Left-Hand Side Terms

Each term on the left captures a different source of change in $f$:

• $\frac{\partial f}{\partial t}$: the explicit time rate of change of the distribution at a fixed phase space point.
• $\mathbf{v} \cdot \nabla_{\mathbf{r}} f$: the spatial streaming term. Particles with velocity $\mathbf{v}$ carry the distribution from one location to another.
• $\frac{\mathbf{F}}{m} \cdot \nabla_{\mathbf{v}} f$: the force term. External forces (gravity, electric fields, etc.) accelerate particles, shifting the distribution in velocity space.

Without collisions, these three terms together just say that $f$ is constant along particle trajectories, which is Liouville's theorem applied to single particles.

Collision Integral

The collision integral $C[f]$ (sometimes written $\left(\frac{\partial f}{\partial t}\right)_{\text{coll}}$) encodes how binary collisions redistribute particles in velocity space.

• It is a nonlinear functional of $f$, because the collision rate depends on the distribution of both colliding partners.
• Evaluating it exactly requires knowing the differential scattering cross-section for the intermolecular potential.
• Analytical evaluation is rarely possible for realistic potentials, which is why approximation schemes (relaxation time, BGK model) are so important.

Conservation Properties

The collision integral satisfies key constraints that reflect the physics of elastic collisions. Specifically, the collisional invariants $\psi = \{1, \, m\mathbf{v}, \, \frac{1}{2}mv^2\}$ satisfy:

$\int C[f] \, \psi \, d^3v = 0$

This guarantees that collisions conserve:

• Particle number (mass)
• Momentum
• Kinetic energy

These conservation properties are what allow you to derive the familiar fluid equations (continuity, Euler/Navier-Stokes, energy) from the Boltzmann equation by taking velocity moments.

Collision Term Analysis

Modeling the collision integral accurately is the central technical challenge of the Boltzmann equation. The assumptions you make here determine both the tractability and the validity of your results.

Binary Collisions

The Boltzmann equation assumes that only two-body collisions matter. This is justified when the gas is dilute enough that the probability of three or more particles being close together simultaneously is negligible.

• Each collision is characterized by pre-collision velocities $(\mathbf{v}_1, \mathbf{v}_2)$ and post-collision velocities $(\mathbf{v}_1', \mathbf{v}_2')$.
• The differential scattering cross-section $\sigma(g, \Omega)$ determines the probability of a given deflection, where $g = |\mathbf{v}_1 - \mathbf{v}_2|$ is the relative speed and $\Omega$ is the solid angle.
• This binary assumption breaks down for dense gases, where multi-particle correlations become significant.

Molecular Chaos Assumption (Stosszahlansatz)

This is the most consequential assumption in the entire derivation. Molecular chaos (German: Stosszahlansatz) asserts that the velocities of two particles about to collide are statistically uncorrelated.

Concretely, it means you can factorize the two-particle distribution:

$f^{(2)}(\mathbf{r}, \mathbf{v}_1, \mathbf{v}_2, t) \approx f(\mathbf{r}, \mathbf{v}_1, t) \, f(\mathbf{r}, \mathbf{v}_2, t)$

Why this matters so much:

1. It closes the equation at the single-particle level. Without it, the evolution of $f^{(1)}$ depends on $f^{(2)}$, which depends on $f^{(3)}$, and so on (the BBGKY hierarchy).
2. It introduces irreversibility into an equation derived from time-reversible microscopic dynamics. The factorization discards the correlations that collisions create, effectively building in a preferred direction of time.
3. It is valid for dilute gases where particles travel many mean free paths between collisions, so any correlations from a previous collision are "forgotten." It breaks down for dense gases or strongly interacting systems.

Equilibrium Solutions

At equilibrium, the distribution function stops changing: $\frac{\partial f}{\partial t} = 0$, and there are no spatial gradients or external forces driving the system. The collision integral alone determines the equilibrium form of $f$.

Maxwell-Boltzmann Distribution

The unique equilibrium solution of the Boltzmann equation for an ideal gas is the Maxwell-Boltzmann distribution:

$f_{\text{eq}}(\mathbf{v}) = n \left(\frac{m}{2\pi k_B T}\right)^{3/2} \exp\left(-\frac{m v^2}{2 k_B T}\right)$

where $n$ is the number density, $m$ is the particle mass, $k_B$ is Boltzmann's constant, and $T$ is the temperature.

• This is the distribution that maximizes entropy subject to fixed total energy and particle number.
• The collision integral vanishes identically when $f = f_{\text{eq}}$, meaning collisions produce no net change: the rate of scattering into any velocity equals the rate of scattering out.

H-Theorem

Boltzmann's H-theorem proves that the quantity

$H(t) = \int f \ln f \, d^3v$

is a monotonically non-increasing function of time under the Boltzmann equation. Since $H$ is the negative of the entropy (up to a constant), this means entropy increases until equilibrium is reached.

• The H-theorem provides a microscopic foundation for the second law of thermodynamics.
• It resolves (at least partially) the puzzle of how irreversible macroscopic behavior emerges from reversible microscopic laws. The resolution lies in the molecular chaos assumption, which breaks time-reversal symmetry.
• $H$ reaches its minimum (entropy reaches its maximum) precisely when $f = f_{\text{eq}}$.

Approach to Equilibrium

A non-equilibrium distribution relaxes toward the Maxwell-Boltzmann form on a timescale set by the collision frequency.

• The relaxation time $\tau$ is roughly the mean time between collisions. Higher density or larger cross-sections mean faster relaxation.
• Different physical quantities can relax on different timescales. For example, the velocity distribution might isotropize (losing directional bias) faster than the energy distribution equilibrates.
• In the relaxation time approximation, the approach is exponential: deviations from equilibrium decay as $e^{-t/\tau}$.

Linearized Boltzmann Equation

When a system is only slightly out of equilibrium, you can linearize the Boltzmann equation around $f_{\text{eq}}$. This dramatically simplifies the math and is the standard route to deriving transport coefficients.

Small Perturbations

Write the distribution as:

$f = f_{\text{eq}}(1 + \phi)$

where $\phi(\mathbf{r}, \mathbf{v}, t)$ is a small dimensionless perturbation ($|\phi| \ll 1$).

• Substituting into the Boltzmann equation and dropping terms of order $\phi^2$ and higher yields a linear integro-differential equation for $\phi$.
• The linearized collision operator is a linear integral operator acting on $\phi$, which is much more tractable than the full nonlinear collision integral.
• This approach is valid for systems with small spatial gradients, weak external fields, or small deviations from equilibrium.

Relaxation Time Approximation (BGK Model)

The simplest model for the linearized collision term replaces the full collision operator with:

$C[f] \approx -\frac{f - f_{\text{eq}}}{\tau}$

This is the Bhatnagar-Gross-Krook (BGK) approximation. It says the distribution relaxes toward equilibrium at a uniform rate $1/\tau$.

• The single parameter $\tau$ captures the overall collision timescale.
• It gives qualitatively correct results for many transport problems and is often the starting point for analytical work.
• Its main limitation: it uses one relaxation time for all processes, whereas in reality, viscosity and thermal conductivity involve different velocity-space modes that relax at different rates.

Chapman-Enskog Expansion

The Chapman-Enskog method is a systematic perturbation expansion that extracts hydrodynamic equations from the Boltzmann equation.

1. Expand $f$ in powers of the Knudsen number $\text{Kn} = \ell / L$, where $\ell$ is the mean free path and $L$ is the macroscopic length scale.
2. At zeroth order ($\text{Kn} \to 0$), you recover the Euler equations of inviscid fluid dynamics, with $f = f_{\text{eq}}$.
3. At first order in $\text{Kn}$, you get the Navier-Stokes equations, and the transport coefficients (viscosity, thermal conductivity) emerge as integrals over the linearized collision operator.
4. Higher orders yield the Burnett and super-Burnett equations, which include corrections for moderately rarefied flows, though these higher-order equations can have stability issues.

The power of this method is that it derives macroscopic fluid equations from first principles, with no free parameters beyond the intermolecular potential.

Transport Coefficients

The Boltzmann equation provides explicit expressions for macroscopic transport coefficients in terms of microscopic quantities (particle mass, cross-section, temperature). These connect the kinetic theory directly to measurable fluid properties.

Viscosity

Shear viscosity $\eta$ measures a fluid's resistance to shear deformation. From kinetic theory for a dilute gas:

$\eta \propto \frac{m \bar{v}}{\sigma}$

where $\bar{v} \sim \sqrt{k_B T / m}$ is the mean thermal speed and $\sigma$ is the collision cross-section.

• For gases, viscosity increases with temperature ($\eta \propto T^{1/2}$ for hard spheres), because faster particles transport more momentum. This is the opposite of liquids.
• Viscosity is independent of pressure for an ideal gas at fixed temperature, because the increased collision rate at higher density is exactly offset by the shorter mean free path.

Thermal Conductivity

Thermal conductivity $\kappa$ quantifies heat transport due to temperature gradients. The kinetic theory result has a similar structure:

$\kappa \propto \frac{c_v \bar{v}}{\sigma}$

where $c_v$ is the specific heat per particle.

• Like viscosity, $\kappa$ increases with temperature for gases.
• The ratio $\eta c_v / \kappa$ is a dimensionless number related to the Prandtl number, and its value depends on the molecular interaction model.

Diffusion Coefficient

The self-diffusion coefficient $D$ describes how particles spread through a gas due to concentration gradients:

$D \propto \frac{\bar{v}}{n \sigma}$

• $D$ increases with temperature (faster particles diffuse more quickly).
• $D$ varies inversely with pressure (or number density $n$) at constant temperature, because more frequent collisions impede particle migration.

Applications of the Boltzmann Equation

The Boltzmann equation and its variants appear across many areas of physics and engineering, wherever you need to describe particle transport beyond the continuum fluid limit.

Rarefied Gas Dynamics

When the Knudsen number $\text{Kn} = \ell / L$ approaches or exceeds unity, the continuum (Navier-Stokes) description breaks down and you must solve the Boltzmann equation directly.

• This regime arises in high-altitude aerodynamics (e.g., spacecraft re-entry), vacuum systems, and microfluidic devices (MEMS).
• Characteristic rarefied-gas phenomena include velocity slip at solid boundaries and temperature jump between a gas and a wall, neither of which appear in standard fluid mechanics.

Plasma Physics

In plasmas, charged particles interact through long-range Coulomb forces, and the Boltzmann equation is coupled to Maxwell's equations for the electromagnetic fields. The resulting system is often called the Vlasov-Boltzmann equation.

• Applications include fusion reactor design, astrophysical plasmas, and industrial plasma processing.
• Kinetic effects like Landau damping (collisionless wave damping) and plasma instabilities require the full kinetic description and cannot be captured by fluid models alone.

Neutron Transport

The neutron transport equation is structurally identical to the Boltzmann equation, with neutrons playing the role of gas particles and nuclear reactions replacing molecular collisions.

• Neutrons undergo scattering, absorption, and fission, each characterized by energy-dependent cross-sections.
• This equation is central to nuclear reactor design, criticality safety analysis, and radiation shielding calculations.

Numerical Methods

The full Boltzmann equation is a nonlinear integro-differential equation in 6-dimensional phase space plus time. Direct analytical solutions exist only for the simplest cases, so numerical methods are essential for realistic problems.

Direct Simulation Monte Carlo (DSMC)

DSMC is a particle-based stochastic method developed by Graeme Bird in the 1960s.

1. Represent the gas by a large number of computational particles, each standing for many real molecules.
2. Advance particle positions according to their velocities (free streaming step).
3. Select collision pairs stochastically within spatial cells, using the collision cross-section to determine outcomes (collision step).
4. Sample macroscopic quantities (density, velocity, temperature) by averaging over particles in each cell.

DSMC is the standard tool for rarefied gas flows and high-speed aerodynamics. Its computational cost scales with the number of particles and becomes expensive at low Knudsen numbers (near-continuum regime).

Lattice Boltzmann Method (LBM)

LBM discretizes the Boltzmann equation on a regular spatial lattice with a small set of discrete velocities.

• It typically uses the BGK collision operator for simplicity.
• In the macroscopic limit (small $\text{Kn}$), LBM recovers the Navier-Stokes equations, so it functions as an alternative CFD method.
• Its strengths are easy handling of complex geometries, natural parallelization, and straightforward extension to multiphase and multicomponent flows.

Discrete Velocity Models

These methods approximate the continuous velocity space with a finite set of discrete velocity vectors.

• The Boltzmann equation reduces to a system of coupled partial differential equations, one for each discrete velocity.
• Standard numerical techniques for hyperbolic conservation laws (finite volume, finite difference) can then be applied.
• The trade-off is between the number of discrete velocities (accuracy in velocity space) and computational cost.

Limitations and Extensions

The standard Boltzmann equation rests on specific assumptions (dilute gas, classical particles, non-relativistic speeds). Relaxing these assumptions leads to important generalizations.

Dense Gases

The binary collision assumption fails when the gas is dense enough that particles are frequently close to multiple neighbors simultaneously.

• The Enskog equation extends the Boltzmann framework to moderately dense gases by accounting for the finite size of particles (excluded volume effects and collisional transfer of momentum/energy).
• For strongly interacting or liquid-like systems, the full BBGKY hierarchy provides a systematic (though difficult) framework that does not assume molecular chaos.

Quantum Boltzmann Equation

At low temperatures or high densities, quantum statistics become important and the classical distribution function must be replaced.

• The Wigner function serves as a quantum analog of the classical phase space distribution.
• For fermions, the collision term must respect the Pauli exclusion principle, modifying the scattering rates with blocking factors $(1 - f)$.
• For bosons, stimulated scattering enhances the collision rates with factors $(1 + f)$, and the equation can describe phenomena like Bose-Einstein condensation.

Relativistic Boltzmann Equation

When particle speeds approach the speed of light, the standard Boltzmann equation must be reformulated using relativistic kinematics.

• The distribution function is defined on the mass shell in 4-momentum space, and the equation is written in a manifestly Lorentz-covariant form.
• Applications include early-universe cosmology, relativistic heavy-ion collisions, and high-energy astrophysical plasmas.
• Additional complications arise from particle creation and annihilation processes, which have no analog in the classical Boltzmann equation.