Part 2: From Boltzmann to Fluid Equations
From Particles to Stars | Statistical Thinking Module 2 | COMP 536
Learning Objectives
By the end of Part 2, you will be able to:
Part 2: From Statistics to Fundamental Physics
2.1 The Boltzmann Equation: The Master Evolution Equation
Collision Integral Collision Integral: Mathematical term in Boltzmann equation accounting for how particle interactions change the distribution function.
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:
\[\boxed{\frac{\partial f}{\partial t} + \vec{v} \cdot \nabla_r f + \frac{\vec{F}}{m} \cdot \nabla_v f = \left(\frac{\partial f}{\partial t}\right)_{\text{coll}}}\]
Let’s understand each term physically:
Phase Space Coordinates In phase space, we have 6 coordinates: 3 position \((r)\) and 3 velocity \((v)\). The gradients \(\nabla_r\) and \(\nabla_v\) are with respect to position and velocity separately.
\(\frac{\partial f}{\partial t}\): 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 \(10^{57}\) particles. The magic happens when we take moments.
::: {.callout-note hint} ## 🤔 Quick Check: Understanding the Boltzmann Terms
Consider the Boltzmann equation:
\(\frac{\partial f}{\partial t} + \vec{v} \cdot \nabla_r f + \frac{\vec{F}}{m} \cdot \nabla_v f = \left(\frac{\partial f}{\partial t}\right)_{\text{coll}}\)
- Which term represents particles flowing through space?
- What happens to the distribution if we set all forces to zero?
- Why does the collision term drive distributions toward Maxwell-Boltzmann?
Answer: 1. The \(\vec{v} \cdot \nabla_r f\) term represents streaming — particles moving changes the local density. 2. Without forces, particles just stream along straight lines (ballistic motion). 3. Collisions maximize entropy subject to conservation laws, and Maxwell-Boltzmann is the maximum entropy distribution. :::
Module 2 focuses on stellar interiors where collision times are SHORT (nanoseconds), so LTE holds and we can often ignore the collision term after it establishes equilibrium.
Module 3 covers stellar dynamics where collision times are essentially INFINITE (longer than universe age), so there’s no thermalization.
Same equation, opposite limits.
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 \(10^{57}\) 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:
- Multiply Boltzmann equation by \(v^n\)
- Integrate over all velocities
- 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:
\[m \int \left[\frac{\partial f}{\partial t} + \vec{v} \cdot \nabla_r f + \frac{\vec{F}}{m} \cdot \nabla_v f\right] d^3v = 0\]
(We set the collision integral to zero since it conserves particle number by definition.)
Let’s work through each term carefully:
Term 1: \[m \int \frac{\partial f}{\partial t} d^3v = m \frac{\partial}{\partial t} \int f d^3v = \frac{\partial \rho}{\partial t}\]
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: \[\boxed{\frac{\partial \rho}{\partial t} + \nabla \cdot (\rho\vec{u}) = 0}\]
This is mass conservation! The density changes only due to flow divergence.
The 0th moment (multiply by \(m\) and integrate over all velocities) extracts mass conservation from the Boltzmann equation:
- Particle number at a point changes only due to flow in/out
- This is the continuity equation from fluid dynamics
- It’s literally probability conservation: total probability = 1 always!
- No assumptions needed except \(f \to 0\) as \(v \to \infty\)
First Moment: Momentum Conservation
Multiply by \(m v_i\) (\(i\)-th component of momentum) and integrate. After working through the algebra:
\[\boxed{\frac{\partial (\rho u_i)}{\partial t} + \frac{\partial}{\partial x_j}(\rho u_i u_j + P_{ij}) = \rho F_i}\]
where \[P_{ij} = \rho \langle (v_i - u_i)(v_j - u_j) \rangle\] 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.
Kronecker Delta: \(\delta_{ij} = 1\) if \(i = j\), and \(\delta_{ij} = 0\) if \(i \neq j\). It’s the identity matrix element.
Isotropic: Same in all directions. For pressure, this means no preferred direction of particle motion.
This simplifies to:
\[\boxed{\rho \frac{D\vec{u}}{Dt} = -\nabla P + \rho \vec{F}}\]
This is the Euler equation — Newton’s second law for fluids!
The 1st moment (multiply by \(v\)) extracts momentum conservation from Boltzmann:
- The Euler equation emerges naturally
- Pressure appears as the variance of velocity: \(P = \rho \langle(v-u)^2\rangle\)
- This isn’t an analogy — pressure IS velocity variance times mass density
- Forces appear on the right side as momentum sources
Second Moment: Energy Conservation
Multiply by \(\frac{1}{2}mv^2\) (kinetic energy) and integrate to get:
\[\boxed{\frac{\partial E}{\partial t} + \nabla \cdot [(E + P)\vec{u}] = \rho \vec{F} \cdot \vec{u}}\]
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 2nd moment (multiply by \(\frac{1}{2}m v^2\)) extracts energy conservation from Boltzmann:
- Energy changes due to flux divergence and work done by forces
- Pressure enters the energy flux naturally
- The \(m v^2\) weighting picks out the kinetic energy content
- Higher moments would give heat flux, viscous stress, etc.
The Beautiful Pattern
Let’s step back and see what we’ve accomplished:
| Moment Operation | Multiply Boltzmann by | Integrate to get | Physical Meaning | Conservation Law |
|---|---|---|---|---|
| 0th moment | \(m\) | \(m \int f d^3v = \rho\) | Mass density | Mass conservation |
| 1st moment | \(mv\) | \(m \int vf d^3v = \rho u\) | Momentum density | Momentum conservation |
| 2nd moment | \(m v^2\) | \(m \int v^2f d^3v \propto E\) | Energy density | Energy conservation |
Key Takeaway: Each moment extracts a different conservation law.
The procedure is universal — it works for any system where particles follow the Boltzmann equation!
These three conservation laws, when applied to a spherical star in equilibrium (\(\frac{\partial}{\partial t} = 0\), spherical symmetry), become three of our four stellar structure equations:
0th moment \(\to\) Mass Continuity: \[\frac{dM_r}{dr} = 4\pi r^2 \rho\]
1st moment \(\to\) Hydrostatic Equilibrium: \[\frac{dP}{dr} = -\frac{GM_r\rho}{r^2}\]
2nd moment \(\to\) Energy Conservation: \[\frac{dL_r}{dr} = 4\pi r^2 \rho \epsilon\]
The fourth equation (energy transport) comes from the radiation field, which also follows Boltzmann statistics for photons!
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 \(\to\) 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 \(10^{58}\) numbers down to \(\sim 10^3\) — a reduction factor of \(10^{55}\)!
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”)
From Module 1, Section 1.1, we learned that temperature IS the variance parameter of the velocity distribution: \(\sigma_v^2 = \frac{k_B T}{m}\)
Now we see this isn’t just a relationship — it’s the fundamental bridge between microscopic motion (variance) and macroscopic observables (pressure)! Module 1, Section 2.6 showed that variance measures spread; here that spread literally creates the pressure that prevents stellar collapse.
With this framework firmly in mind, you’re ready to see it applied to real stellar physics.
Where we’ve been: You’ve learned the universal technique of taking moments — multiplying by powers of velocity and integrating — to transform the Boltzmann equation into conservation laws. Each moment extracts different physics: mass (0th), momentum (1st), energy (2nd).
Where we’re going: Part 3 will apply this machinery to stellar interiors. You’ll see how Local Thermodynamic Equilibrium allows one temperature to control everything (pressure, ionization, radiation), and how the four stellar structure equations emerge naturally from our moment-taking procedure. The mathematics becomes less abstract and more physical as we connect these statistical principles to observable stellar properties.
The key insight to carry forward: Stellar structure equations aren’t empirical — they’re the exact statistical behavior of \(10^{57}\) particles expressed through moments of their distribution.