Part 2: From Boltzmann to Fluid Equations

Learning Objectives¶

By the end of Part 2, you will be able to:

• Write the Boltzmann equation and explain each term’s physical meaning

• Apply the moment-taking procedure to derive conservation laws from statistics

• Derive the continuity, momentum, and energy equations from first principles

• Connect pressure to velocity variance and temperature to distribution width

• Recognize that fluid dynamics is emergent statistical behavior

Part 2: From Statistics to Fundamental Physics¶

2.1 The Boltzmann Equation: The Master Evolution Equation¶

Priority: 🟡 Standard Path

The Boltzmann equation is the master equation governing how probability distributions evolve in phase space. Think of it as Newton’s $F = ma$ but for probability clouds rather than individual particles. While Newton tells us how one particle’s position and velocity change over time, Boltzmann tells us how the probability of finding particles at various positions and velocities evolves. It’s the fundamental equation that bridges the microscopic world of individual particles to the macroscopic world of fluid dynamics and thermodynamics.

Imagine tracking not one particle but the probability cloud of where particles might be. This cloud flows through space (particles move), deforms under forces (acceleration changes velocities), and gets scrambled by collisions (randomization). The Boltzmann equation captures all three processes in one elegant framework.

The Boltzmann equation governs how distribution functions evolve in phase space:

Let’s understand each term physically:

• $\frac{\partial f}{\partial t}$: Local time change - How the distribution changes at a fixed point in phase space. Like watching the density of a crowd change while standing still.

• $\vec{v} \cdot \nabla_r f$: Streaming/advection

  • Particles moving in space change local density. Imagine wind blowing smoke—the distribution moves but keeps its shape. If particles move right with velocity $v$, the distribution at your location depletes as they flow past.

• $\frac{\vec{F}}{m} \cdot \nabla_v f$: Force-driven evolution

  • Forces change particle velocities, reshaping the distribution in velocity space. Like gravity pulling all particles downward, shifting the velocity distribution toward negative $v_z$.

• $\left(\frac{\partial f}{\partial t}\right)_{\text{coll}}$: Collision redistribution

  • Collisions scramble velocities, driving the distribution toward Maxwell-Boltzmann. This is nature’s way of maximizing entropy — randomization through molecular chaos.

This equation is exact but unsolvable for 10⁵⁷ particles. The magic happens when we take moments.

Remember from Part 1 that the collision time between particles is much shorter than the star’s dynamical timescale in stellar interiors ($\tau_\text{coll} \ll \tau_\text{dyn}$)? That’s why we can often set the collision integral to zero locally — collisions have already done their work establishing the Maxwell-Boltzmann distribution. The distribution has thermalized so thoroughly that it maintains its equilibrium shape even as the star evolves. This is the magic of LTE: the collision term has already won the race, so we can ignore it in our macroscopic equations.

2.3 The Moment-Taking Machine: From Boltzmann to Fluid Equations¶

Priority: 🔴 Essential

Now comes the magic trick that transforms statistical mechanics into the equations you know and love. We’re going to multiply the unsolvable Boltzmann equation by different powers of velocity and integrate. Each multiplication extracts different physics—like using different filters on the same photograph reveals different features. The blue filter shows the sky, the red filter shows the sunset, the infrared filter shows the heat. Similarly, multiplying by 1 extracts mass flow, by $v$ extracts momentum flow, by $v^2$ extracts energy flow.

This procedure seems almost too simple to work, yet it transforms an equation tracking 10⁵⁷ individual particles into the handful of smooth equations that govern stars, galaxies, and gas clouds. Watch carefully—this is where statistics becomes physics.

The Universal Procedure:

1. Multiply Boltzmann equation by $v^n$

2. Integrate over all velocities

3. Get evolution equation for the $n$-th moment

Let’s do this explicitly for the first few moments.

Zero-th Moment: Mass Conservation¶

Multiply the Boltzmann equation by particle mass $m$ (a constant) and integrate:

(We set the collision integral to zero since it conserves particle number by definition.)

Let’s work through each term carefully:

Term 1:

where $\rho = n m$ is the mass density in g/cm³.

Term 2: $m \int \vec{v} \cdot \nabla_r f d^3v = m \nabla_r \cdot \int \vec{v} f d^3v = \nabla \cdot (\rho\vec{u})$

where $\vec{u} = \langle \vec{v} \rangle$ is the mean velocity in cm/s.

Term 3 (Force term): $m \int \frac{\vec{F}}{m} \cdot \nabla_v f d^3v = \int \vec{F} \cdot \nabla_v f d^3v$

Using integration by parts: $= -\int f (\nabla_v \cdot \vec{F}) d^3v + \underbrace{\left[\vec{F} \cdot f\right]_{v=-\infty}^{v=+\infty}}_{=0}$

The boundary term vanishes because $f$ must decay faster than any polynomial as $|v| \to \infty$ (required for finite mass and energy). For uniform forces (like gravity), $\nabla_v \cdot \vec{F} = 0$, so the entire term vanishes.

Mathematical note: This is where the assumption of “no particles at infinite velocity” enters. It’s physically reasonable but mathematically crucial.

Result - The Continuity Equation:

This is mass conservation! The density changes only due to flow divergence.

First Moment: Momentum Conservation¶

Multiply by $m v_i$ ($i$-th component of momentum) and integrate. After working through the algebra:

where

is the pressure tensor in dyne/cm².

For isotropic pressure (same in all directions): $P_{ij} = P\delta_{ij}$ where $\delta_{ij}$ is the Kronecker delta.

This simplifies to:

This is the Euler equation — Newton’s second law for fluids!

Second Moment: Energy Conservation¶

Multiply by $\frac{1}{2}mv^2$ (kinetic energy) and integrate to get:

where $E$ is the energy density in erg/cm³. The pressure $P$ appears naturally in the energy flux — pressure does work on flowing fluid!

The Beautiful Pattern¶

Let’s step back and see what we’ve accomplished:

Key Takeaway: Each moment extracts a different conservation law.

The procedure is universal — it works for any system where particles follow the Boltzmann equation!

Part 2 Synthesis: The Moment-Taking Framework¶

Priority: 🔴 Essential

You’ve just learned one of the most powerful techniques in physics: transforming unsolvable microscopic equations into tractable macroscopic ones through taking moments. This brief synthesis will cement your understanding before we see it in action.

The Universal Recipe¶

Building on the statistical foundation from Module 1 — where we learned that temperature is a distribution parameter and pressure emerges from ensemble averages — we’ve discovered a recipe that works for any system of particles:

Step 1: Start with the distribution $f(r,v,t)$

• Describes probability of finding particles at position $r$ with velocity $v$

• Contains all microscopic information but is impossibly complex

Step 2: Write the Boltzmann equation

• Governs how $f$ evolves in time

• Exact but unsolvable for realistic systems

Step 3: Take moments (multiply by $v^n$ and integrate)

• $n = 0$: Extracts mass/number density

• $n = 1$: Extracts momentum density

• $n = 2$: Extracts energy/pressure

• Each moment throws away details but keeps essential physics → conservation laws!

Step 4: Get conservation laws

• Continuity equation (mass conservation)

• Momentum equation (Newton’s laws for fluids)

• Energy equation (thermodynamics)

Why This Works: Information Compression¶

Taking moments is fundamentally about information compression. Consider what we’re doing:

• Full information: $f(r,v,t)$ requires tracking $\sim 10^{57} \times 6$ numbers

• After 0th moment: $\rho(r,t)$ requires $\sim 10^3$ numbers (spatial grid points)

• After 1st moment: $u(r,t)$ adds velocity field

• After 2nd moment: $P(r,t)$ adds pressure field

We’ve compressed 10⁵⁸ numbers down to $\sim 10^3$ — a reduction factor of 10⁵⁵!

The “lost” information? The precise velocity of particle number 8,745,293,048,571,293 at this exact instant. We don’t care, and neither does nature at macroscopic scales.

Key Insight: Pressure IS Variance¶

The most profound realization from our moment-taking:

$P = \rho \langle(v - u)^2\rangle = nm \cdot \text{Var}(v)$

Pressure isn’t just “related to” velocity spread — it IS mass density times velocity variance. This identity (not approximation!) means:

• High temperature = large velocity variance = high pressure

• Zero temperature = zero variance = all particles moving together = no pressure

• This works whether “particles” are atoms (giving gas pressure) or stars (giving dynamical “pressure”)

With this framework firmly in mind, you’re ready to see it applied to real stellar physics.