Concorde

• 1. Introduction

  ›

  On the 7th of February 1996, British Airways Concorde G-BOAD flew from New York to London in 2 hours, 52 minutes and 59 seconds, measured from take-off to touchdown. A tailwind of 175 mph helped. Even so, that time belongs to an aircraft unlike any other airliner. Sud Aviation and the British Aircraft Corporation developed Concorde jointly as a supersonic passenger jet, and it is now retired. Only 20 were ever built, and 18 of them sit on display in Europe and North America. How did two governments commit to a single machine? What did it take to make an aluminium airliner outrun the speed of sound? And why was a design that drew so many airline options built in such small numbers?

  320212

• 2. The Slender Delta

  ›

  In the early 1950s, Arnold Hall, director of the Royal Aircraft Establishment, or RAE, asked Morien Morgan to form a committee on supersonic transport. The group met in February 1954 and delivered its first report in April 1955. Robert T. Jones of NACA had shown that supersonic drag depends strongly on wingspan. The committee therefore studied short, thin, trapezoidal wings, like those on the Lockheed F-104 Starfighter and the planned Avro 730 bomber. Its baseline design looked like an enlarged Avro 730.

  "Some horribly large aeroplanes" were the result when the committee tried to design for existing runways. A short wingspan produces little lift at low speed, so take-offs run long and landings come in fast. Lifting such an aircraft off would demand enormous engine power and fuel. The group judged supersonic transport infeasible and proposed only continued low-level study of supersonic aerodynamics.

  Soon afterwards, Johanna Weber and Dietrich Küchemann of the RAE published reports on a new wing shape, which Britain called the slender delta. Their colleague Eric Maskell wrote a report titled "Flow Separation in Three Dimensions" that deepened understanding of separated airflow. At a high angle of attack, a delta wing throws up strong vortices over its upper surface. Each vortex lowers the air pressure there and creates lift. Chuck Yeager had noticed the effect in the Convair XF-92, but its value went unappreciated. Weber suggested using it to improve low-speed performance.

  Weber also noted that vortex lift grows with the length of wing it acts over. That suggested stretching the wing along the fuselage as far as possible. Such a layout kept good supersonic performance and gave reasonable take-off and landing speeds. The price was a very nose-high attitude, which raised doubts about low-speed handling.

  Test pilot Eric Brown remembered Küchemann presenting the idea at a meeting that Morgan attended. Morgan, Brown said, immediately seized on it as the solution to the supersonic transport problem. Brown regards that moment as the birth of the Concorde project.

  On the 1st of October 1956, the Ministry of Supply asked Morgan to form a new body, the Supersonic Transport Aircraft Committee, or STAC. At its first meeting on the 5th of November 1956, it chose to fund a test-bed for the slender delta's low-speed behaviour. That contract eventually produced the Handley Page HP.115, which showed safe control at speeds as low as 69 mph. That is about one-third the speed of the F-104 Starfighter.

  STAC reasoned that supersonic lift-to-drag ratios are about half those of subsonic designs, so more thrust is needed. Even so, faster aircraft could fly more paying trips in a given time, so fewer would be needed on a route. The committee sketched a transatlantic model at about Mach 2 and a shorter-range one at Mach 1.2. Morgan estimated £75 to £90 million to develop the 150-passenger transatlantic version, with service starting in 1970.

  Three wing shapes were under study at the RAE. These were the classic straight-edge delta, the gothic delta rounded like a gothic arch, and the ogival wing, compound-rounded into an ogee. A practical problem forced the choice. The wing's centre of pressure should sit close to the aircraft's centre of gravity, but the gravity point drifts during design and as fuel burns. With a delta running most of the fuselage length, the wing could not simply be slid fore or aft. Measured against these shifts, the ogee shape came to the fore.

  In 1965, NASA modified a Douglas F5D Skylancer to mimic the chosen wing, and the test aircraft landed at noticeably lower speeds than a standard delta. Simulations at NASA's Ames centre also predicted a sudden pitch change on entering ground effect. Ames test pilots later joined French and British pilots in a cooperative test and found the simulations correct. That finding was added to pilot training.

  79101214

• 3. Treaty and Name

  ›

  In April 1960, Pierre Satre, technical director of Sud Aviation, was sent to Bristol to discuss a partnership. Bristol's engineers were surprised to find that the French company had designed a similar aircraft and reached the same economic conclusions as STAC. The explanation emerged later. The STAC report, marked "For UK Eyes Only", had been passed secretly to France to win political favour. Sud made minor changes and presented it as its own work.

  France had run its own design contest in the late 1950s. Sud Aviation's Super-Caravelle won with a medium-range design, deliberately sized to avoid competing with the transatlantic American aircraft the French assumed were being drawn up. Nord Aviation had proposed a ramjet-powered Mach 3 aircraft, and Dassault a jet-powered Mach 2 design. France had no modern large jet engines and had already decided to buy a British one, as it had for the subsonic Caravelle. A top speed of around Mach 2 was chosen, which would also speed development and let the aircraft fly before the Americans.

  Through 1961 the two teams kept meeting. The British still wanted a 150-passenger transatlantic aircraft, while France deliberately avoided those routes. Yet the two designs grew very similar. A single design emerged that differed mainly in fuel load. More powerful Olympus engines then allowed either version to use only four.

  Peter Thorneycroft, the UK Minister of Aviation, told the cabinet that France was far more serious about a partnership than any American company. The American firms had shown no interest, probably because they expected government funding and feared giving away technological leadership. The Treasury took a negative view. It pointed to the industry's "over-optimistic estimating", including the recent history of the TSR.2, and warned that the cost estimate would prove much too low.

  The Committee on Civil Scientific Research and Development reviewed the project between July and September 1962. Its October report rejected the economic arguments and expected no direct positive return. Still, it said the project deserved consideration, because everyone else was going supersonic and Britain feared being locked out of future markets. Britain was also pressing for admission to the European Economic Community, and that became the main rationale. The work was therefore set up as a treaty between the two countries, not a commercial agreement. It included a clause, requested by the UK government, imposing heavy penalties for cancellation. The treaty was signed on the 29th of November 1962.

  On the 25th of January 1963, Charles de Gaulle vetoed the UK's entry into the European Community in a speech. That same month, at a press conference by de Gaulle, the aircraft was first called Concorde. The 18-year-old son of F.G. Clark, publicity manager at BAC's Filton plant, had suggested the name. It comes from the French word concorde, which has an English equivalent, concord, and both mean agreement, harmony or union.

  Harold Macmillan changed the spelling to Concord in response to a perceived slight by de Gaulle. At the French roll-out in Toulouse in late 1967, British Minister of Technology Tony Benn announced that he would change it back. A nationalist uproar followed, and it died down when Benn said the final "e" stood for "Excellence, England, Europe, and Entente (Cordiale)". A Scotsman wrote to complain, since part of the aircraft was made in Scotland. Scotland supplied the nose cone, and Benn replied that it was also "E" for "Écosse", the French name for Scotland. He added that he might have included "e" for extravagance and "e" for escalation. In British usage the type is simply "Concorde", with no article.

  1516171928

• 4. Heat and Altitude

  ›

  Hiduminium R.R. 58, an aluminium alloy, was used throughout the airframe because it was relatively cheap and easy to work with. The highest temperature it could sustain over the aircraft's life was 127 °C. That limit capped the top speed at Mach 2.02, since faster flight would have ruled out aluminium. Cruising at that speed also suited the engines, which burned about 4800 gallons an hour.

  Kinetic heating from the boundary layer warmed the skin in supersonic flight. Windows and panels were warm to the touch by the end of a flight. Apart from the engine bay, the nose was the hottest part of the structure. The airframe heated and cooled twice per flight, once climbing and accelerating, and again descending and landing. Engineers built a test rig that repeatedly heated and cooled a full-size section of wing, and they tested metal samples at intervals. The airframe was designed for a life of 45,000 flying hours.

  The fuselage grew by as much as 300 mm as it heated. A gap opened on the flight deck between the flight engineer's console and the bulkhead. On some aircraft making a retiring supersonic flight, flight engineers wedged their caps into the gap, and the caps were trapped when the airframe shrank again.

  Concorde used its fuel as a heat sink for the air conditioning, and the same method cooled the hydraulics. A visor kept hot air from flowing over the cockpit skin. Even the paint was regulated. Most of the surface had to be highly reflective white, which cut the skin temperature by 6 to 11 degrees Celsius.

  A shift of about 2 m in the centre of pressure occurred as the aircraft passed its critical Mach number. Trim controls could have countered it, but at such speeds they would have added unacceptable drag. Instead, fuel was pumped along the aircraft during acceleration and deceleration, so that it acted as an auxiliary trim control. Large forces in turns distorted the structure. Engineers solved this by varying the ratio between inboard and outboard elevon deflections, and only the innermost elevons, on the stiffest part of the wing, worked at high speed. The narrow fuselage also flexed, which rear passengers could see looking along the cabin.

  People aboard received almost twice the flux of extraterrestrial ionising radiation as those on a conventional long-haul flight. Early speculation held that this would raise the likelihood of skin cancer. The shorter flight time normally gave a lower overall dose, though unusual solar activity could raise it. The flight deck carried a radiometer, and if radiation grew too high, Concorde descended below 47000 ft.

  The cabin was pressurised to the equivalent of 6000 ft, at the low end of the usual range, while the aircraft cruised as high as 60000 ft. Subsonic airliners typically cruise below 44000 ft. At Concorde's altitude, the plastic oxygen masks of other jets would not have been effective. Concorde therefore had smaller windows, a reserve air supply and a rapid descent procedure. The Federal Aviation Administration noted the higher operating altitude and concluded that rapid descent was the best response to pressure loss.

  5476798091

• 5. Engines and Intakes

  ›

  A symposium titled "Supersonic-Transport Implications" at the Royal Aeronautical Society on the 8th of December 1960 aired competing views on the engine. Speakers weighed podded against buried installation, and turbojets against ducted fans. The aircraft had to fly long distances to pay its way, which demanded an efficient powerplant. Turbofans were rejected because their larger cross-section caused excessive drag. Rolls-Royce proposed the RB.169, but a wholly new engine for a single aircraft would have been extremely costly. Instead the existing BSEL Olympus Mk 320 turbojet, already flying in the BAC TSR-2 strike bomber prototype, was chosen.

  Dr. Seddon of the RAE favoured a buried installation over a podded one. One worry was that putting two or more engines behind a single intake could turn an intake failure into a double or triple engine failure. Each engine therefore got its own intake, and the nacelles were paired with a splitter plate between them. Only above Mach 1.6 was an engine surge likely to affect its neighbour. That is why Concorde "had to be certified aerodynamically as a twin-engined aircraft above Mach 1.6".

  The intakes had to slow supersonic air to subsonic speeds efficiently, with low distortion to prevent engine surge, across all likely temperatures. The variable features included front and rear ramps, a dump door, an auxiliary inlet and a ramp bleed to the exhaust nozzle. The wing's boundary layer lay ahead of the intake, and two-thirds of it was diverted. During pushovers the remaining layer thickened and caused surging. Wind tunnel testing defined leading-edge changes ahead of the intakes that solved the problem.

  The Air Intake Control Units used a digital processor, the first with full authority over an essential system in a passenger aircraft. BAC's Electronics and Space Systems division developed it after the analogue units on the prototypes, built by Ultra Electronics, proved insufficiently accurate. Ultra Electronics also developed the thrust-by-wire engine control system.

  On a subsonic airliner, a failed engine adds drag and yaws the aircraft toward the dead side. At supersonic speed that could theoretically have broken the airframe, and computer simulations predicted considerable problems. In practice, Concorde could shut down both engines on one side at Mach 2 without difficulty. The auxiliary spill door opened and the ramps extended fully, deflecting air downward past the engine. Pilots were routinely trained for double engine failure.

  Noise proved stubborn. SNECMA advanced silencer design during the programme, but by 1974 the spade silencers projecting into the exhaust were reported ineffective. Reheat, the afterburners, was used only at take-off and to pass through the transonic range between Mach 0.95 and 1.7.

  5859686971

• 6. Nose and Undercarriage

  ›

  At rotation, Concorde rose to an angle of attack of about 18 degrees, at 199 knots indicated airspeed. Before rotation the wing produced almost no lift, which stressed the main undercarriage in ways nobody expected during development. The result was a major redesign. The legs had to be unusually tall and strong, and a small set of wheels was added aft to prevent tailstrikes. The main units swing toward each other to stow, and because of their height they first contract telescopically.

  The four main tyres on each bogie were inflated to 232 psi, and the twin nose wheels to 191 psi. The nose assembly carried a spray deflector to stop standing water from being thrown into the engine intakes. The tyres were rated to a maximum runway speed of 250 mph.

  Dunlop developed the brakes, the first carbon-based brakes used on an airliner. Carbon saved 1200 lb compared with equivalent steel. Each wheel's multiple discs were cooled by electric fans. After a typical landing at Heathrow, brake temperatures reached 300 to 400 °C.

  The normal landing speed was 170 mph. On approach, Concorde sat on the backside of the drag curve, where raising the nose increases the rate of descent. It was therefore flown largely on the throttle, with an autothrottle to ease the pilots' workload. Landing needed at least 6000 ft of runway, and Cardiff Airport was the shortest on which Concorde landed with commercial passengers. On the 20th of August 1977, G-AXDN made its final landing at Duxford Aerodrome, whose runway was just 6000 ft at the time. It was the last aircraft to land there before the runway was shortened later that year.

  Marshall's of Cambridge developed the drooping nose. At high angles of attack, the long pointed nose blocked the pilots' view, so it had to droop. A moving visor retracted into the nose before the nose was lowered. For taxiing and take-off, the nose sat 5 degrees below horizontal. After clearing the airport, the nose and visor were raised. Before landing the visor retracted again and the nose dropped to 12.5 degrees below horizontal for maximum visibility. After touchdown the nose came up to the 5 degree position to avoid damage from ground vehicles. It was raised fully before engine shutdown, so that condensation inside the radome would not pool and seep onto the pitot and air data probes.

  The United States Federal Aviation Administration objected to the restricted visibility through the visor on the first two prototypes. Those had been designed before suitable high-temperature window glass existed. A redesigned visor went onto the pre-production and production aircraft. Triplex developed the nose window and visor glass, which had to endure temperatures above 100 °C in supersonic flight.

  5657678663

• 7. Orders and Rivals

  ›

  Advertisements in Aviation Week & Space Technology during the late 1960s predicted a market for 350 aircraft by 1980. The consortium planned a long-range and a short-range version, but customers showed no interest in the short-range one, and it was dropped. Options poured in from the major airlines of the day. Pan Am, BOAC and Air France were the launch customers, with six aircraft each. When the prototype first flew, the list stood at 74 options from 16 airlines.

  The programme estimate in 1962 was £70 million. By 1976 the bill stood between £1.5 and £2.1 billion. Costs spiralled to more than six times the original projections, and the unit cost reached £23 million in 1977. That was the main reason the production run was far smaller than expected. During the early 2000s, Flight International called Concorde "one of aerospace's most ambitious but commercially flawed projects".

  The 1973 oil crisis and the 1973-74 stock market crash left airlines wary of aircraft that burned a lot of fuel. New wide-bodies such as the Boeing 747 had also made subsonic flying much more efficient. With a full load, Concorde achieved 15.8 passenger miles per gallon. The Boeing 707 reached 33.3, the 747 reached 46.4, and the McDonnell Douglas DC-10 reached 53.6. A trend toward cheaper tickets led airlines such as Qantas to question whether Concorde suited the market.

  Pan Am cancelled its options on the 31st of January 1973, part of a wave of cancellations. The crash of the Soviet Tupolev Tu-144 at the Paris Le Bourget air show had shocked potential buyers. Public concern over the sonic boom, take-off noise and pollution also turned opinion against supersonic aircraft. By 1976 the remaining buyers came from four countries: Britain, France, China and Iran. Only Air France and British Airways took up their orders, for seven aircraft each, and the two governments took a cut of any profits.

  The United States, India and Malaysia all ruled out supersonic Concorde flights over noise concerns, though some restrictions were later relaxed. Professor Douglas Ross characterised the limits imposed under President Jimmy Carter's administration as an act of protectionism for American manufacturers. The sonic boom also meant supersonic flight over land caused complaints, which confined Concorde to transoceanic routes.

  Western European journalists nicknamed the Tupolev Tu-144 "Concordski" for its outward similarity. Soviet espionage allegedly stole Concorde blueprints to help its design. It carried passengers from November 1977, but a second crash in May 1978 ended its passenger career after only 55 flights, which averaged 58 passengers. Jean Rech of Sud Aviation blamed its shorter range on a very heavy powerplant with an intake twice as long as Concorde's. The Soviet government cancelled the programme on the 1st of July 1983.

  The American entry in the race had two main designs, the swing-wing Boeing 2707 and the compound-delta Lockheed L-2000. Both were larger, with seating for up to 300 people, and Boeing's was selected. Concorde first flew in 1969, the year Boeing began building mockups after redesigning to a cropped delta wing. The cost of that change helped kill the project. The American government cancelled it in 1971, having spent more than $1 billion without building any aircraft. Earlier tests with the XB-70 Valkyrie, the B-58 Hustler and the Oklahoma City sonic boom trials had already shown that booms reach the ground, which raised the same environmental concerns that hindered Concorde.

  2627291790

• 8. Service, Crash, Retirement

  ›

  On the 21st of January 1976, Air France flew from Paris-Roissy and British Airways from London Heathrow to begin scheduled service. Washington Dulles followed on the 24th of May, and New York JFK from the 17th of October 1977. Subsonic jets needed eight hours from Paris to New York and seven on the return. Concorde's average transatlantic flight took just under 3.5 hours.

  British Airways also flew London to Bahrain, Miami and Barbados. Air France flew Paris to Dakar and Rio de Janeiro, and Paris to the Azores and Caracas. Noise-related bans and low profitability dogged the routes. British Airways later repositioned Concorde as a super-premium service, and it then became profitable.

  With no other civil traffic at about 56000 ft, Concorde had dedicated oceanic tracks. Their co-ordinates stayed fixed, because high-altitude winds vary much less than those lower down. Concorde was cleared in a 15000 ft block, so it could climb slowly as fuel burned off. One recollection of flying it describes 747s far below "almost appearing to go backwards", and calls the aircraft "an absolute delight to fly".

  Braniff International Airways operated Concordes at subsonic speed between Dulles and Dallas Fort Worth from January 1979 until May 1980. It used its own crews, insurance and operator's licence. Stickers with a US registration covered the French and British registrations on each rotation. Singapore Airlines had its livery on the left side of G-BOAD, with its insignia on the cabin fittings and "Singapore Girl" stewardesses sharing duty with British Airways attendants. Crew, operations and insurance stayed with British Airways throughout. The arrangement began with three flights between the 9th and the 13th of December 1977, and resumed from the 24th of January 1979 to the 1st of November 1980.

  The aircraft also carried the famous. Pope John Paul II flew on Concorde in May 1989, and Elizabeth II took it to Barbados for her Silver Jubilee in 1977. Concorde was used to observe the total solar eclipse of the 11th of August 1999. A charter flew from Heathrow to Sydney on the 13th of February 1985 in 17 hours, 3 minutes and 45 seconds, including refuelling stops. In August 1995, Air France Concorde F-BTSD circled the world eastbound in 31 hours, 27 minutes and 49 seconds, with six refuelling stops. In November 2003, G-BOAG flew from New York to Seattle in 3 hours, 55 minutes and 12 seconds, mostly supersonic over sparsely populated Canadian territory.

  On the 25th of July 2000, Air France Flight 4590, registered F-BTSC, crashed at Gonesse after leaving Charles de Gaulle Airport for New York JFK. All 100 passengers and nine crew died, along with four people on the ground. It remains the only fatal accident involving Concorde. The official investigation by France's BEA traced the cause to a metallic strip that fell from a Continental Airlines DC-10 which had taken off minutes earlier. It punctured a tyre on the left main bogie. The tyre exploded, a piece of rubber hit a fuel tank, and the resulting leak caught fire.

  The crew shut down engine number 2 after a fire warning, and engine number 1 was surging and producing little power. The aircraft could not gain altitude or speed. It pitched up, descended suddenly, rolled left and hit the Hôtelissimo Les Relais Bleus Hotel tail-low.

  Independent investigators and former British Airways Chief Concorde Pilot Mike Bannister dispute the official findings. They argue for an "accident chain" of maintenance and operational errors. Evidence from the British team indicated that a critical spacer sleeve was omitted from the left main landing gear during maintenance. That left the bogie misaligned and made the aircraft veer toward the runway edge before it struck the debris. Bannister also alleged that the aircraft was over its maximum structural take-off weight, and that the crew shut down engine number 2 too early, at a critically low altitude. A French court found Continental criminally liable in 2010. An appeals court overturned the criminal convictions in 2012, while keeping civil responsibility on the company and the aircraft's manufacturer.

  Before that day, Concorde had arguably been the safest operational passenger airliner in the world, with no passenger deaths. Its tyre damage rate from 1995 to 2000 was nonetheless 30 times that of subsonic airliners. The BEA counted 57 tyre failure incidents from 1976 onward. The worst came on the 14th of June 1979, when F-BVFC lost two tyres on its take-off roll at Washington Dulles. Debris penetrated three fuel tanks, and the aircraft was leaking 4 kilograms of fuel per second when it landed. Rudder panels had also separated in flight in 1989 and 1992, without injuries.

  Both airlines grounded their fleets after the crash. Safety improvements included more secure electrical controls, Kevlar lining on the fuel tanks and burst-resistant tyres. On the 17th of July 2001, Bannister piloted the first modified aircraft from Heathrow on a 3 hour 20 minute flight toward Iceland. He reached full cruise speed and altitude before returning to RAF Brize Norton, and crowds and live television watched. Commercial operations resumed on the 7th of November 2001, with G-BOAE and F-BTSD flying to New York JFK, where Mayor Rudy Giuliani greeted the passengers.

  In 2003, Air France and British Airways announced Concorde's retirement. They cited rising maintenance costs, low passenger numbers after the 2000 crash and the slump in air travel after the September 11 attacks. Air France flew its last commercial flight on the 30th of May 2003. British Airways retired its fleet on the 24th of October 2003-27 years after commercial operations began. Grandstands were erected at Heathrow for the final day.

  In 2006, a BBC and Design Museum contest called the Great British Design Quest drew 212,000 votes. Concorde won, beating the Mini, the Jaguar E-Type, the Tube map and the Supermarine Spitfire. The recollection quoted earlier ends with the observation that, even in the 21st century, the aircraft "remains unique".

  13202174