🌀 Aerodynamics: The Science That Shapes Aircraft

What you’ll learn

1. Where the Trouble LivesExplain d'Alembert's paradox and Prandtl's boundary layer as the division of the flow into two regions, and state the course's through-line.In 1752 d'Alembert proved rigorously that an inviscid fluid exerts zero drag — 'a singular paradox' that divorced theory from engineering for 150 years. On 12 August 1904 Prandtl resolved it by splitting the flow: outside a thin layer viscosity genuinely doesn't matter and potential flow is exactly right, while inside it the no-slip condition creates an enormous velocity gradient, so a tiny viscosity produces forces that are not tiny at all. The course's argument follows: air will follow almost any shape for almost nothing provided the change is gradual, and every refusal — near a surface, or near Mach 1 — is what an aircraft's shape exists to prevent.
2. Inside the Boundary LayerDescribe the boundary layer's structure, the origin of skin friction, and the opposed virtues of laminar and turbulent layers.The boundary layer grows from nothing at the leading edge to a few centimetres by the trailing edge; its thickness is conventionally the height at which flow reaches 99% of freestream, and for a laminar layer δ ≈ 5.0·x/√Re. Shear within it, summed over the wetted area, is skin friction drag — the dominant term on a slender body, meaning an airliner's biggest single resistance is air rubbing on paint. A laminar layer has a gentle profile and low friction; a turbulent one is 'fuller', with more wall shear but far more momentum near the surface. Those opposed virtues set up every choice in the chapters that follow.
3. TransitionExplain boundary-layer transition, its critical Reynolds number, and why natural laminar flow is so hard to keep in service.Every layer starts laminar and transitions downstream as it thickens and becomes unstable; for a smooth flat plate transition begins near Re_x ≈ 5×10⁵, though the real value depends on freestream turbulence, surface finish, pressure gradient and vibration. Transition is a one-way door, and a rivet head, panel gap, scratch, frost or insect can trip it early. Natural laminar flow holds a favourable (falling) pressure gradient far aft to keep the layer stable, and it works on hangared, hand-finished composite sailplanes. The P-51 Mustang is the cautionary case: a laminar-flow section whose laminar run was largely lost to wartime tolerances and field conditions.
4. Separation: Running UphillExplain flow separation as a momentum failure in an adverse pressure gradient, and why a turbulent layer is worth its extra friction.Over the front of a wing, falling pressure accelerates and stabilises the flow; over the back, pressure must recover to freestream, giving an adverse gradient — a hill. Air outside the layer can trade speed for pressure and climb it; air inside has been robbed by friction since the leading edge, and at some point the slowest air stops and reverses. The layer lifts off, levering the outer flow with it, and the resulting wake means the pressure never recovers — pressure drag, which on a separated body can exceed skin friction tenfold. A turbulent layer's near-wall momentum lets it climb far further, so designers pay a few percent of friction to buy later separation, which is what vortex generators do in hardware.
5. Stall, ProperlyDistinguish leading-edge, trailing-edge and thin-airfoil stall, and explain spanwise stall progression, washout, pitch-up and deep stall.Two wings can share a CLmax and a stall angle and fly completely differently, because what matters is where separation begins and how it spreads. Trailing-edge stall creeps forward gently with warning; leading-edge stall bursts a bubble and drops the whole upper surface at once; thin-airfoil stall arrives early with a low CLmax. Thickness is what makes the pressure recovery gradual. Along the span, a root-first stall gives nose-drop, buffet warning and working ailerons, so designers force it with washout, stall strips and section variation at a permanent cruise cost. Pitch-up on swept wings and T-tail deep stall are configuration failures, not airfoil ones.
6. The Drag PolarAssemble the drag polar C_D = C_D0 + C_L²/(π·AR·e), explain the drag bucket, and account for why airliners do not cruise at L/D max.Parasite drag rises with V²; induced drag — the energy permanently carried away in the trailing vortex sheet — is 'inversely proportional to the square of the airspeed (at a given lift)', so it falls. Their sum is a bucket whose minimum, at L/D max, falls where the two are equal. The polar's four symbols name the whole design argument: C_D0 (wetted area, finish, separation), C_L² (why slow flight is draggy), AR (the strongest lever — Wikipedia lists 33.5 for an ASH 31 glider against 9.5 for a 787 and 1.55 for Concorde), and e, an honest fudge measuring the shortfall from an elliptical lift distribution. Airliners cruise fast of L/D max because crew, lease, slots and cycles are spent per hour, not per mile.
7. Air Stops Being AirExplain compressibility physically as the collapse of the air's advance warning, and place the Mach 0.3 convention and Prandtl–Glauert correctly.Air ahead of a wing learns the wing is coming via pressure waves travelling at the speed of sound; that forewarning is why subsonic flow is so agreeable. As Mach rises the warning time collapses, and at Mach 1 the aircraft arrives with its own announcement. Mach 0.3 is a convention about acceptable error — density variations reach a few percent there — not a physical threshold. Prandtl–Glauert's factor 1/√(1−M²) amplifies pressures with Mach and predicts infinity at M = 1, which is the linearised theory announcing that its small-disturbance assumption has died. The infinity was a signpost that a great many people in the 1940s read as a barrier.
8. The Drag Rise Is RealExplain the critical Mach number, why nothing bad happens at it, and how shock formation produces wave drag, buffet, Mach tuck and drag divergence.A wing accelerates air over its upper surface, so local flow reaches Mach 1 while the aircraft is well subsonic — the critical Mach number. Nothing bad happens there; the bill comes at the back of the supersonic pocket, which must return to subsonic and can only do so through a shock. That shock produces wave drag (an irreversible entropy-generating loss) and, worse, shock-induced separation, since it is the steepest possible adverse pressure gradient. Together they give drag divergence: 'a rapid increase in drag from about Mach 0.8, and it is the fuel cost of the drag that typically limits the airspeed.' The 1940s symptoms — dead controls, buffet, Mach tuck — were a design problem, and the XS-1 in 1947 solved it rather than broke anything.
9. SweepExplain sweep via the independence principle and the cosine effect, enumerate its costs, and place the supercritical airfoil as the alternative.Busemann proposed swept wings for supersonic flight at the 1935 Volta Conference and was largely ignored. The mechanism is that only the flow component perpendicular to the leading edge negotiates the airfoil's curvature: 'a wing with a 45 degree sweep will see a reduction in effective curvature to about 70% of its straight-wing value. This has the effect of increasing the critical Mach by 30%.' The costs are paid at low speed near the ground — spanwise flow fattening the tip boundary layer, tip stall and divergent pitch-up (the F-100's 'Sabre dance'), reduced CLmax, a longer and heavier spar, and Dutch roll. Whitcomb's supercritical airfoil — flat top, large leading-edge radius, aft camber — instead makes the shock weak and far aft, letting designers use less sweep; it flew on a T-2C and then a TF-8A Crusader, and is on the 757, 767, 777, A300, A310 and C-17.
10. Area RulingExplain the Whitcomb area rule via the F-102 case, and identify it as the through-line's most literal statement.The YF-102 could not reach Mach 1 despite a Mach 1.2 design requirement, because wind-tunnel predictions had been 'overly optimistic'. After Busemann's late-1951 'streampipes' talk, Whitcomb realised in 1952 at NACA Langley that transonic air objects not to any component but to abrupt changes in the aircraft's total cross-sectional area along its length. The F-102A's fuselage was indented beside the wings and volume added at the rear, which 'reduced the transonic drag significantly' and let it make Mach 1.2 — the aircraft went faster because it was made smaller in the right place. Küchemann reached the same principle independently; the Sears–Haack body is the related ideal, though 'not theoretically optimum' transonically since it derives from Prandtl–Glauert and Ackeret theory.
11. Wave DragExplain wave drag as a tax on volume, correct the sonic-boom misconception, and account for vortex lift as controlled separation.Supersonically the air cannot be forewarned, disturbances are confined to the Mach cone, and — the big change — thickness itself becomes drag, since every bit of volume must be shocked aside irreversibly. Wave drag scales roughly with the square of thickness/chord, so supersonic design is about slenderness rather than boundary-layer management. The sonic boom is not an event at Mach 1 but a continuously trailing cone, a carpet of overpressure wherever the aircraft goes — which is why Concorde could not fly supersonically over land. Concorde's 12.2-foot rear extension, supersonically area-ruled at Mach 2, 'reduced wave drag by 1.8%'. Its ogee delta makes vortex lift by turning separation into a stable, repeatable, chartable component.
12. High Lift, and a Lie You Were ToldRefute the 'high-energy air' account of the slot on Smith's grounds, present his five effects, and separate a Fowler flap's three benefits.A wing optimised for Mach 0.85 cannot land: at a third of the speed it must be nine times better at making lift, so it changes shape twice a flight. The universal textbook account — that the slot ducts high-energy air to re-energise the boundary layer — is false in both halves. Per Wikipedia's leading-edge slat article, following A.M.O. Smith's 1975 Wright Brothers Lecture: 'the slat does not give the air in the slot a high velocity (it actually reduces its velocity) and also it can not be called high-energy air since all the air outside the actual boundary layers has the same total heat.' Smith's five effects — slat, circulation, dumping, off-the-surface pressure recovery, fresh boundary layer — include three explainable by pure potential flow, and not one is 'add energy'. A Fowler flap adds camber, adds AREA, and opens a slot; the area is the one people forget.
13. Designing to the Drag RiseAccount for an airliner's cruise Mach number as a position taken near the drag-divergence cliff, and close the course's through-line against a real wing.Below drag divergence, drag rises gently with V² and is negotiable; at M_dd the curve turns a corner as wave drag and shock-induced separation appear together. So a cruise Mach is not a smooth optimum but a position near a cliff, held with margin for weight, temperature, climb, ATC and a dirty wing — and everything in the course's back half exists to move the cliff. Read against a real wing: sweep is a cosine buying critical Mach; the supercritical section keeps the shock weak so less sweep is needed; winglets chase e and AR; vortex generators refuel a boundary layer; fences fight spanwise drift; canoe fairings hold machinery that makes the wing physically larger. Not one makes the aircraft faster or lighter. Every one makes an abrupt change gradual.

Questions this course answers

Grounded in trusted sources

• Wikipedia — D'Alembert's paradox (1752; d'Alembert 1768 on 'a strictly vanishing resistance, a singular paradox'; Prandtl's 1904 resolution): https://en.wikipedia.org/wiki/D%27Alembert%27s_paradox
• Wikipedia — Boundary layer (Prandtl's paper presented 12 August 1904 at the third International Congress of Mathematicians, Heidelberg; no-slip condition; 99%-of-freestream thickness definition; displacement thickness; δ ≈ 5.0x/√Re): https://en.wikipedia.org/wiki/Boundary_layer
• Wikipedia — Lift-induced drag (C_Di = C_L²/(π·AR·e); 'induced drag is inversely proportional to the square of the airspeed (at a given lift)'): https://en.wikipedia.org/wiki/Lift-induced_drag
• Wikipedia — Aspect ratio (aeronautics) (ASH 31 AR=33.5; Eta motor glider AR=51.33; Dash 8 Q400 AR=12.8; Boeing 787 and Airbus A350 AR=9.5; A380 AR=7.8; Piper PA-28 Cherokee AR=5.6; Concorde AR=1.55): https://en.wikipedia.org/wiki/Aspect_ratio_(aeronautics)
• Wikipedia — Transonic (transonic range Mach 0.8–1.2; speed of sound '343 m/s at sea level'; 'Transonic airspeeds see a rapid increase in drag from about Mach 0.8, and it is the fuel cost of the drag that typically limits the airspeed'): https://en.wikipedia.org/wiki/Transonic
• Wikipedia — Swept wing (Busemann's 1935 Volta Conference proposal; 'a wing with a 45 degree sweep will see a reduction in effective curvature to about 70% of its straight-wing value. This has the effect of increasing the critical Mach by 30%'; spanwise flow, tip stall, pitch-up and the F-100 'Sabre dance'; 'sweeping it increases the length of the spars running along it from root to tip. This tends to increase weight and reduce stiffness'): https://en.wikipedia.org/wiki/Swept_wing
• Wikipedia — Supercritical airfoil (Whitcomb; 'flattened upper surface, highly cambered ("downward-curved") aft section' and larger leading-edge radius; shocks 'farther aft than traditional airfoils'; T-2C Buckeye and TF-8A Crusader testbeds; Boeing 757/767/777, Airbus A300/A310, C-17): https://en.wikipedia.org/wiki/Supercritical_airfoil
• Wikipedia — Area rule (Whitcomb 1952 at NACA Langley's 8-foot High-Speed Tunnel; Busemann's late-1951 'streampipes' talk; the YF-102 could not reach Mach 1 against a Mach 1.2 design requirement, tunnel predictions 'overly optimistic'; the F-102A's indented fuselage and added aft volume 'reduced the transonic drag significantly'; Küchemann; Sears–Haack body 'not theoretically optimum' transonically; Concorde's rear fuselage extended 12.2 feet, 'reduced wave drag by 1.8%'): https://en.wikipedia.org/wiki/Area_rule

Every Wunder lesson is built from real, reputable sources — never invented.

Related Science courses