3.5 Prandtl-Meyer expansion waves

Prandtl-Meyer expansion theory

Prandtl-Meyer expansion waves describe how supersonic flow speeds up and changes direction when it expands around a convex corner or through a diverging nozzle. They're one of the few cases in compressible aerodynamics where you get an exact analytical solution, which makes them foundational for designing supersonic nozzles, analyzing flow over airfoils, and building intuition about high-speed flow behavior.

The entire theory rests on three assumptions: the flow is isentropic (no entropy change), irrotational (no vorticity), and inviscid (no viscosity). These are idealizations, but they hold well enough in many practical supersonic flows to give accurate predictions.

Prandtl-Meyer function

The Prandtl-Meyer function $\nu(M)$ represents the total angle through which a flow must turn to accelerate from Mach 1 to a given Mach number $M$. It's the central tool for all expansion wave calculations.

$\nu(M) = \sqrt{\frac{\gamma+1}{\gamma-1}} \tan^{-1} \sqrt{\frac{\gamma-1}{\gamma+1}(M^2-1)} - \tan^{-1} \sqrt{M^2-1}$

Here $\gamma$ is the specific heat ratio (1.4 for air). A few reference values worth remembering:

• At $M = 1$: $\nu = 0°$ (by definition, since the function measures turning from Mach 1)
• At $M = 2$: $\nu \approx 26.4°$
• At $M \to \infty$: $\nu_{\max} = \frac{\pi}{2}\left(\sqrt{\frac{\gamma+1}{\gamma-1}} - 1\right) \approx 130.5°$ for air

Tabulated values of $\nu(M)$ appear in most compressible flow textbooks and are the fastest way to solve problems on exams.

Mach number vs turning angle

The relationship between the turning angle $\theta$ and the upstream/downstream Mach numbers is straightforward:

$\theta = \nu(M_2) - \nu(M_1)$

where $M_1$ is the Mach number before the expansion and $M_2$ is the Mach number after. Because $\nu$ increases monotonically with $M$, a larger turning angle always means a larger increase in Mach number.

Typical problem-solving steps:

1. Look up (or compute) $\nu(M_1)$ from the table or formula.
2. Add the turning angle: $\nu(M_2) = \nu(M_1) + \theta$.
3. Use the table (or invert the formula) to find $M_2$.
4. Apply isentropic relations to get downstream pressure, temperature, and density.

The maximum possible turning angle occurs when $M_2 \to \infty$, which corresponds to the flow expanding into a vacuum. In practice, you'll never reach this limit, but it sets an upper bound on how much a supersonic flow can turn through expansion.

Centered expansion waves

When supersonic flow hits a sharp convex corner, a centered expansion fan forms at the corner. This fan consists of an infinite number of Mach waves radiating outward from the corner point, each one turning the flow by an infinitesimal amount.

The fan is bounded by two Mach waves:

• The leading wave at angle $\mu_1 = \sin^{-1}(1/M_1)$ relative to the upstream flow direction
• The trailing wave at angle $\mu_2 = \sin^{-1}(1/M_2)$ relative to the downstream flow direction

Across the fan, pressure, density, and temperature all decrease smoothly and continuously. This is the opposite of an oblique shock, where properties jump discontinuously.

Characteristics of the expansion fan

• Isentropic: Entropy stays constant across the entire fan, so stagnation pressure is preserved.
• Irrotational: The velocity field has zero curl and can be described by a potential function.
• Reversible: In principle, the flow could be compressed back to its original state by passing through a symmetric convergent section.
• Uniform downstream flow: After the fan, the flow is parallel to the downstream surface with a single, well-defined Mach number $M_2 > M_1$.

The contrast with shock waves is worth emphasizing: shocks are irreversible and increase entropy, while expansion fans are isentropic. This is why expansions are "friendly" in supersonic flow design and shocks are something you try to minimize.

Supersonic flow over convex corners

When supersonic flow meets a convex corner (the surface turns away from the flow), an expansion fan forms to smoothly redirect and accelerate the flow. Unlike oblique shocks at concave corners, where properties change abruptly, expansion fans produce gradual, continuous changes in all flow properties.

Expansion wave geometry

The Mach angle $\mu$ at any point in the flow is:

$\sin \mu = \frac{1}{M}$

Since $M$ increases across the fan, $\mu$ decreases. This means the Mach waves tilt more steeply near the leading edge (lower Mach) and become shallower near the trailing edge (higher Mach).

The total turning angle is still governed by:

$\theta = \nu(M_2) - \nu(M_1)$

For more complex geometries (curved walls, interacting fans), the method of characteristics traces individual Mach wave paths through the flow field to construct the complete solution.

Mach wave angle

The Mach wave angle $\mu$ deserves extra attention because it controls the physical width of the expansion fan:

• At $M = 1$: $\mu = 90°$ (Mach wave is perpendicular to the flow)
• At $M = 2$: $\mu = 30°$
• At $M = 5$: $\mu \approx 11.5°$
• As $M \to \infty$: $\mu \to 0°$

Higher Mach numbers produce narrower fans. At the theoretical vacuum limit, the fan collapses to a single line.

Weak vs strong expansion waves

• Weak expansion waves involve small turning angles and produce modest increases in Mach number. The flow properties change only slightly across the fan.
• Strong expansion waves involve large turning angles, producing significant Mach number increases and large drops in pressure, temperature, and density.

The strength depends on two things: the geometry of the corner (how sharply the surface turns) and the initial Mach number. A flow already at high Mach will experience a smaller relative change in Mach number for the same turning angle compared to a flow just above Mach 1.

Isentropic expansion process

Because the expansion fan is isentropic, all the standard isentropic flow relations apply. Stagnation properties remain constant through the fan, so you can calculate local conditions at any point if you know the local Mach number.

Compressible flow equations

These three relations connect local static properties to stagnation (total) properties through the Mach number:

Pressure ratio: $\frac{p}{p_0} = \left(1 + \frac{\gamma-1}{2}M^2\right)^{-\frac{\gamma}{\gamma-1}}$

Density ratio: $\frac{\rho}{\rho_0} = \left(1 + \frac{\gamma-1}{2}M^2\right)^{-\frac{1}{\gamma-1}}$

Temperature ratio: $\frac{T}{T_0} = \left(1 + \frac{\gamma-1}{2}M^2\right)^{-1}$

Here $p_0$, $\rho_0$, and $T_0$ are the stagnation pressure, density, and temperature. As $M$ increases through the expansion, all three static quantities ($p$, $\rho$, $T$) decrease.

Stagnation properties in expansion

Stagnation properties are what you'd measure if you brought the flow to rest isentropically. The key point for expansion fans: stagnation pressure, density, and temperature are all constant across the fan, because the process is isentropic.

This is a major practical advantage. If you know the stagnation conditions (from, say, a reservoir upstream) and the local Mach number, you can immediately compute every local flow property using the equations above.

Compare this to flow through a shock, where stagnation pressure drops due to entropy generation. Expansion fans preserve stagnation pressure, which is why well-designed supersonic nozzles rely on expansion rather than shocks to accelerate the flow.

Expansion ratio

The area-Mach number relation for isentropic flow connects the local cross-sectional area $A$ to the critical area $A^*$ (the area where $M = 1$):

$\frac{A}{A^*} = \frac{1}{M}\left[\frac{2}{\gamma+1}\left(1+\frac{\gamma-1}{2}M^2\right)\right]^{\frac{\gamma+1}{2(\gamma-1)}}$

This relation is most directly relevant in nozzle flows rather than corner flows, but the underlying physics is the same: a larger area ratio corresponds to a higher Mach number on the supersonic branch of the solution. For a given initial Mach number and turning angle, the expansion ratio determines the final Mach number and all downstream properties.

Applications of Prandtl-Meyer flow

Supersonic nozzle design

Supersonic nozzles use carefully shaped contours to create a series of expansion waves that gradually accelerate the flow to the desired exit Mach number. The nozzle wall is essentially a sequence of convex corners (or a smooth curve approximating them), and Prandtl-Meyer theory dictates the turning angle needed at each point.

The design goal is to produce uniform, parallel flow at the nozzle exit with minimal losses. This requires canceling any reflected waves inside the nozzle so that the exit flow is shock-free. Rocket engines, supersonic wind tunnels, and high-speed jet engines all rely on this approach.

Minimum length nozzles

A minimum length nozzle achieves the target exit Mach number in the shortest possible axial distance. The design uses a sharp throat followed by an aggressive initial expansion, creating a strong centered expansion fan right at the throat.

The downstream contour is then shaped to cancel the reflected waves and straighten the flow. These nozzles are compact, which makes them attractive for applications with tight space constraints like small wind tunnel test sections or upper-stage rocket engines. Prandtl-Meyer theory combined with the method of characteristics is the standard approach for computing the wall contour.

Thrust vectoring

Thrust vectoring controls the direction of engine thrust to steer a vehicle. One approach injects a secondary flow through ports in the nozzle wall, creating local expansion (and sometimes shock) waves that deflect the primary exhaust stream.

By adjusting the injection location and mass flow rate, engineers can control the deflection angle and thus the thrust direction. Prandtl-Meyer theory helps predict how much the primary flow will turn in response to a given injection configuration, which is essential for designing the control system.

Numerical methods for expansion waves

Prandtl-Meyer theory gives clean analytical solutions for simple geometries (sharp corners, straight walls), but real-world problems often involve curved surfaces, interacting wave systems, and non-ideal effects. Numerical methods handle these complexities.

Method of characteristics

The method of characteristics (MOC) is tailor-made for supersonic flow. It works by tracing characteristic lines (Mach waves) through the flow field and computing properties at their intersections.

How it works:

1. Identify the initial data line where flow properties are known (e.g., the nozzle throat or upstream uniform flow).
2. From each point, draw two characteristic lines at angles $\pm \mu$ relative to the local flow direction.
3. At each intersection of characteristics, use the compatibility equations (derived from the governing PDEs) to solve for the local Mach number and flow angle.
4. March downstream, building up the complete flow field point by point.

MOC naturally captures expansion fans because the characteristic lines follow the Mach wave angles. It's the standard tool for designing supersonic nozzle contours.

Finite difference schemes

Finite difference methods discretize the flow domain onto a grid and approximate the partial derivatives in the Euler equations using algebraic differences between neighboring grid points.

• Explicit schemes calculate new values directly from known values at the previous step. They're simple to implement but have stability limits on the time step (CFL condition).
• Implicit schemes solve a coupled system of equations at each step. They're more computationally expensive per step but allow larger step sizes.

These methods are more versatile than MOC and can handle mixed subsonic-supersonic flows, complex geometries, and viscous effects when needed.

Boundary conditions for expansion

Correct boundary conditions are critical for getting meaningful numerical results:

• Inflow: Specify the upstream Mach number, pressure, density, and flow direction.
• Outflow: For supersonic exit flow, all information travels downstream, so no conditions need to be imposed at the exit (characteristic boundary condition).
• Solid walls: The flow must be tangent to the wall (zero normal velocity). For inviscid simulations, wall pressure comes from the Prandtl-Meyer relations. For viscous simulations, the no-slip condition (zero velocity at the wall) replaces the tangency condition.

Getting these wrong, especially at outflow boundaries, is a common source of numerical errors in supersonic flow simulations.

Limitations of Prandtl-Meyer theory

Assumptions and simplifications

Prandtl-Meyer theory assumes:

• Steady, inviscid, adiabatic flow (no time dependence, no viscosity, no heat transfer)
• Ideal gas with constant $\gamma$
• Irrotational, isentropic flow (no shocks or entropy-producing mechanisms)

These assumptions hold reasonably well for many supersonic flows at moderate Mach numbers and away from walls where boundary layers are thin. They start to break down when viscous effects, heat transfer, or extreme temperatures become significant.

Real gas effects

At very high Mach numbers (roughly $M > 5$, the hypersonic regime), temperatures become high enough that the ideal gas assumption fails. Real gas effects include:

• Variable specific heats: $\gamma$ is no longer constant as vibrational energy modes of the gas molecules become excited.
• Chemical reactions: At sufficiently high temperatures, air molecules dissociate ($\text{O}_2 \to 2\text{O}$, $\text{N}_2 \to 2\text{N}$) and can even ionize.
• Non-equilibrium effects: The gas may not reach thermodynamic equilibrium quickly enough, so the flow chemistry depends on the flow history.

These effects alter the expansion process and can produce significant deviations from the standard Prandtl-Meyer predictions.

Viscous and heat transfer effects

In real flows, boundary layers grow along surfaces and can interact with expansion fans. Viscous effects that Prandtl-Meyer theory ignores include:

• Boundary layer displacement: The effective wall shape differs from the physical wall, modifying the expansion geometry.
• Flow separation: If the boundary layer separates (due to adverse pressure gradients elsewhere in the flow), the entire expansion pattern changes.
• Heat transfer: Significant wall heating or cooling alters the temperature profile near the surface and can affect the expansion.

For flows where these effects matter, computational fluid dynamics (CFD) with turbulence models and real-gas thermodynamics replaces the analytical Prandtl-Meyer approach.