Concorde never floated gently onto a runway. It arrived nose-high, droop nose lowered so the pilots could see over it, touching down at around 170 mph at a far steeper angle of attack than any conventional airliner would use. Above that pale triangle, invisible to the passengers, two huge spiralling vortices were holding the aircraft up.

The same basic triangle shaped the Dassault Mirage III, IL Avro Vulcan and the Convair interceptors that guarded North America. It is one of the most recognisable shapes in aviation, and one of the most misunderstood. The delta wing looks like a swept wing with the gap filled in, but it behaves very differently. It is a set of trade-offs that made perfect sense for Mach 2 and very little sense for a slow, tight dogfight.

Here is the delta wing explained: why designers chose it, what it gave them, what it cost them, and why today's fighters still use it with canards bolted on the front.

Lippisch's triangle

The practical delta wing begins with Alexander Lippisch, a German designer who coined the name Delta for a series of tailless aircraft. His Delta I glider flew in 1931, and four more Delta designs followed by 1939. They were not easy to handle at low speed and none saw widespread use, but they put the shape on the map.

During the Second World War Lippisch worked on a supersonic delta interceptor, the P.13a, and built a test glider for it, the DM-1. Allied troops captured both the glider and its designer. The DM-1 was completed for the United States and shipped to NACA's Langley laboratory, where it was tested and modified to reduce its drag.

Millennium 7 on the delta wing dilemma: why the shape suits some missions brilliantly and others badly.

According to NASA's own history of the type, conferences with Lippisch convinced Convair's engineers that a delta would work for their XF-92 interceptor project. The interceptor was cancelled, but its flying mock-up survived as the Convair XF-92A, which officially first flew on 18 September 1948 and became the test bed for America's first operational delta fighters.

It was not a pleasant aeroplane. The XF-92A suffered severe pitch-up, sometimes beyond 6 g, and its handling became legendary at Edwards. Yet it proved the configuration was practical, and Convair carried the lessons into the F-102 Delta Dagger, IL F-106 Delta Dart e il B-58 Hustler.

Thin, stiff and full of fuel

The delta's first gift is structural. Because the wing root runs a long way along the fuselage, a wing that is very thin as a percentage of its chord can still be physically deep at the root. Deep spars are stiff spars, so the delta can be built stronger, stiffer and lighter than a swept wing of similar lifting capability.

That long root also leaves a lot of volume inside the wing for fuel, landing gear and equipment. It made the shape cheap and robust to build, a real factor in the success of both the MiG-21 and the Mirage family. On supersonic designs, engineers usually spent that advantage on an even thinner aerofoil to cut drag.

Built for Mach 2

Speed is the delta's reason to exist. A strongly swept leading edge sits behind the shock cone at transonic and low supersonic speeds, which keeps the airflow over the wing behaving more like subsonic flow and holds wave drag down. A thin wing helps further, and the delta lets a wing be thin without becoming flimsy.

The shape alone was not enough. Convair's YF-102 first flew in October 1953 and could not exceed Mach 1 in level flight because its transonic drag was far higher than predicted. Convair rebuilt it around Richard Whitcomb's area rule, pinching the fuselage into a Coke-bottle waist, and the YF-102A went supersonic the day after its first flight in December 1954.

In France, Dassault's tailless Mirage III first flew in November 1956, and a pre-production Mirage IIIA reached Mach 2 in level flight in October 1958. The thin delta had become the standard recipe for a Mach 2 interceptor: a big engine, a sharp nose, and a triangle behind it.

Vortex lift: the delta's hidden trick

At low speed, any wing needs a higher angle of attack to keep lifting. On a conventional wing, flow separation at high angles brings the stall. On a sharply swept delta, the air spilling around the leading edge rolls up into a pair of strong vortices over the upper surface. Their low pressure adds lift and delays the stall.

Chuck Yeager noticed the effect in the XF-92A, but it was the Royal Aircraft Establishment's Dietrich Küchemann and Johanna Weber who turned it into a design tool. Their slender delta work showed that the vortex lift grows with the length of leading edge it acts along, which pointed straight towards a long, slender wing running far along the fuselage.

Real Engineering's study of Concorde, including how its slender ogival wing worked at Mach 2 and on approach.

That is Concorde's ogival, or ogee, delta, its leading edge curving in a smooth double sweep from nose to tip. It cruised efficiently at Mach 2, then relied on vortex lift to fly slowly enough to land, with no flaps or slats at all. Pitch and roll came from six elevons on the trailing edge.

The price: drag, speed and no flaps

Every gift has a bill. A delta has a low aspect ratio, short span and broad chord, and NASA's own formula for induced drag shows why that hurts: for the same wing area, a lower aspect ratio means more drag for every unit of lift. In a hard, sustained turn, a pure delta bleeds energy faster than a fighter with a higher aspect ratio wing.

Landing is the second bill. Without vortex lift the wing produces little lift at low angles, so deltas approach nose-high and fast. Concorde rotated to about 18 degrees on take-off and needed tall, strong landing gear to allow for it, and its normal landing speed was around 170 mph.

The third bill belongs to tailless deltas in particular. With no horizontal tail, the elevons must both trim and control the aircraft, so conventional trailing-edge flaps cannot be used. On the Mirage III that meant a relatively long take-off run and a high landing speed, while the large, lightly loaded wing gave a bumpy ride at low altitude.

The Vulcan: a delta for the bomb bay

Avro chose the delta for a different reason. To meet a demanding RAF bomber specification, its team found that an otherwise conventional jet with a 45-degree swept wing would have been far too heavy. Shortening the span while keeping the wing area, and filling in the gap, produced a deep, thick triangle that could swallow the engines, the fuel and the bomb load.

Avro knew Lippisch's work, but followed its own logic, and de-risked the shape with the small Avro 707 research jets. The Vulcan prototype VX770 first flew on 30 August 1952 with Avro's Superintendent of Flying, Roly Falk, at the controls.

Even the Vulcan's delta needed fixing. Near its limiting Mach number the original straight leading edge buffeted, so Avro gave the production wing a kinked and drooped leading edge, the so-called phase 2 wing, first tried on an Avro 707A.

The Vulcan entered RAF service in 1956 as part of Britain's nuclear deterrent and flew its only combat missions in 1982, the long-range Black Buck raids of the Falklands War. The RAF retired it in 1984, and the restored XH558 kept the great delta flying at air shows until 2015.

Imperial War Museums on the Avro Vulcan, Britain's best-known V bomber and its giant delta wing.

Tails, canards and cranks: fixing the delta

Designers soon began editing the pure triangle. The Soviet answer was the tailed delta, adopted by the TsAGI research institute to improve high angle of attack handling and centre of gravity range. The MiG-21 added a conventional tailplane to its 57-degree delta, which freed the trailing edge for real flaps, and with more than 11,000 built it is widely recorded as the most-produced supersonic jet in history.

Sweden went the other way. The Saab 35 Draken used a double delta with a sharply swept inner section, and the Saab 37 Viggen of 1967 pioneered the close-coupled canard, a small delta foreplane whose vortex strengthens the airflow over the main wing. That layout returned in the Saab Gripen, IL Dassault Rafale e il Eurofighter Typhoon, all of them fly-by-wire designs.

The United States tried a cranked arrow. General Dynamics' F-16XL, first flown on 3 July 1982, replaced the F-16's wing and tailplanes with a delta more than twice the size of the original wing, improving supersonic lift-to-drag ratio and adding a great deal of internal fuel. It lost the Enhanced Tactical Fighter contest to the F-15E Strike Eagle and ended its days as a NASA research aircraft.

Even the Mirage evolved. Israel's IAI Kfir added canards to the Mirage-derived airframe, and Dassault's later Mirage 2000 paired a tailless delta with fly-by-wire controls.

Why the delta never went away

The pure tailless delta of the 1950s was a specialist: superb at climbing hard and running fast in a straight line, less happy in a long turning fight or on a short runway. Its descendants keep the delta's strengths, the stiff thin structure, the fuel volume, the vortex lift and the low supersonic drag, and use canards, tails, cranks and computers to pay down its weaknesses.

Lippisch's triangle has now been flying for almost a century. Look at a Rafale, a Gripen or a Typhoon on the runway and you are still looking at his delta wing, with a few hard-won corrections attached.

Sources: NASA (Dryden fact sheet FS-2003-06-080, Glenn Research Center Beginner's Guide to Aeronautics, NASA Image Library), Royal Air Force Flying Review (via Vulcan to the Sky Trust), Heritage Concorde, Imperial War Museums, Wikipedia (Delta wing, Alexander Lippisch, Concorde, Avro Vulcan, Convair F-102, Dassault Mirage III, MiG-21, F-16XL)