Fluid Dynamics: Conservation, Flow and Stability: Petschek Reconnection

Petschek Reconnection

Slow shocks add 2-D structure → faster rate $v_{\text{in}}/v_A \sim 1/\ln\text{S}$, but requires localized resistivity.

Modern view (plasmoid instability)

At very high $\text{S}$, Sweet-Parker sheet itself is unstable to plasmoid formation; effective reconnection rate becomes $\sim 0.01\,v_A$, independent of $\text{S}$. Solves the long-standing reconnection-rate problem.

Applications: solar flares (releasing $10^{32}$ erg in minutes), coronal mass ejections, magnetospheric substorms, sawtooth crashes in tokamaks, magnetar bursts.

MHD Instabilities

MRI's importance: explained accretion disk turbulence/angular momentum transport — Galaxy-scale and quasar physics depends on it.

Dynamos

Question: how does a conducting fluid sustain a magnetic field against ohmic decay?

Cowling's antidynamo theorem: axisymmetric fields cannot be sustained — symmetry breaking required.

Mean-field dynamo equations: $$\partial_t\mathbf{B} = \nabla\times(\boldsymbol\alpha\mathbf{B}) + \nabla\times(\mathbf{v}\times\mathbf{B}) - \nabla\times(\beta\nabla\times\mathbf{B})$$

with $\alpha$-effect (helicity from rotation + convection) producing poloidal from toroidal; differential rotation ($\Omega$-effect) producing toroidal from poloidal. Closed loop → exponential growth.

Realizations:

• Earth: $\alpha$-$\Omega$ dynamo in outer core; reverses on $\sim 10^5$ yr timescale.
• Sun: 11-year cycle from $\alpha$-$\Omega$ at base of convective zone (tachocline).
• Galaxies: $\alpha$-$\Omega$ in disks; large-scale fields amplified over Gyr.

Workflow / Process

flowchart TD
    A[Fluid problem] --> B{Compressible?}
    B -->|M < 0.3| C[Incompressible]
    B -->|M >> 0.3| D[Compressible; check for shocks]
    C --> E{Reynolds?}
    E -->|Re << 1| F[Stokes flow: ∇P = μ∇²v]
    E -->|Re ~ 1-10³ | G[Laminar, viscous]
    E -->|Re >> 10³| H[Turbulent: Reynolds-avg / LES / K41]
    A --> I{Body force / rotation?}
    I -->|Gravity, stratification| J[Buoyancy, Schwarzschild]
    I -->|Coriolis dominant| K[Geostrophic / Taylor-Proudman]
    A --> L{Magnetic field?}
    L -->|Yes| M[MHD: Rm, vA, frozen-in]
    M --> N[Reconnection if topology changes]
    A --> O{Waves?}
    O -->|Sound| P[c_s, Mach]
    O -->|Surface| Q[ω²=gk tanh kh]
    O -->|Internal| R[N, dispersion]
    O -->|Alfvén| S[vA = B/√μ₀ρ]
    D --> T[Shocks: Rankine-Hugoniot]
    H --> U[K41 spectrum k^-5/3]
    J --> V[Ra > Ra_c → convection]

Comparison Tables

Dimensionless Numbers

Wave Modes Summary

Boundary Layer Profiles

Common Flow Geometries

Common Mistakes

• ❌ Applying Bernoulli across streamlines when flow isn't irrotational.
• ❌ Confusing $v_{\text{phase}}$ and $v_g$. Energy moves at $v_g$; phase fronts at $v_{\text{phase}}$.
• ❌ Using incompressible approximation for $M > 0.3$. Density variations matter.
• ❌ Ignoring viscosity at no-slip boundaries. Even at high Re, thin boundary layers carry essential physics (drag, separation).
• ❌ Forgetting Coriolis in large-scale flows. Earth weather/ocean require it; cyclones, gyres are geostrophic.
• ❌ Treating turbulence as random "noise." It has structure — coherent eddies, log law, $k^{-5/3}$ inertial range.
• ❌ Applying K41 outside the inertial range. Spectrum departs at injection scale (small $k$) and dissipation scale (large $k$).
• ❌ Using Stokes drag in non-Stokes regime. Reynolds-dependent corrections needed when Re > 1.
• ❌ Confusing gauge vs dynamical pressure. $P$ is set by EOS in compressible flow; in incompressible, $P$ is a Lagrange multiplier enforcing $\nabla\cdot\mathbf{v} = 0$ (solved from Poisson eq.).
• ❌ Bernoulli with viscosity. Bernoulli equation assumes ideal; in real flow there's head loss along the streamline.
• ❌ Assuming shocks are isentropic. They aren't — entropy jumps across the shock (2nd law).
• ❌ Neglecting $\nabla\cdot\mathbf{B} = 0$ in numerical MHD. Spurious monopoles destroy solutions; constrained-transport schemes essential.
• ❌ Computing MHD as fluid + force. $\mathbf{B}$ has its own evolution equation; not just a force on $\mathbf{v}$.
• ❌ Treating Alfvén waves as compressional. Pure Alfvén is incompressible — only fast/slow MS waves compress fluid.
• ❌ Forgetting baroclinic vorticity generation. Misaligned $\nabla\rho$ and $\nabla P$ create $\boldsymbol\omega$ — drives sea breeze, tornado vortex genesis.
• ❌ Hydrostatic with constant density for tall atmosphere. Use scale height $H$ for ideal-gas case.
• ❌ Confusing Rayleigh number critical values. Depends on boundary conditions (rigid vs free).

Expert Insights

Fluid dynamics is one continuum theory, with three dimensional numbers (Re, Ma, dimensionless gravity / rotation / magnetic) selecting regimes. Master the regime classification first; the equations are always the same.

Vorticity is a quasi-Lagrangian invariant. In ideal flow, vorticity lines move with the fluid, just like magnetic flux in ideal MHD. This is no coincidence — both are conserved-circulation theorems.

3-D turbulence cascades to small scales; 2-D turbulence cascades to large scales (inverse cascade). Atmospheric dynamics is approximately 2-D on large scales — hence huge weather systems.

The closure problem of turbulence will not be "solved." What we have is a hierarchy of approximations (RANS, LES, DNS), each appropriate to a range of Re. Choose your closure to match Re and required accuracy.

K41 is approximate but extraordinarily predictive. The $k^{-5/3}$ slope is observed from labs to galaxies, across 7+ decades in length scale. Intermittency corrections are small (10–20%).

Sound waves are linearized fluid waves; shocks are nonlinear sound — when compression steepens faster than it spreads.

Tsunamis are shallow-water waves. That's why they're so destructive — they carry their full energy at near-jet speeds and amplify near shore.

Rotation makes flows 2-D (Taylor-Proudman) at low Rossby. This is geophysical intuition: weather is approximately 2-D motion on a globe.

Ekman pumping is how surface winds drive ocean circulation. The Ekman layer is thin (10–100 m) but couples atmosphere to ocean.

Magnetic field amplification by stretching is universal. In dynamos (Earth, Sun, galaxies), in accretion disks (MRI), in plasma jets. Frozen-in flux + differential motion = amplification.

The MRI changed astrophysics in 1991. Before: no agreed mechanism for angular-momentum transport in accretion disks. After: weak $\mathbf{B}$ + differential rotation drives turbulence sufficient for accretion rates we observe.

Reconnection is the key non-ideal effect in MHD. Without it, frozen-in flux means no topology change ever, no energy release from $\mathbf{B}$. Solar flares, magnetic substorms, fusion sawteeth — all reconnection.

The Sun's surface granulation is convection viewed up close. Each granule = top of a convective cell, $\sim 1000$ km across, lasting $\sim 10$ minutes.

Pattern formation theory connects fluids to broader nonlinear science. Bénard cells, Liesegang rings, Turing patterns — all manifestations of universal amplitude equations.

The Navier-Stokes existence-and-smoothness problem ($10^6 prize from Clay Foundation) is unsolved. We don't know if smooth initial data evolve to a finite-time singularity in 3-D. (In 2-D, Leray proved smooth solutions exist for all time.)

In MHD, what matters is $\beta = P/(B^2/2\mu_0)$. $\beta \gg 1$: gas-dominated (most of ISM, dense plasmas). $\beta \ll 1$: field-dominated (solar corona). $\beta\sim 1$: equipartition (lots of interesting physics).

Bernoulli explains lift only partially. Real airfoil lift involves circulation generation (Kutta condition) — a vorticity story, not just pressure on streamlines.

Surface tension is a free-energy effect — molecules at a surface have fewer neighbors than in the bulk. The Young-Laplace equation is just minimization of surface energy at fixed volume.

Compressible-flow theory is one of the most quantitatively successful subfields of physics — Mach numbers in jet engines, shock standoffs in reentry, supersonic combustion are all predicted by the equations of Chapter 17 to engineering tolerances.

Troubleshooting

Cheatsheet

=== FOUNDATIONS ===
Material deriv:   D/Dt = ∂_t + v·∇
Continuity:       ∂_t ρ + ∇·(ρv) = 0
Euler:           ρ Dv/Dt = −∇P + ρg
Navier-Stokes:   ρ Dv/Dt = −∇P + μ∇²v + (ζ+μ/3)∇(∇·v) + ρg
Energy:          ρT Ds/Dt = Φ − ∇·q + Q
                 Φ = 2μ σ_ij σ_ij + ζ(∇·v)²
Bernoulli:       (1/2)v² + ∫dP/ρ + Φ = const (steady, ideal, barotropic)
Hydrostatic:     ∇P = ρg
Stokes drag:     F = 6πμaU
Poiseuille:      Q = πa⁴(−dP/dz)/(8μ)
Surface tension: ΔP = γ(1/R₁ + 1/R₂)
Capillary length: ℓ_c = √(γ/ρg)

Reynolds:    Re = ρvL/μ = vL/ν
Mach:        M = v/c_s
Froude:      Fr = v/√(gL)

=== VORTICITY ===
ω = ∇×v
Dω/Dt = (ω·∇)v − ω(∇·v) + ∇ρ×∇P/ρ² + ν∇²ω
Kelvin: dΓ/dt = 0 (ideal, barotropic, conservative force)
Rotating frame: + Coriolis −2Ω×v + centrif. −Ω×(Ω×r)
Coriolis param: f = 2Ωsinλ
Geostrophic:   −∇P/ρ = 2Ω×v
Taylor-Proudman: (Ω·∇)v = 0 at low Ro
Ekman δ = √(2ν/f)

=== TURBULENCE ===
Reynolds decomp: v = v̄ + v'
Reynolds stress: τ^R_ij = −ρ⟨v'_i v'_j⟩
Kolmogorov:
  η = (ν³/ε)^(1/4)
  v_η = (νε)^(1/4)
  τ_η = (ν/ε)^(1/2)
  L/η ~ Re^(3/4)
Spectrum:    E(k) = C_K ε^(2/3) k^(−5/3)
4/5 law:    ⟨(δv_∥)³⟩ = −(4/5) ε ℓ
Log law:     U⁺ = (1/κ) ln y⁺ + B, κ≈0.41, B≈5

=== WAVES ===
Sound:           ω = c_s k,  c_s = √(γP/ρ)
Surface gravity: ω² = gk tanh(kh) + (γ/ρ)k³ tanh(kh)
  Shallow:      ω² = gk² h, v = √(gh)
  Deep:         ω² = gk
Capillary min:   v = (4gγ/ρ)^(1/4)
Internal gravity: ω = N cos θ
Brunt-Väisälä:  N² = −(g/ρ)(dρ/dz)_ad
Rossby:         ω = −β k_x/(k² + κ²)

=== COMPRESSIBLE ===
c_s² = γP/ρ
M = v/c_s
Stagnation:
  T₀/T = 1 + (γ−1)M²/2
  P₀/P = [T₀/T]^(γ/(γ−1))
  ρ₀/ρ = [T₀/T]^(1/(γ−1))

Normal shock (γ=5/3):
  P₂/P₁ = 1 + 2γ(M₁²−1)/(γ+1)
  ρ₂/ρ₁ = (γ+1)M₁²/[(γ−1)M₁² + 2]
  M₂² = [(γ−1)M₁² + 2]/[2γM₁² − (γ−1)]
  Strong limit: ρ₂/ρ₁ → (γ+1)/(γ−1) = 4

De Laval throat: M = 1 at A_min (choked)
Sedov-Taylor:    R_s(t) = (ξEt²/ρ₀)^(1/5)

=== CONVECTION ===
Schwarzschild stable: −dT/dz < g/c_p
Adiabatic lapse:      Γ_ad = g/c_p ≈ 9.8 K/km
Boussinesq: ρ_eff(T) = ρ₀(1 − α ΔT) in buoyancy term only

Ra = gα ΔT d³/(νκ)
Pr = ν/κ
Nu = q_total/q_cond
Ra_c (rigid-rigid) = 1707.76
Critical k_c d ≈ 3.117  →  λ ≈ 2d

=== MHD ===
Lorentz density: (1/μ₀)(∇×B)×B = −∇(B²/2μ₀) + (B·∇)B/μ₀
Induction:       ∂_t B = ∇×(v×B) + η_m ∇²B
Magnetic press:   P_B = B²/2μ₀
β = P/P_B

Alfvén speed:    v_A = B/√(μ₀ρ)
Shear Alfvén:    ω = v_A k_∥
Fast/slow MS:    ω² = (c_s² + v_A²)/2  ± (1/2)√[(c_s² + v_A²)² − 4c_s²v_A²cos²θ] · k²

Magnetic Re:    Rm = vL/η_m
Lundquist:      S = Lv_A/η_m
Sweet-Parker:   v_in/v_A = 1/√S
Plasmoid:       v_in/v_A ~ 0.01 (high S)

Glossary

• Adiabatic — No heat exchange; $Ds/Dt = 0$ in ideal flow.
• Alfvén speed ($v_A$) — Propagation speed of magnetic-tension waves.
• Alfvén's theorem — Frozen-in flux for ideal MHD.
• Baroclinic — $\nabla\rho \times \nabla P \ne 0$; generates vorticity.
• Barotropic — $P = P(\rho)$ only; $\nabla\rho \times \nabla P = 0$.
• Bernoulli equation — Energy conservation along streamline in ideal flow.
• Boussinesq approximation — Density variations only in buoyancy term.
• Boundary layer — Thin region near wall where viscosity matters even at high Re.
• Brunt-Väisälä frequency ($N$) — Oscillation frequency for buoyancy in stratified medium.
• Capillary length — $\ell_c = \sqrt{\gamma/\rho g}$; gravity-capillary crossover scale.
• Choked flow — Mass flux saturates when $M=1$ at throat.
• Closure problem — Reynolds-averaged equations need unknown $\overline{v'_i v'_j}$.
• Compressible — $\nabla\cdot\mathbf{v} \ne 0$; density changes.
• Coriolis force — Pseudo-force $-2\boldsymbol\Omega\times\mathbf{v}$ in rotating frame.
• Couette flow — Linear-profile flow between moving plates.
• De Laval nozzle — Converging-diverging nozzle; produces supersonic flow.
• Dispersion relation — $\omega(k)$ for a wave; encodes propagation properties.
• Dynamo — Mechanism sustaining magnetic field via fluid motion.
• Ekman layer — Viscous boundary layer in rotating frame, thickness $\sqrt{2\nu/f}$.
• Eulerian description — Fields at fixed spatial points.
• Equation of state — Closes fluid system; e.g., $P(\rho, T)$.
• Free-surface — Boundary between fluid and lighter fluid (or vacuum), pressure-free.
• Frozen-in — Magnetic field lines move with conducting fluid.
• Geostrophic — Coriolis balances pressure gradient; geophysical flows.
• Group velocity ($v_g$) — Speed of energy / wavepacket propagation, $d\omega/dk$.
• Hagen-Poiseuille — Pressure-driven laminar flow in pipe; $\propto a^4$.
• Helmholtz vortex theorems — Vortex lines material; tubes have constant strength.
• Incompressible — $\nabla\cdot\mathbf{v} = 0$.
• Inertial range — Intermediate scales in turbulence where K41 applies.
• Internal waves — Buoyancy-driven waves in stratified medium.
• Inviscid — $\mu = 0$; Euler equations.
• Irrotational — $\boldsymbol\omega = 0$; admits velocity potential.
• K41 — Kolmogorov 1941 theory of turbulence.
• Kelvin's theorem — Circulation conserved on material loop in ideal flow.
• Knudsen number — $\ell/L$; continuum valid when Kn $\ll 1$.
• Kolmogorov scale ($\eta$) — Dissipation length scale in turbulence.
• Lagrangian description — Follow material elements.
• Lift — Force perpendicular to flow on a body with circulation (Kutta-Joukowski).
• Log law — Boundary-layer mean profile $U^+ \sim \ln y^+$.
• Lundquist number ($S$) — MHD analog of Re using $v_A$.
• Mach number ($M$) — $v/c_s$.
• Magnetic pressure / tension — Lorentz force decomposition.
• Magnetorotational instability (MRI) — Drives accretion-disk turbulence.
• Material derivative — $D/Dt = \partial_t + \mathbf{v}\cdot\nabla$.
• Mean free path ($\ell$) — Average distance between molecular collisions.
• Mixing length ($\ell_m$) — Phenomenological scale in convection / turbulence models.
• Navier-Stokes — Newtonian viscous incompressible (or compressible) fluid eqs.
• No-slip condition — Fluid velocity matches solid boundary.
• Nusselt number (Nu) — Ratio of total to conductive heat flux.
• Phase velocity ($v_p$) — $\omega/k$; speed of wave crests.
• Plasma $\beta$ — $P/(B^2/2\mu_0)$; gas vs magnetic pressure.
• Poiseuille flow — Pressure-driven pipe flow.
• Polytropic — $P\propto\rho^\gamma$.
• Potential flow — Irrotational + incompressible; $\nabla^2\phi = 0$.
• Prandtl number (Pr) — $\nu/\kappa$.
• Rayleigh-Bénard convection — Heated-from-below thermal convection.
• Rayleigh number (Ra) — Buoyancy / diffusion ratio.
• Reconnection — Magnetic-topology change via local non-ideal effects.
• Reynolds decomposition — Mean + fluctuation: $\mathbf{v} = \bar{\mathbf{v}} + \mathbf{v}'$.
• Reynolds number (Re) — $vL/\nu$.
• Reynolds stress — Apparent stress from turbulent fluctuations.
• Riemann invariants — Quantities propagating along characteristics in 1-D hyperbolic flow.
• Rossby number — Inertia/Coriolis ratio.
• Rossby waves — Vorticity waves from gradient of Coriolis parameter.
• Schwarzschild criterion — Stability of stratification: lapse rate vs adiabatic.
• Sedov-Taylor — Self-similar blast wave from point explosion.
• Shock — Discontinuity from converging characteristics; entropy increases across.
• Sound speed ($c_s$) — $\sqrt{(\partial P/\partial\rho)_S}$.
• Stagnation properties — Reversibly bringing flow to rest; $T_0, P_0, \rho_0$.
• Stokes flow — Re → 0 limit; linear, reversible, scallop theorem.
• Streamline — Curve everywhere tangent to $\mathbf{v}$; in steady flow = particle path.
• Surface tension ($\gamma$) — Free energy per unit area of interface.
• Sweet-Parker reconnection — Resistive sheet reconnection: rate $1/\sqrt{S}$.
• Taylor-Proudman theorem — Rotation makes slow flow 2-D.
• Tensor stress — In fluid: pressure + viscous strain-rate × viscosity.
• Tsunami — Long-wavelength shallow-water surface wave.
• Turbulence — Chaotic, multiscale, statistically self-similar nonlinear flow.
• Velocity potential ($\phi$) — Exists if irrotational; $\mathbf{v} = \nabla\phi$.
• Viscosity — $\mu$ (dynamic) or $\nu = \mu/\rho$ (kinematic).
• Vortex tube — Bundle of vortex lines; strength = $\int\boldsymbol\omega\cdot d\mathbf{S}$.
• Vorticity ($\boldsymbol\omega$) — $\nabla\times\mathbf{v}$.

Final Takeaways

1. Three conservation laws + EOS + constitutive (viscosity) = all of fluid dynamics. Everything is regime-selection on this single set.
2. The material derivative is the bridge between Eulerian fields and Lagrangian particle dynamics. Master it before everything else.
3. Reynolds number rules. Re determines whether you're in Stokes (linear, reversible), laminar, or turbulent (chaotic, dissipative) regime.
4. Bernoulli is powerful but conditional. Steady, ideal, barotropic, along a streamline. Otherwise pick another tool.
5. Vorticity is the protagonist of incompressible flow. Kelvin's theorem makes it nearly conserved; vortex stretching is the 3-D effect distinguishing turbulence from waves.
6. Turbulence has universal small-scale statistics (Kolmogorov $k^{-5/3}$) but large-scale structure depends on geometry. Closure modeling is the practical art.
7. Compressibility introduces shocks — discontinuities that conserve mass/momentum/energy but increase entropy. Rankine-Hugoniot is non-negotiable.
8. Rotation makes flows two-dimensional. Geostrophic, Taylor-Proudman, Rossby — geophysical fluid mechanics built on this.
9. Buoyancy + diffusion = convection; governed by Rayleigh number. From Bénard cells to Solar granules to galactic interstellar medium.
10. MHD = fluid dynamics + Maxwell. Frozen-in field is the master concept; reconnection is the master non-ideal effect.
11. The fluid equations connect virtually every later Part: plasma (microscopic basis of MHD), GR (relativistic hydrodynamics), elastodynamics (acoustic limit), kinetic theory (closure-from-below).
12. Most "outstanding problems" in classical physics live here: turbulence closure, reconnection rates, dynamos, NS smoothness. Fluid dynamics is the last frontier of 19th-century physics — still surprising us in the 21st.
13. Dimensional analysis is your friend. Re, Ma, Fr, Ra, Rm — these numbers determine which terms in the equations dominate. Never compute without first identifying the regime.
14. Symmetries enforce structure. Galilean invariance + Newton + thermo gives N-S; rotation gives Coriolis; isotropy + locality + dissipation gives K41. The whole field is symmetries + closures.

Next: Part VI — Plasma Physics. Microscopic origin of fluid (Vlasov, Boltzmann), wave-particle resonances, Landau damping, instabilities, nonlinear regime. Foundation of fusion, ionospheric, astrophysical plasmas.