Aerodynamics — Engineering Reference

1. At a glance

Aerodynamics is the engineering science of air flow around bodies and through ducts. It is the specialization of fluid-mechanics that handles the regimes most relevant to flight, propulsion, and high-speed vehicles: from the slow incompressible flow over a low-speed glider, through the subsonic cruise of an airliner, the transonic shock-bound flow over a jet wing, the supersonic regime of fighters and missiles, up to hypersonic re-entry where the air itself dissociates and ionizes. The discipline splits along Mach number rather than geometry:

• Low-speed / incompressible (M < 0.3) — gliders, light aircraft, wind turbines, ground vehicles, sports projectiles, buildings, UAVs.
• Subsonic compressible (0.3 < M < 0.7) — turboprops, regional jets at cruise, large-fan aero-engines.
• Transonic (0.7 < M < 1.2) — modern airliners, business jets, transport aircraft.
• Supersonic (1.2 < M < 5) — military fighters (F-15, F-22), missiles (AMRAAM, BrahMos), Concorde, SR-71.
• Hypersonic (M > 5) — re-entry vehicles (Apollo CM, Shuttle, Dragon, Starship), scramjets (X-43, X-51), ballistic-missile RVs.

Every aircraft is designed by people who decide first about aerodynamics and then about everything else. Wing planform, fuselage cross-section, control-surface area, engine inlet geometry — these decisions cascade into structures (loads, flutter), propulsion (thrust required, fuel volume), flight controls (stability, handling qualities), and operations (takeoff field length, cruise range). The design-stack placement is mission analysis → aerodynamics → propulsion sizing → weights/structures → flight controls → certification.

2. Why it matters

Drag costs fuel. Lift carries payload. Flutter kills airplanes. Stall makes them fall. Compressibility creates shocks that double or triple drag. Re-entry vehicles see stagnation temperatures above 1500 K. Three classes of error are universal in early-stage aircraft design: (1) under-estimated drag, so the engines can’t push the airplane to the design speed and the certified range collapses; (2) under-estimated C_L_max with high-lift devices deployed, so the certified takeoff and landing speeds exceed the runway capability; (3) ignored compressibility effects, so the airframe encounters drag-divergence or buffet years before the marketing brochure said it would. A correctly executed aerodynamic analysis at the configuration stage — drag polar, lift curve, stability derivatives, compressibility correction, high-lift performance — is short. But it has to be right, because the airplane is already half-built around its assumptions.

The economic stakes are large. A 1 % drag reduction on a long-haul widebody is worth roughly $0.5–1 M per aircraft per year in fuel. Boeing’s 787 wingtip raked extensions, Airbus’s A350 sharklet program, the McDonnell Douglas / Boeing winglet retrofits on the 737NG — each is an aerodynamics-driven product decision whose return is measured in tens of billions across a global fleet.

3. First principles

Aerodynamics builds directly on the conservation laws of fluid-mechanics (continuity, momentum / Navier–Stokes, energy) plus an equation of state for the gas. For air at engineering conditions the ideal-gas law P = ρRT (R = 287 J/(kg·K), γ = c_p/c_v = 1.4) is adequate up to about T = 1500 K; above that, real-gas chemistry (vibrational excitation, dissociation, ionization) becomes important and a hypersonic-equilibrium or non-equilibrium model is required.

International Standard Atmosphere (ISA, ICAO Doc 7488)

Aerodynamic performance is always quoted against a reference atmosphere; for civil aviation that is the ICAO ISA. Selected values:

Altitude	T (K)	P (kPa)	ρ (kg/m³)	c (m/s)	µ (×10⁻⁵ Pa·s)
Sea level	288.15	101.325	1.2250	340.3	1.789
5 000 ft (1.52 km)	278.4	84.31	1.0551	334.4	1.741
10 000 ft (3.05 km)	268.3	69.68	0.9046	328.3	1.692
20 000 ft (6.10 km)	248.5	46.56	0.6526	316.0	1.594
30 000 ft (9.14 km)	228.7	30.09	0.4585	303.2	1.493
35 000 ft (10.67 km)	218.8	23.84	0.3796	296.5	1.442
40 000 ft (12.19 km)	216.65	18.75	0.3016	295.1	1.421
50 000 ft (15.24 km)	216.65	11.60	0.1865	295.1	1.421
65 000 ft (19.81 km)	216.65	5.529	0.0889	295.1	1.421
82 000 ft (25.00 km)	221.55	2.549	0.0401	298.4	1.448

Tropopause sits at 11 km (36 089 ft); above it the temperature is constant at 216.65 K until 20 km. ISA+15 and ISA+20 are hot-day offsets used for hot-day certification performance (Denver in summer, Middle Eastern airports).

Governing dimensionless groups

• Reynolds number Re = ρVL/µ. Sets the relative importance of inertia and viscosity, governs boundary-layer transition and skin friction.
• Mach number M = V/c, c = √(γRT). Sets the role of compressibility. M < 0.3 → incompressible; 0.3 < M < 0.8 → subsonic compressible; 0.8 < M < 1.2 → transonic; 1.2 < M < 5 → supersonic; M > 5 → hypersonic.
• Knudsen number Kn = λ/L. Governs rarefied / free-molecular flow at very high altitude (above ~80 km).
• Prandtl number Pr ≈ 0.72 for air. Sets thermal-boundary-layer thickness relative to velocity boundary layer.

Re-vs-M regime map

Vehicle class	Cruise V (m/s)	Cruise alt	Re (chord)	M_cruise	Dominant physics
Insect / MAV	1–10	sea level	10² – 10⁴	< 0.05	Viscous, low-Re separation bubbles
Hand-launched UAV	15–30	< 500 m	10⁵ – 5×10⁵	0.05	Laminar separation, transition critical
Light GA (Cessna 172)	60	5 000 ft	4×10⁶	0.18	Fully turbulent BL, incompressible
Sailplane	30–55	varies	1–3×10⁶	0.09–0.17	Laminar-flow buckets, low Re_critical
Turboprop regional	160	25 000 ft	2×10⁷	0.50	Subsonic compressible
Narrowbody jet (737/A320)	240	35 000 ft	3×10⁷	0.78	Transonic, supercritical wing
Widebody jet (777/787/A350)	250	39 000 ft	5×10⁷	0.85	Transonic, shock-bound upper surface
Supersonic fighter (F-22, M=2)	600	50 000 ft	3×10⁷	2.0	Oblique shocks, expansion fans
Concorde / SR-71 cruise	600 / 980	55 000 ft	5×10⁷	2.0 / 3.2	Wave drag dominant, fuselage area-ruled
Scramjet (X-43A, X-51)	2 500–3 200	30 000 m	10⁷	7–10	Real-gas chemistry, viscous interaction
Re-entry capsule (Apollo)	11 000 → 0	80 → 0 km	varies	36 → 0	Stagnation heating, dissociation

2D vs 3D distinction

An airfoil is the 2D cross-section of a wing. Its properties (c_ℓ, c_d, c_m) are per unit span, depend on α, Re, M, and surface condition. A wing is the 3D shape with finite span; its coefficients (C_L, C_D, C_M) include induced-drag and three-dimensional effects (downwash, tip vortices, wash-in/wash-out). Confusing the two is the most common first-year error in aircraft design.

Lift and drag definitions

L = ½ · ρ · V² · S · C_L
D = ½ · ρ · V² · S · C_D
M = ½ · ρ · V² · S · c̄ · C_M     (pitching moment; c̄ = mean aerodynamic chord)

S is planform area (wing area for wings, frontal area for bluff bodies). The dynamic pressure q = ½ρV² is the universal scaling. Coefficients are dimensionless and only weakly Re-dependent at moderate to high Re — this is what makes wind-tunnel testing useful.

Drag bookkeeping

Industry convention (per AGARD AR-256 and AIAA standard practice) decomposes total drag into a small set of additive components — useful both for design trade studies and for assigning responsibility for fixes:

Component	Symbol	Physical origin	Typical share, transport cruise
Friction drag	C_D,f	Skin friction over all wetted area	45–55 %
Pressure (form) drag	C_D,p	Pressure-recovery loss in turbulent wake	5–15 %
Induced drag	C_D,i	Vortex sheet downwash on finite wing	35–45 %
Wave drag	C_D,w	Shock losses above M_DD	0–5 %
Interference drag	C_D,int	Junctions (wing-fuselage, pylon, nacelle)	2–5 %
Trim drag	C_D,trim	Horizontal-tail download contribution	2–4 %
Excrescence / roughness	C_D,exc	Rivets, seams, gaps, antennas, drains	3–6 %
Cooling / bleed / leakage	C_D,cool	Internal mass-flow losses through aircraft	1–3 %

The classical preliminary-design split is C_D = C_D,0 + k·C_L². The constants are usually back-fit from the full breakdown above against measured polars.

4. Subsonic / low-speed aerodynamics

Thin-airfoil theory (Munk 1922, Glauert 1926)

For a thin, cambered airfoil in incompressible flow:

c_ℓ = 2π (α − α_0)            (lift slope = 2π per radian = 0.110 per degree)

α_0 is the zero-lift angle (negative for positive camber). The theory predicts the lift slope exactly; real airfoils achieve about 5–6 per radian (≈ 0.10 per degree) because of viscous boundary-layer thickening. Real airfoils also have a finite c_ℓ_max (≈ 1.2–1.6 for a clean section at Re ≈ 10⁶) set by leading-edge separation or trailing-edge stall.

Airfoil families:

Family	Era / origin	Key feature	Example aircraft
NACA 4-digit	NACA, 1933	Simple algebraic camber + thickness; e.g. 2412	Cessna 152/172, J-3 Cub
NACA 5-digit	NACA, 1935	Reflexed camber; higher c_ℓ_max	DC-3, Beech Bonanza
NACA 6-series	NACA, 1940s	Laminar-flow bucket; e.g. 65₂-415	P-51 Mustang, A-4 Skyhawk
NASA LS(1)	NASA Langley, 1970s	High c_ℓ_max for GA; e.g. LS(1)-0417	Piper Tomahawk, Beechcraft Skipper
Supercritical	Whitcomb / NASA, 1965+	Flat top, rear-loaded; high M_DD	737NG/MAX, 777, 787, A320, A350
Eppler / Wortmann FX	Eppler 1960s, Wortmann 1970s	Sailplane-optimized	ASW-27, Discus, Ventus
Selig / Drela	UIUC / MIT 1990s+	Low-Re UAV / pico-AC	MQ-9 sensor pods, MIT Daedalus

Pressure distribution on the upper surface drives c_ℓ_max — leading-edge suction peak height and pressure-recovery gradient govern separation onset.

Finite-wing theory — Prandtl lifting-line (1918)

A finite wing sheds a trailing vortex sheet that induces a downwash velocity, tilting the local flow vector and producing induced drag. For an elliptically-loaded wing (Spitfire planform, ideal case):

C_L,wing = a₀ · α_eff
α_eff   = α_geom − C_L/(π·AR)        (downwash correction)
C_L     = (a₀ · α)/(1 + a₀/(π·AR))   (3D lift slope)
C_D,i   = C_L²/(π·AR·e)              (induced drag)

with aspect ratio AR = b²/S and Oswald efficiency factor e ≈ 0.7–0.95 (1.0 for elliptical loading, lower for tapered or swept wings). The total drag splits as parasite (zero-lift) + induced:

C_D = C_D,0 + k · C_L²,   k = 1/(π·AR·e)

This is the drag polar — the single most-used equation in airplane preliminary design. Range and endurance maximize at distinct C_L points on the polar:

Range (jet, Breguet):           V·C_L/C_D maximum  →  C_L_opt = √(C_D,0 / (3k))
Endurance (jet):                C_L/C_D maximum    →  C_L_opt = √(C_D,0 / k)
Range (prop, Breguet):          C_L/C_D maximum    →  C_L_opt = √(C_D,0 / k)
Endurance (prop):               C_L^1.5/C_D maximum→  C_L_opt = √(3·C_D,0 / k)

For airliners the cruise C_L is typically chosen 5–10 % below (C_D,0/(3k))^0.5 to leave margin for buffet onset; gliders pick (C_D,0/k)^0.5 exactly for maximum L/D and minimum sink polar tangent.

High-lift devices

A clean airfoil’s c_ℓ_max ≈ 1.5 is insufficient for safe takeoff and landing. Flaps and slats raise it dramatically:

Device	Δc_ℓ_max	Examples
Plain flap	+0.5	Piper Cub, basic GA trainers
Split flap	+0.7	DC-3, early WWII fighters
Single-slotted Fowler	+1.1	737, A320, most regional jets
Double-slotted Fowler	+1.6	727, 757, A300, MD-80
Triple-slotted Fowler	+1.9	747-classic, A380
Leading-edge slat (Krueger)	+0.4	737NG/MAX, 757, A320 inboard
Leading-edge slat (slotted)	+0.6	747, 777, A350, A380
Boundary-layer suction	+1.5	F-104 (BLC), research aircraft
Blown / coanda flap	+2.0	YC-14, NASA QSRA, AG-2

A modern airliner deploying full slats + triple-slotted Fowler reaches landing C_L_max ≈ 2.8–3.0, which is what lets a 240-tonne 777 land at 140 kt.

Wing planform parameters

• Aspect ratio AR = b²/S. Sailplanes 20–35, airliners 8–11, fighters 3–5, delta wings 2–3.
• Taper ratio λ = c_tip/c_root. Typically 0.25–0.4 for airliners; reduces root bending moment and induced drag near elliptical optimum.
• Sweep Λ. 25–35° for airliners (transonic), 40–60° for supersonic fighters.
• Dihedral Γ. 1–7° for stability; anhedral (negative dihedral) on high-wing transports (C-5, An-124) to balance the destabilizing wing-position effect.
• Twist (washout). Tip incidence less than root by 2–5° to delay tip stall and approach elliptical loading.

Stall types

Different airfoil-thickness and Re combinations produce qualitatively different stall behavior. Knowing which type a section exhibits is a key go/no-go for handling-qualities risk.

Stall type	Typical airfoils / Re	Behavior
Trailing-edge stall	Thick (> 14 %) sections, Re > 10⁶	Gradual; separation creeps forward from TE
Leading-edge stall	Moderate-thickness (6–14 %), Re ≈ 10⁵–10⁶	Abrupt; short bubble at LE bursts catastrophically
Thin-airfoil stall	Thin sections (< 6 %), sharp LE	Long bubble that gradually lengthens; gentle
Combined	Many real wings	Both mechanisms compete; depends on Re

T-tail aircraft with thick wing roots are particularly prone to deep stall because the high tailplane sits in the wing wake at high α (BAC 1-11 fatal prototype crash, 1963).

Boundary-layer transition

Smooth flat plate transition Re_x ≈ 5 × 10⁵ to 3 × 10⁶ depending on free-stream turbulence and pressure gradient. Natural laminar flow (NLF) airfoils maintain laminar BL to 60–70 % chord, achieving c_d,0 ≈ 0.004 vs 0.008 for fully turbulent. Practical NLF requires meticulous surface finish (≤ 5 µm waviness) — bug strikes and rain trip transition immediately. Used on sailplanes, the Cirrus SR22, business jets (Honda HA-420). Hybrid Laminar Flow Control (HLFC) uses suction through leading-edge perforations to delay transition past sweep-induced cross-flow instability — flown on the Boeing 787-9 vertical tail and the A340-300 HLFC fin demonstrator.

5p. Transonic regime and compressibility

Critical and drag-divergence Mach numbers

M_crit is the free-stream Mach number at which the local flow first reaches M = 1 somewhere on the body. Below M_crit the flow is everywhere subsonic. Above M_crit, supersonic pockets form on the upper surface, terminated by a recompression shock that thickens the boundary layer.

M_DD (drag-divergence) is the free-stream Mach at which C_D rises by 0.0020 above its subsonic value (Boeing definition) or where dC_D/dM = 0.10 (DATCOM definition). M_DD typically lies 0.05–0.08 above M_crit.

Prandtl–Glauert compressibility correction

For thin airfoils in subsonic flow up to M ≈ 0.7:

C_p = C_p,inc / √(1 − M²)
C_L = C_L,inc / √(1 − M²)

This is the Prandtl–Glauert rule (Prandtl 1922, Glauert 1928). Above M ≈ 0.7 the linearization breaks down and full nonlinear transonic small-disturbance theory or Euler/Navier–Stokes CFD is required.

Supercritical airfoils (Whitcomb 1965)

Conventional airfoils peak in suction near 25 % chord; the supersonic pocket grows rapidly with Mach. Whitcomb’s supercritical airfoil flattens the upper surface (low curvature, broad weak supersonic pocket) and adds rear loading via a strongly cambered aft lower surface. M_DD shifts from ≈ 0.74 (1960s wings) to ≈ 0.83 (modern 787 / A350 sections) at the same t/c. Every modern airliner wing uses a supercritical section.

Area rule (Whitcomb 1952)

At transonic Mach, wave drag is minimized when the aircraft’s cross-sectional area distribution S(x) along the longitudinal axis matches the Sears–Haack body — smooth and bell-shaped. The Convair F-102 Delta Dagger could not exceed M = 1 in level flight until its fuselage was waisted (“Coke-bottled”) at the wing position; redesigned as the F-102A it cleanly accelerated through Mach 1. Subsequently applied to the F-106, T-38, B-1, F-14, and every supersonic fighter since.

Swept-wing principle (Busemann 1935, Jones 1945)

For a wing swept at angle Λ, the velocity component normal to the leading edge is V·cos Λ. If V·cos Λ < V·M_crit_unswept, no shock forms even though V > V_crit. Practical airliner sweep is 25–35°; pushing further loses C_L_max and complicates structures. Aft sweep is standard; forward sweep (X-29, Su-47) gives aeroelastic divergence problems and is unused in production.

6p. Supersonic aerodynamics

Above M = 1, disturbances cannot propagate upstream; the flow field is bounded by Mach lines at angle µ = arcsin(1/M) to the local flow direction. Compression turns produce oblique shocks (Rankine–Hugoniot jumps in P, ρ, T); expansion turns produce smooth Prandtl–Meyer expansion fans.

Shock-expansion theory and Ackeret linear theory

For a thin 2D wedge or diamond airfoil at small α:

Ackeret (linear, M > 1, thin):
   c_p = ± 2θ / √(M² − 1)        (compression / expansion)
   c_ℓ = 4α / √(M² − 1)
   c_d,wave = 4α² / √(M² − 1) + (wave drag due to thickness)

L/D drops sharply at supersonic Mach. A clean, ideal Mach-2 wing achieves L/D_max ≈ 8 (vs ≈ 18 for a transonic airliner wing). Concorde’s L/D in cruise was 7.4; the SR-71’s was about 6.

Delta wings and vortex lift (Polhamus 1966)

Highly swept (Λ > 60°) delta wings shed a stable leading-edge vortex at moderate α. The vortex reattaches over the upper surface and generates non-linear lift beyond the linear lift curve — the Polhamus suction analogy gives:

C_L = K_p·sin α·cos²α + K_v·cos α·sin²α

K_p is potential-flow lift coefficient, K_v is vortex-lift coefficient. Aircraft using this regime extensively: Concorde, F-16XL, F-22 (with leading-edge extensions / LEX), Eurofighter, JAS-39, Mirage 2000.

Engine inlets

• Fixed pitot / normal-shock inlet — used up to M ≈ 1.5 (F-16, Mirage 2000-9). One normal shock; total-pressure recovery drops rapidly with M.
• External-compression ramp — multiple oblique shocks plus a terminal normal shock. F-15, F-18, F-22, F-35.
• Mixed-compression (internal + external) — used at M > 2.5. SR-71 spike, Concorde rectangular ramp, MiG-25/31 splitter plate.
• Variable geometry — required above M ≈ 2 to keep the terminal shock at the throat across the speed range.

Inlet unstart (shock expelled from inlet) is catastrophic at high Mach — the SR-71 routinely lost engines this way; recovery required throttle chop, manual restart, often after a 30° yaw excursion.

7p. Hypersonic and re-entry aerodynamics

Above M ≈ 5 the flow chemistry matters. At the bow shock, kinetic energy converts to thermal energy proportional to V²; stagnation temperature behind the shock follows T₀ = T_∞ · (1 + (γ−1)/2 · M²). For an Apollo CM at re-entry V = 11 km/s, T₀ ≈ 11 000 K — far above air’s dissociation onset. Real-gas (Park 2-T or 5-species air-chemistry) models are mandatory.

Newtonian theory (Newton 1687, applied by Lees 1955)

For a blunt body in hypersonic flow, the very simple modified Newtonian approximation gives:

C_p = C_p,max · sin²θ          (θ = local body angle to free-stream)
C_p,max = (2 − γ) / γ · M²  (in the limit M → ∞; for air ≈ 1.84 at M = ∞)

Despite its crudeness, modified Newtonian predicts pressure distribution on blunt re-entry capsules to within ~10 % — sufficient for first-cut sizing.

Stagnation-point heating (Fay & Riddell 1958)

q̇_w = K · √(ρ_∞ / R_nose) · V_∞³

K ≈ 1.83 × 10⁻⁸ (W/m²) / (kg/m³)^0.5 · (m/s)^−3 · m^0.5 for laminar air stagnation flow with cold wall. Heating scales as V_∞³ and inversely with √R_nose — which is why blunt-body re-entry vehicles (Apollo, Soyuz, Dragon) have small radius of curvature but not too small; the trade is heating rate vs deceleration rate vs vehicle mass.

Thermal Protection Systems (TPS)

TPS	Service	Density / type
Inconel X-750 (hot structure)	X-15 leading edges, SR-71 forebody	8 000 kg/m³ metallic
Phenolic-impregnated nylon	Apollo CM, Mercury, Gemini	540 kg/m³ ablative
AVCOAT 5026-39	Apollo CM, Orion	510 kg/m³ ablative
LI-900 / HRSI silica tiles	Space Shuttle Orbiter underside	144 kg/m³ reusable
RCC (reinforced carbon-carbon)	Shuttle nose + wing leading edges	1 580 kg/m³ reusable
PICA / PICA-X	Stardust, Mars Science Lab, Dragon 2	270 kg/m³ ablative, partially reusable
3DMAT	Orion compression pad	woven 3D quartz composite

Aerocapture and aerobraking

Aerobraking uses repeated atmospheric passes at low dynamic pressure to circularize orbits (Magellan at Venus, Mars Reconnaissance Orbiter). Aerocapture uses a single, controlled high-dynamic-pressure pass to capture from hyperbolic approach (proposed for Mars Sample Return, Titan missions; not yet flown).

Hypersonic vehicle classes

• Ballistic re-entry capsules — Apollo CM, Soyuz, Dragon 2, Orion. Blunt body, L/D ≈ 0.3. Cross-range capability minimal; landing-site ellipse hundreds of km.
• Lifting body — X-24, M2-F2, HL-10. L/D ≈ 1.5 unpowered; tested 1960s–70s as Shuttle precursors.
• Winged orbiter — Shuttle Orbiter, Buran, X-37B, Dream Chaser. L/D ≈ 4–5 during hypersonic glide; significant cross-range (Shuttle ≈ 2 000 km).
• Hypersonic glide vehicle (HGV) — Avangard, DF-ZF, X-51 (boost-glide). Sustained M = 5–20 in upper atmosphere; uses dynamic boundary-layer transition control and aerodynamic skip-glide trajectories.
• Air-breathing scramjet — X-43A (M = 9.6 record, 2004), X-51 Waverider (M = 5.1 sustained, 2013), HiFire program. Engine and airframe inseparable; entire forebody compresses the inlet flow.

7pp. Aero / stealth / RCS coupling

Modern combat-aircraft external shaping is jointly optimized for aerodynamics and radar cross-section (RCS — covered in electromagnetics-engineering). The two disciplines often conflict but occasionally align.

• Aligned planform edges — RCS demands grouping of leading-edge, trailing-edge, and panel-line angles into a few discrete azimuths (the “F-22 facet directions” are 42° and 48° from the longitudinal axis). Aerodynamically this constrains sweep choice; supercruise at M = 1.6 with Λ = 42° is the F-22 compromise.
• Serpentine inlets — block line-of-sight to compressor face (a large RCS source) but introduce 4–8 % total-pressure loss and flow distortion. F-22, F-35, B-2, B-21 all use them.
• Tailless flying-wing configurations (B-2, X-47B, B-21) eliminate vertical-tail RCS contribution at the cost of weak directional stability — managed by split drag-rudders and fly-by-wire.
• Internal weapons bays preserve smooth fuselage RCS but introduce cavity flow — Rossiter modes, intense unsteady pressures, store-separation difficulties. F-22, F-35, B-2 all required extensive captive-carry CFD and tunnel test for bay-open conditions.
• Trailing-edge sawtooths (B-2, F-117 originally) align trailing-edge spikes with the same azimuth set as leading edges, accepting a small drag penalty for RCS uniformity.

8p. Stability and control

Longitudinal static stability

Aircraft is statically stable in pitch when a pitch-up disturbance generates a restoring nose-down moment. The criterion:

dC_M / dα < 0

Equivalently, the center of gravity must lie ahead of the neutral point (NP). The static margin:

SM = (x_NP − x_CG) / c̄

Practical SM: airliners 5–25 %, sailplanes 10–20 %, fighters −10 % to +5 % (relaxed static stability, computer-controlled — F-16 onwards). Negative SM gives higher maneuverability and lower trim drag but requires fly-by-wire stabilization.

Aerodynamic center and neutral point — typical numbers

Configuration	x_ac / mac (airfoil)	x_NP / mac (aircraft)	Static margin (typical)
Subsonic thin airfoil	0.25	n/a	n/a
Supersonic thin airfoil	0.50	n/a	n/a
Light GA aircraft	n/a	0.40–0.50	10–20 %
Transport (737, A320)	n/a	0.45–0.55	12–25 %
Long-range twin (777, 787)	n/a	0.50–0.60	8–18 %
4th-gen fighter (F-15)	n/a	0.20–0.30	0–5 %
Relaxed-stability fighter (F-16, F-22)	n/a	0.20–0.40	−10 % to +5 %

CG range is bracketed by forward limit (set by elevator authority at landing flare with full-aft stick) and aft limit (set by minimum static margin and stall recovery). Operational CG envelope rarely exceeds 25 % c̄ between forward and aft limits; loading procedures (weight-and-balance manifests) enforce this on every flight.

Lateral-directional modes

Mode	Description	Time scale
Roll subsidence	Pure roll damping; first-order	0.1–0.5 s
Spiral	Slow bank-angle divergence (or convergence)	20–200 s
Dutch roll	Coupled roll-yaw oscillation	2–6 s period
Phugoid (long.)	Long-period altitude/airspeed exchange	30–90 s
Short period (long.)	Rapid α / pitch-rate oscillation	1–4 s

Dutch roll is the most troublesome — swept wings with high dihedral effect couple roll and yaw such that the airplane “wallows.” Mitigation: yaw damper (rate-gyro signal to rudder), well-tuned vertical-tail sizing.

Handling qualities

Quantified by the Cooper-Harper rating scale (1969), 1 (excellent) to 10 (uncontrollable). MIL-STD-1797A (and the civil equivalents in 14 CFR Part 25 / 23) specify required dynamic-mode characteristics (short-period damping ratio, Dutch-roll frequency, control-force gradients).

Control-surface effectiveness

Surface	Primary axis	Effectiveness (typical)	Failure-mode concern
Elevator	Pitch	C_M,δe ≈ −1.0 to −2.0 /rad	Trim-tab runaway, jam
Stabilator	Pitch	All-moving; ~2× elevator authority	Mistrim, runaway
Aileron	Roll	C_L,δa ≈ 0.10–0.20 /rad	Reversal at high q (B-47, 737 MAX trim)
Rudder	Yaw	C_n,δr ≈ −0.05 to −0.15 /rad	Hardover, blowback
Spoiler (roll)	Roll	Augments aileron at high q	Stuck deployed
Speedbrake / spoiler	Drag / descent	ΔC_D up to 0.08	Asymmetric deployment
Canard	Pitch (FCS-prim)	C_L,δc ≈ 0.3–0.5 /rad	Vortex interference with wing
Thrust vectoring	Pitch / yaw	Axial T component × moment arm	Engine-out limits

Aileron reversal is the textbook gotcha: at high q the aileron’s twist of the elastic wing produces lift opposing the commanded roll, eventually reversing it (B-47 had this above M = 0.85 — fixed by adding spoilers as the primary high-speed roll device).

9p. Computational aerodynamics

Panel and vortex-lattice methods

For low-Mach external aero, panel methods discretize the body surface into source/doublet panels and solve the Laplace equation. Cheap, fast, useful for preliminary design.

• PMARC (NASA Ames, 1980s) — research and education.
• VSAERO (Analytical Methods Inc, 1980s) — commercial panel code with viscous coupling.
• AVL (Drela / MIT, 1990s) — vortex-lattice, the canonical light-aircraft + sailplane stability and trim tool. Free.
• VSPAero — OpenVSP’s built-in vortex-lattice / panel solver. Free.
• Tornado (Melin) — MATLAB-based vortex-lattice.
• XFOIL (Drela / MIT, 1989) — 2D airfoil analysis with integral boundary-layer coupling; the universal sailplane / UAV airfoil tool. Free.
• MSES (Drela) — multi-element 2D Euler with BL coupling; high-lift design.

CFD — Navier–Stokes solvers

Commercial RANS: ANSYS Fluent, Siemens Star-CCM+, Numeca FINE, AcuSolve, CFD++. Production aircraft design workhorses.

Open-source: OpenFOAM (broad solver library, requires Linux/CLI fluency), SU2 (Stanford, adjoint-capable, popular for shape optimization), Code_Saturne (EDF).

NASA in-house: FUN3D (unstructured, adjoint), CFL3D (structured), OVERFLOW (overset, helicopter and store-separation), NSU3D, VULCAN (combustion / hypersonic).

LES / high-order: charLES (Cascade Technologies — Stanford spinout), HiOSh, FLEXI (Stuttgart). 50–1000× cost of RANS, used for jet noise, combustion stability, separated-flow unsteady aero.

Hybrid RANS-LES (DES, DDES, IDDES — Spalart 1997+) — RANS in attached BL, LES in separated wake. The current sweet spot for landing-gear noise, store separation, flutter precursor, leading-edge vortex flows.

Adjoint and optimization. SU2 and FUN3D both have continuous-adjoint capabilities — compute the gradient of a scalar (drag, lift) with respect to thousands of shape parameters at the cost of one extra flow solve. Combined with OpenMDAO (NASA) and MACH-Aero (UMich MDOlab), these underpin modern transonic wing optimization (twist, planform, section camber).

Verification, validation, uncertainty quantification (VV&UQ)

AIAA G-077-1998 and ASME V&V 20-2009 distinguish:

• Verification — “solving the equations right.” Mesh-independence study (grid convergence index, Roache 1994), iterative convergence, code-to-code comparison.
• Validation — “solving the right equations.” Comparison against trusted experimental data with quantified uncertainty in both code and experiment.
• Uncertainty quantification — propagating known input uncertainties (geometry tolerance, BC uncertainty, model coefficients) into output uncertainty bars.

The AIAA Drag Prediction Workshop (DPW, biennial since 2001) and High-Lift Prediction Workshop (HLPW) are the community benchmarks; the spread in C_D predictions across ~50 participating groups for the same geometry (NASA CRM Common Research Model) has narrowed from ±15 counts in DPW-I (2001) to ±3 counts in DPW-VII (2022) — a measure of how far RANS practice has matured.

10p. Wind tunnels

Facility	Type	Mach range	Test section
NASA Langley 14×22	Low-speed	0–0.3	4.3 m × 6.7 m
Boeing BFR (Seattle)	Low-speed	0–0.3	2.4 m × 3.7 m
NTF (NASA Langley)	Cryogenic transonic	0–1.2	2.5 m × 2.5 m, T to 100 K
ETW (Cologne)	Cryogenic transonic	0.15–1.35	2.4 m × 2.0 m
ONERA S1MA (Modane, FR)	Transonic	0.05–1.0	8 m circular
NASA Glenn 8×6	Supersonic	0.36–2.0	2.4 m × 1.8 m
AEDC 16T / 16S	Trans/supersonic	0.06–4.75	4.9 m × 4.9 m
AEDC Tunnel 9 (White Oak)	Hypersonic blowdown	7, 8, 10, 14	1.5 m circular
LENS-I / LENS-II (CUBRC)	Hypersonic shock	3–22	up to 2.4 m
HEG (DLR Göttingen)	Hypersonic shock	6–10	nozzle exit ~0.6 m

Diagnostics. Force/moment balance (internal sting balance, external platform balance), pressure-sensitive paint (PSP) for global Cp, particle image velocimetry (PIV), Schlieren and shadowgraph for shocks, surface oil-flow for separation lines, infrared thermography for BL transition, tufts for qualitative separation.

Flight-test instrumentation

Beyond the wind-tunnel, every new airframe goes through dedicated flight test. Key instrumentation:

• Air-data boom — nose-mounted swiveling pitot-static and α/β vanes; the only way to get clean free-stream conditions independent of fuselage upwash.
• Trailing-cone static — calibration reference for static pressure (drag a cone on a tube 30 m behind the aircraft).
• Production pitot-static system — calibrated against the boom.
• Strain-gauged control surfaces — measure hinge moments, actuator loads.
• In-flight thrust measurement — gas-path instrumentation (NASA Glenn method), tail-pipe rakes.
• Wake-rake / wake-momentum drag measurement at low Mach (Jones method, 1936) — direct momentum-deficit integration in the wing wake gives section drag.
• Telemetry + flight-test engineer “team” — real-time downlink during envelope expansion. Required for flutter, stall, supersonic-acceleration test cards.

Typical envelope-expansion sequence: low-speed handling, stall series, stick-pusher demonstration, climb performance, cruise speeds, dive (V_MO + 10 %), flutter sweeps in fine increments, supersonic acceleration (military), engine-out and OEI handling, ice-shape handling, natural-icing flights. A clean-sheet civil transport requires 3 000–4 000 flight hours over 12–18 months to certificate.

Corrections. Wall interference, blockage, sting-support interference, Mach-shift in cryogenic tunnels, Reynolds-number scaling to flight. Cryogenic tunnels (NTF, ETW) match flight Re by lowering T, raising ρ — at the cost of operational complexity and run cost.

10pp. Rotors, propellers, and turbomachinery aero

Although pumps-turbomachinery covers axial-compressor and turbine design in detail, the aerodynamic foundations sit here.

Propeller / rotor blade-element momentum theory (Glauert 1935)

A propeller is treated as a stack of 2D airfoil sections at radius r, each operating at local angle of attack:

α(r) = β(r) − arctan( (V_∞ + v_i) / (Ω·r − v_θ) )

with β the blade-pitch angle, v_i the axial induced velocity (Froude), v_θ the swirl induced velocity. Thrust and torque integrate spanwise; advance ratio J = V_∞ / (n·D) (n = rev/s, D = diameter) and propeller efficiency η_p = J·C_T / C_P parametrize off-design performance.

Typical propeller peak efficiency η_p ≈ 0.85–0.92 at design J; loss mechanisms are tip compressibility (M_tip > 0.85 causes wave drag), blockage by spinner/nacelle, and root-section profile drag where the airfoil is fat for structural reasons.

Helicopter rotor aerodynamics adds: cyclic pitch control, blade flapping (Sikorsky 1930s), retreating-blade stall (limits forward speed to about 200 kt for conventional helicopters), advancing-blade compressibility (M_tip,advancing < 0.92 typical). Coaxial / co-rotating configurations (Ka-50, Sikorsky X2) cancel torque without a tail rotor and push the forward-speed envelope.

Wind-turbine aerodynamics

The same blade-element-momentum framework with Betz limit (Betz 1919): the maximum fraction of free-stream kinetic energy extractable by an open rotor is C_P,max = 16/27 ≈ 0.593. Modern utility-scale wind turbines (GE Haliade-X, Vestas V236, Siemens Gamesa 14-236 DD) achieve C_P ≈ 0.50 — within 85 % of theoretical maximum.

Tip-speed ratio λ = (Ω·R)/V_∞ is the key parameter; modern 3-blade turbines optimize at λ ≈ 7–9. Pitch control (full-span, hydraulic or electric) regulates power above rated wind speed by feathering blades.

10ppp. Ground-vehicle and built-environment aerodynamics

Automotive

Highway cars sit at Re ≈ 5 × 10⁶ (bluff body, frontal area ≈ 2–3 m², V ≈ 30 m/s). Drag breakdown for a typical sedan:

Component	Fraction of total C_D
Forebody and front-end	15–20 %
Underbody	10–25 %
Wheels and wells	15–25 %
Cooling system (through-flow)	5–10 %
Mirrors, antennas, trim	5–10 %
Wake / base drag (rear)	30–45 %

This is why Kammback (truncated tail) and boat-tail rear shaping dominate efficiency-focused EVs (Tesla Model S C_D = 0.208, Lucid Air C_D = 0.197, Mercedes EQS C_D = 0.20). Active grille shutters, retractable spoilers, underbody panels, and rear diffusers all attack wake drag.

Truck and rail

Class-8 tractor-trailer fleet drag (C_D = 0.6–0.9 unmitigated) is the largest single fuel cost in long-haul trucking; SAE J1321 fuel-economy test method quantifies aerodynamic-aid savings. Side-skirts, boat-tails, and trailer gap-reducers each cut 3–6 % fuel. High-speed rail uses tapered nose-cones (Shinkansen N700, ICE 3, TGV) — Re ≈ 10⁸; the dominant drag source is pressure (wake) at low speed, skin-friction-of-side-walls at high speed; tunnel-entry micro-pressure wave (sonic boom in tunnel) is the design driver for very-high-speed tunnel-running trains.

Buildings and bridges

Wind loading on buildings: ASCE 7 procedure with C_p,external × q_h × area for cladding; gust-effect factor G accounts for resonance of flexible structures. Tall slender buildings (Burj Khalifa, Shanghai Tower) use chamfered, twisted, or porous plan forms to disrupt coherent vortex shedding. Long-span bridges (Tacoma Narrows lesson, 1940) require wind-tunnel verification of aeroelastic stability — flutter, torsional divergence, vortex-induced vibration (VIV), and galloping all must be cleared. The Akashi-Kaikyo bridge deck went through six wind-tunnel iterations before construction.

11p. Worked examples

Example A — Cessna 172 cruise drag

Problem. A Cessna 172 in cruise: W = 1100 kg, S = 16.2 m², AR = 7.32, e = 0.75, C_D,0 = 0.027. Cruise V = 60 m/s (≈ 116 kt) at 5000 ft (ρ = 0.9 kg/m³). Find C_L, C_D, L/D, drag, cruise power.

Step 1 — Cruise C_L. q = ½ · 0.9 · 60² = 0.5 · 0.9 · 3600 = 1620 Pa. C_L = W·g / (q·S) = 1100 · 9.81 / (1620 · 16.2) = 10 791 / 26 244 = 0.411.

Step 2 — Induced drag and total C_D. k = 1 / (π · AR · e) = 1 / (π · 7.32 · 0.75) = 1 / 17.25 = 0.0580. C_D,i = k · C_L² = 0.0580 · 0.169 = 0.0098. C_D = C_D,0 + C_D,i = 0.027 + 0.0098 = 0.0368.

Step 3 — L/D. L/D = C_L / C_D = 0.411 / 0.0368 = 11.2.

Step 4 — Drag force. D = q · S · C_D = 1620 · 16.2 · 0.0368 = 966 N.

Step 5 — Cruise power required at the propeller. P_aero = D · V = 966 · 60 = 57.9 kW (78 hp). With prop efficiency η_p ≈ 0.78 at cruise: P_shaft = P_aero / η_p = 74.3 kW ≈ 99 hp shaft.

The Lycoming O-320 produces 150 hp at sea-level full throttle; at 5000 ft and 75 % power setting it makes ≈ 112 hp shaft. The 99 hp requirement matches comfortably — typical fleet 172s cruise at 60–65 % power for fuel economy, exactly aligned with this analysis.

Example B — Transonic shock onset on a NACA 0012

Problem. NACA 0012 (12 % thick symmetric) at α = 0 in cruise: V_∞ = 250 m/s, T = 220 K (≈ 36 000 ft). Determine M_∞, estimate whether the upper surface is shocked.

Step 1 — Speed of sound and M_∞. c = √(γRT) = √(1.4 · 287 · 220) = √88 396 = 297.3 m/s. M_∞ = 250 / 297.3 = 0.841.

Step 2 — Incompressible minimum C_p on the airfoil. From standard NACA 0012 data: min C_p,inc ≈ −0.40 at the suction peak (≈ 11 % chord, α = 0).

Step 3 — Prandtl–Glauert correction. β = √(1 − M²) = √(1 − 0.707) = √0.293 = 0.541. C_p,local = C_p,inc / β = −0.40 / 0.541 = −0.739.

Step 4 — Critical C_p at M_∞ = 0.841. The isentropic relation gives C_p,crit (the C_p at which local M = 1):

C_p,crit = (2/(γ·M_∞²)) · { [ (1 + 0.5(γ−1)·M_∞²) / (1 + 0.5(γ−1)) ]^(γ/(γ−1)) − 1 }
        = (2/(1.4·0.707)) · { (1.1414 / 1.2)^3.5 − 1 }
        = 2.020 · { 0.9512^3.5 − 1 }
        = 2.020 · { 0.8404 − 1 } = 2.020 · (−0.1596) = **−0.322**.

Step 5 — Compare. C_p,local = −0.739 is more negative than C_p,crit = −0.322 → local flow is supersonic → the airfoil is shocked. M_∞ = 0.841 lies well above M_crit (which for a NACA 0012 is ≈ 0.74 at α = 0).

Conclusion. A modern airliner cannot cruise on a NACA 0012 at this Mach; a supercritical section (e.g. RAE 2822 or proprietary Boeing / Airbus sections) is required to push M_DD above 0.84.

Example C — Boeing 787-9 takeoff field-length sizing

Problem. A 787-9 at MTOW = 254 t at a hot-and-high airport (Denver, 5 433 ft, ISA+20, ρ = 0.95 kg/m³). S = 360 m², C_L_max,TO = 2.4 (full-extended flaps + slats), C_D,TO = 0.13 at lift-off, T_static = 2 × 320 kN = 640 kN per engine spec. Estimate V_rotation and ground-roll distance to lift-off.

Step 1 — Stall and rotation speed. V_stall_TO = √( 2·W·g / (ρ·S·C_L_max) ) = √( 2 · 254 000 · 9.81 / (0.95 · 360 · 2.4) ) = √( 4.985 × 10⁶ / 820.8 ) = √(6 074) = 77.9 m/s = 151 kt. V_R ≈ 1.10 · V_stall = 85.7 m/s ≈ 167 kt.

Step 2 — Average effective thrust during roll. Modern high-bypass turbofans lose thrust with V: T(V) ≈ T_static · (1 − 0.5·M). At Denver M_R ≈ 85.7/345 ≈ 0.249, so T(V_R) ≈ T_static · 0.876. Average T̄ over the roll ≈ 0.94 · T_static = 1 203 kN total.

Step 3 — Average drag during roll. At V_R, q_R = 0.5·0.95·85.7² = 3 489 Pa. D_R = q_R · S · C_D = 3 489 · 360 · 0.13 = 163 kN. Average D over roll ≈ D_R / 2 (since D ∝ V²) ≈ 82 kN. Rolling friction (concrete runway, µ_r ≈ 0.02): F_r = 0.02 · 254 000 · 9.81 = 49.8 kN.

Step 4 — Ground-roll distance. Net average accelerating force: F_net = T̄ − D̄ − F_r = 1 203 − 82 − 50 = 1 071 kN. Average acceleration: ā = F_net / m = 1 071 000 / 254 000 = 4.22 m/s². Ground roll: s = V_R² / (2·ā) = (85.7)² / 8.44 = 870 m ≈ 2 850 ft.

Add 35 ft screen-height climb-out (≈ +400 m at climb-out gradient ≈ 5°) → certificated TOFL ≈ 1 270 m. The published 787-9 hot/high TOFL at these conditions is ≈ 2 500 m once OEI (one-engine-inoperative) and 35 % regulatory margin are applied — consistent with this rough cut once the safety overheads are layered on.

Example D — Apollo CM peak stagnation heating

Problem (Example D). Apollo Command Module re-entering after a lunar return. At peak heating: V_∞ = 11 000 m/s, ρ_∞ = 5 × 10⁻⁴ kg/m³ (≈ 65 km altitude), R_nose = 4.7 m (effective spherical-cap radius). Find stagnation-point heat flux.

Step 1 — Fay–Riddell (engineering form). q̇_s = K · √(ρ_∞ / R_nose) · V_∞³, K = 1.83 × 10⁻⁸ W·s³·m^−5.5·kg^−0.5

Step 2 — Substitute. √(ρ_∞ / R_nose) = √(5 × 10⁻⁴ / 4.7) = √(1.064 × 10⁻⁴) = 1.031 × 10⁻² (kg/m³)^0.5 / m^0.5. V_∞³ = (11 000)³ = 1.331 × 10¹² m³/s³. q̇_s = 1.83 × 10⁻⁸ · 1.031 × 10⁻² · 1.331 × 10¹² = 2.51 × 10⁶ W/m² = 2.51 MW/m² = 251 W/cm².

This matches Apollo flight data (peak ≈ 270 W/cm² recorded on Apollo 4). Integrated over the 8-minute peak-heating phase, total heat load is ≈ 30 MJ/m² at the stagnation point — the AVCOAT ablator handles this by char-layer formation and pyrolysis-gas blowing, losing about 20 % of its ≈ 60 mm thickness during entry.

12p. Edge cases and gotchas

1. Stall — flow separation when α exceeds critical (~12–16° for typical airfoils, lower for highly swept wings). Deep stall of T-tail aircraft (BAC 1-11, DC-9 prototype) occurs when the wake of a stalled wing blankets the high tailplane, eliminating pitch recovery authority. Mitigation: stick-pusher, ventral fin, certification stall-recovery demonstration.
2. Buffet boundary — at transonic, shock-induced separation oscillates, exciting structural modes. Airliners certify a buffet-free envelope with a 1.3-g margin to onset.
3. Flutter — coupled bending-torsion wing oscillation at a critical V_flutter. Killed the Helios solar-electric UAV (2003), grounded the F-15 fleet briefly in the 1970s, and modes of it appeared in the 737 MAX MCAS context. Always verify flutter speed exceeds V_dive by 15 % per FAR 25.629.
4. Aeroelastic divergence — static, low-frequency wing twist runaway. Sets the absolute upper speed of forward-swept wings.
5. Wing-tip vortex strength scales as W/(ρ·V·b). A 380-tonne A380 trailing a 90-tonne CRJ at 6 nm separation can roll the smaller aircraft inverted — the ICAO wake-turbulence categories (Heavy / Medium / Light / Super) exist for this.
6. Ground effect — within ½-span of the ground, induced drag drops ~30–40 % and lift rises. Beneficial for takeoff acceleration; treacherous if not anticipated during landing flare (the airplane “floats”).
7. Reynolds-number sensitivity — small UAVs at Re ≈ 10⁵ have radically different C_L_max, c_d, and stall behavior than airliners at Re ≈ 10⁷–10⁸. Re-scaling from wind-tunnel sub-scale to full-scale is the dominant uncertainty in early aero predictions.
8. Boundary-layer transition — tripped vs untripped wind-tunnel data differ by factors of 2 in drag. Always document and match in flight test.
9. Engine-inlet distortion — separated, swirling, or shock-bound flow at the compressor face causes surge / stall. F-22 has a serpentine inlet shaped both for stealth (line-of-sight blockage) and for distortion management.
10. Icing — ice accretion on leading edges collapses C_L_max by 30–50 %. FAR 25 Appendix C / O ice envelopes define certification icing conditions; anti-ice (hot-air bleed, electrothermal) and de-ice (pneumatic boot, expulsive) systems are required.
11. Compressibility-correction validity — Prandtl–Glauert breaks above M ≈ 0.7; Karman–Tsien (1941) and Laitone (1951) extend it modestly, but transonic CFD or wind-tunnel data are required for design.
12. Sonic boom — N-wave overpressure typically 0.5–2 psf (24–100 Pa) for fighters, the reason Concorde was banned overland. The NASA / Lockheed X-59 QueSST reshapes the longitudinal pressure signature into a sequence of weak shocks (“soft thump” ≈ 75 PLdB) to enable supersonic over-land flight.
13. Aero-thermal coupling at hypersonic — vehicle skin temperatures of 1500–2500 K change material moduli, geometry (thermal bowing), and even local Re via temperature-dependent viscosity. The X-43A’s wedge-shape forebody had to be flight-tested twice (failed first time on the Pegasus booster fin issue, succeeded Mach-7 and Mach-9.6).
14. Wind-tunnel scale-up — Reynolds-number mismatch, sting interference, blockage, model surface finish all introduce systematic uncertainty. The “C_D walk” between wind-tunnel prediction and flight-test measurement is typically ±5–10 counts (1 count = 0.0001) and is closely guarded by every manufacturer.
15. Trim drag — the horizontal stabilizer carries a download in conventional aft-tail configurations, requiring extra wing lift and extra induced drag. Canard and three-surface configurations rearrange this but introduce other compromises.

13p. Reference aircraft data

Aircraft	M_cruise	AR	Sweep Λ_¼	S (m²)	C_L_cruise	L/D_max	W_MTOW (t)
Cessna 172	0.18	7.3	0°	16.2	0.41	11	1.10
Cessna Citation X	0.92	8.3	37°	49	0.50	13	16.0
Boeing 737-800	0.785	9.4	25°	124.6	0.53	18	79.0
Boeing 787-9	0.85	11.0	32°	360	0.55	21	254
Airbus A350-900	0.85	9.5	31.9°	442	0.55	21	280
Airbus A380-800	0.85	7.5	33.5°	845	0.55	20	575
F-16C Block 50	0.9 / 2.0	3.2	40°	27.9	varies	9 / 4	19.2
F-22A	1.6 supercruise	2.36	42°	78.0	varies	9 / 5	38.0
Concorde	2.04	1.6	75° ogival	358.6	0.10	7.4	185
SR-71	3.2	1.7	60°	167.2	varies	6	78.0
Apollo CM (entry)	36 → 0	n/a	n/a	12.0	n/a (blunt)	0.35	5.8
Space Shuttle Orbiter (entry)	25 → 0.3	2.27	81/45°	249.9	n/a	4 (HSC)/4.5 (LSC)	110

13pp. Representative airfoils and their use

Designers do not invent airfoils from scratch except for the most demanding programs. The vast majority of GA, UAV, sailplane, and even modern transport work picks from a known catalog and tunes via inverse design.

Airfoil	t/c	c_ℓ_max (Re = 3×10⁶)	M_DD (α = 0)	Notable use
NACA 0009	0.09	1.32	0.78	Symmetric tail surfaces, ailerons
NACA 0012	0.12	1.55	0.74	Helicopter rotors, fins, T-tails
NACA 2412	0.12	1.65	0.73	Cessna 172, 152, Piper PA-28
NACA 4412	0.12	1.70	0.71	Piper Cherokee, Auster, older GA
NACA 23012	0.12	1.79	0.71	Beech Bonanza, DC-3, Bf 109
NACA 65₂-415	0.15	1.45	0.70	P-51 Mustang (laminar bucket)
NASA LS(1)-0417	0.17	2.05	0.66	Piper Tomahawk, NASA experimental GA
RAE 2822	0.121	1.62	0.78	Transonic CFD benchmark
NASA SC(2)-0714	0.14	1.60	0.82	Supercritical research
Eppler E387	0.09	1.20 (Re = 5×10⁵)	n/a	Sailplane / UAV low-Re reference
Selig S1223	0.121	2.20 (Re = 3×10⁵)	n/a	High-lift low-Re; HALE UAVs
Drela DAE 11 / 21	0.082	1.05 (Re = 2×10⁵)	n/a	MIT Daedalus, micro-air-vehicles

13ppp. Typical preliminary-design aerodynamic workflow

A clean-sheet new aircraft (or a major derivative) typically proceeds:

1. Mission analysis. Range, payload, speed, field length → design point on the constraint diagram (W/S vs T/W).
2. Initial sizing (Raymer, Roskam, Torenbeek methods). Empirical weight fractions; first-pass S, b, AR, Λ, MTOW.
3. Airfoil selection. Pick from catalog or commission inverse design (Drela MSES-style) at the cruise design point.
4. Vortex-lattice / panel-method analysis. AVL or VSPAero for stability derivatives, span loading, control authority. Iterate planform and twist.
5. RANS CFD at cruise and high-lift. Drag polar, M_DD, buffet boundary, high-lift performance. Ten to thirty geometries explored.
6. Wind-tunnel test. Subscale model (typically 1/20 to 1/8) in a transonic tunnel. Force/moment polars, surface PSP, hinge moments.
7. Aeroelastic and flutter analysis. Coupled CFD-FEM (CFD/CSD); flutter speeds verified by wind-tunnel flutter model.
8. High-Re / cryogenic tunnel confirmation. NTF or ETW run at flight-Re to bound the wind-tunnel-to-flight extrapolation.
9. Flight test. Envelope expansion, performance verification, stability and control, certification icing, hot/high.
10. Service-life follow-on. Drag-cleanup retrofits (winglets, fairings, riblets) come years after entry-into-service as economics demand.

The whole loop for a clean-sheet airliner consumes 5–7 years from authority-to-offer to type certificate, of which aerodynamics-driven work spans about the first 4 years.

14. Cross-references

• fluid-mechanics — governing equations, Reynolds and Mach scaling, boundary-layer theory (Sections 7 and 9p there).
• thermodynamics — gas-dynamic stagnation properties, isentropic relations, compressible energy framing.
• heat-transfer — stagnation-point convective heating, hypersonic ablator design.
• materials-aluminum — 2024-T3 / 7075-T6 / 7050-T7651 airframe alloys.
• materials-composites — CFRP wings (787, A350), GLARE fuselage skin (A380).
• materials-ceramics — Shuttle silica tiles (LI-900, FRSI), RCC leading edges, ultra-high-temperature ceramics for scramjets.
• electromagnetics-engineering — radar cross-section (RCS) shaping interacts with aero shaping in stealth aircraft design.
• vibration-dynamics — flutter, buffet, control-surface freeplay.
• propulsion — planned companion note; inlets, nozzles, turbofans, turbojets, ramjets, rockets.
• pumps-turbomachinery — planned; shares blade-row analysis methods with axial compressors and turbines.
• scientific — planned; mesh and case-setup file formats (CGNS, OpenFOAM dict, Plot3D, Tecplot).
• aerospace-defence — planned; KML, X-Plane / MSFS scenery files, AIRAC navigation data, FCS / flight-control specs.

15. Citations

1. Anderson, J. D. Fundamentals of Aerodynamics, 7th ed. McGraw-Hill, 2023. ISBN 978-1264123599. The canonical undergraduate text.
2. Anderson, J. D. Modern Compressible Flow with Historical Perspective, 4th ed. McGraw-Hill, 2020. ISBN 978-1260570823.
3. Anderson, J. D. Hypersonic and High-Temperature Gas Dynamics, 3rd ed. AIAA Education Series, 2019. ISBN 978-1624105142.
4. Bertin, J. J.; Cummings, R. M. Aerodynamics for Engineers, 7th ed. Pearson, 2021. ISBN 978-1108481892.
5. Drela, M. Flight Vehicle Aerodynamics. MIT Press, 2014. ISBN 978-0262526449. Modern, mathematically rigorous treatment from the AVL/XFOIL author.
6. Katz, J.; Plotkin, A. Low-Speed Aerodynamics, 2nd ed. Cambridge University Press, 2001. ISBN 978-0521665520.
7. Schlichting, H.; Truckenbrodt, E. Aerodynamics of the Airplane. McGraw-Hill, 1979. Out of print but cited universally.
8. Liepmann, H. W.; Roshko, A. Elements of Gasdynamics. Wiley, 1957 (Dover reprint 2001). ISBN 978-0486419633. The compressible-flow classic.
9. Hoerner, S. F. Fluid-Dynamic Drag. Self-published, 1965. The empirical drag bible — still indispensable.
10. Hoerner, S. F.; Borst, H. V. Fluid-Dynamic Lift, 2nd ed. Hoerner Fluid Dynamics, 1985.
11. Whitcomb, R. T. “A Study of the Zero-Lift Drag-Rise Characteristics of Wing-Body Combinations near the Speed of Sound.” NACA Report 1273, 1952. The area-rule paper.
12. Whitcomb, R. T.; Clark, L. R. “An Airfoil Shape for Efficient Flight at Supercritical Mach Numbers.” NASA TM X-1109, 1965. The supercritical-airfoil paper.
13. Drela, M. “XFOIL: An Analysis and Design System for Low Reynolds Number Airfoils.” Lecture Notes in Engineering 54, Springer 1989.
14. Polhamus, E. C. “A Concept of the Vortex Lift of Sharp-Edged Delta Wings Based on a Leading-Edge-Suction Analogy.” NASA TN D-3767, 1966.
15. Fay, J. A.; Riddell, F. R. “Theory of Stagnation Point Heat Transfer in Dissociated Air.” Journal of the Aeronautical Sciences, vol. 25(2), 1958, pp. 73–85.
16. Cooper, G. E.; Harper, R. P. “The Use of Pilot Rating in the Evaluation of Aircraft Handling Qualities.” NASA TN D-5153, 1969.
17. MIL-STD-1797A (2004) — Flying Qualities of Piloted Aircraft. The military handling-qualities specification.
18. USAF Stability and Control DATCOM (1978, with updates through 1999). The empirical preliminary-design handbook.
19. 14 CFR Part 25 (Transport Category Airworthiness Standards) — FAA airworthiness regulation for transport aircraft.
20. 14 CFR Part 23 (Normal Category Airworthiness) — light-aircraft equivalent, performance-based since 2017 revision.
21. FAR Part 33 — Airworthiness Standards: Aircraft Engines.
22. NASA Technical Reports Server (NTRS) — searchable archive of NACA and NASA reports (https://ntrs.nasa.gov).
23. AIAA standards portfolio — modern industry standards for terminology, wind-tunnel uncertainty, CFD verification & validation (AIAA G-077, AIAA S-117).