The lower right wear strip was missing a rivet and the strip itself was distorted. Compared with the original part, the strip was too long and prevented the fan door from closing. There was considerable wearing around the other strips. The investigators asked Continental to remove the reverser cowls and put them in storage.
According to the manufacturer, the disassembly and repair of the wear strips is classified as a “minor repair”. It requires no special tools and requires no particular inspection after completion. New wear strips are made in the workshop. Shims are supposed to be placed between the strip and the cowl, but these are often left out, leaving too much play between the forward and aft cowls. Investigators found this to be the case on a number of aircraft. In normal conditions, this was satisfactory. But during take-off when the engine is at maximum thrust, the pressure inside the cowl is very high, which could explain the loss of the wear strip.
The plane’s maintenance log shows that the left wear strips on engine three were replaced in Tel Aviv by Israel Aircraft Industries during a routine check on 11 June 2000. Further work on the engine’s reverser cowl was carried out in Houston, where the lower left wear strip was changed. The technician who completed the maintenance report said that he had noticed a twisted wear strip that was sticking out of the cowl.
It is not easy to spot that the wear strip is missing when the cowl doors are closed. But the cowl doors on engine three were opened at least once between 9 July and 3 September 2000. None of the maintenance documents refers to the wear strips during this period.
The strip found on the runway was subjected to examination at the laboratory. It was found to be made of an alloy of titanium instead of stainless steel as the manufacturers specify. Along with the twelve rivet holes needed to hold it on, it had another seventeen holes that had been filled in with epoxy. Black marks on it and debris found jammed in one of the rivets were subjected to spectroscopic analysis and were found to be similar to the material of Concorde’s tyres.
More laboratory tests showed that the primer and mastic on the strip found on the runway were the same as those found on the aircraft’s cowl. Drill holes in the cowl correspond to those on the strip, though some of the rivets still in place in the cowl did not correspond to those on the strip. These were thought to have been left behind from a previous strip. However, the torn and unstuck areas of the mastic did correspond and the investigators concluded that the strip found on the runway did indeed come from the Continental Airlines DC-10.
The investigators also examined the two pieces of tank five found at the crash site and the three pieces found on the runway. At the crash site, one piece had a hole in it, which showed that the impact occurred from the outside towards the inside of the tank, from the left to the right and more or less from the rear towards the front. The puncture showed a clear petal-shaped structure, indicating a high-energy penetration not due to the final impact. Analysis was unable to provide details of the penetrating object, but its probable trajectory shows that it could have come from the left main landing gear.
However, the pieces found on the runway showed clear signs of rupture – from inside outwards. This excludes the possibility that it had been caused directly by a puncture and a great deal of theoretical work and practical experimentation had to be done to explain it. It was then calculated that, in principle, if a fuel tank was hit by a heavy object that did not rupture it, a pressure wave would be transmitted through the fuel inside that could rupture the tank at a weak point.
Tests were done at Centre d’Essais Aéronautiques de Toulouse – the Toulouse aeronautical test centre – where they fired pieces of tyre at test boxes containing a liquid with similar characteristics to kerosene. One was made out of a panel from tank five, taken from a Concorde while the plane was suspended from service following the crash. It was not possible to make a box that completely represented tank five as the material it was made out of, AU2GN, was no longer available.
Due to the large number of variables, it was not possible to reproduce the rupture of the tank found in the accident, but measurements taken during the tests demonstrated that it was possible to rupture the tank that way. A piece of tyre weighing just 4.5 kg travelling at a speed of around 140 metres per second (504 km/h; 313 mph) – easily achieved by a tyre burst when the wheel was rotating at speed – could have caused such a rupture.
Tank five was loaded with 7.2 tons of fuel before take-off. After the accident, the gauge indicated that two tons were left in the tank. This meant that several dozen kilograms of fuel were leaking from the tank every second – ten times greater than the loss of fuel after the accident in Washington, DC. This would mix with air and catch fire. But what ignited the fuel?
A surge in the engine can cause a flame in the air intakes. But this was rejected as the source of the ignition because the engine surge happened after the fuel streaming from with ruptured tank caught fire.
A spark from an electrical circuit in the main landing gear, damaged by debris from the burst tyre, could have been the source of ignition. Such damage was noted in the Washington incident. However, modifications were made after that incident and no further damage to the cables there was reported.
The fuel could have been ignited on contact with the hot walls of the engine or with gas coming out of the reheat (afterburner). Yet there was no trace of fire damage on the engines. However, the fan blades of engine one showed signs of damage from impact with small pieces of stainless steel, while damaged fan blades in engine two showed it had been hit by something soft. There had been nineteen cases of damage to engines due to tyre debris; six led to a loss of thrust during take-off.
The surge and subsequent loss of thrust from engine one was thought to have been caused by the debris it sucked in, while the surge in engine two was blamed on hot gases entering the engine. Engine two was then closed down and the fire handle was pulled in an attempt to douse the fire. Engines three and four experienced a surge, too. But this was due to the roll and high angle of attack. All damage to the engines was found to have come from their impact with the ground.
Attempts were made to retract the landing gear. In the cockpit voice recorder, a voice can be heard saying “the landing gear is not retracting” and “I’m trying”. Examination of the wreckage did not bring to light the cause. But there were indications that the hydraulic system had failed.
It was also calculated that, if the pilot had aborted the take-off, the plane would have still being travelling at between 74 and 115 knots when it reached the end of the runway – with similarly catastrophic results. The landing gear would have collapsed and, with a fire under one wing, the whole plane would have burst into flames immediately.
The Operations Manual for Concorde said: “The decision to abort take-off before V
1
must only be taken in case of a significant loss of thrust or fire on an engine, or with the certainty that the aircraft will be unable to fly (loss of an essential structural element, for example . . .). In all other cases it is preferable to continue the take-off” and “In case of a failure on take-off, no action will be taken before 400 feet AAL [Above Aerodrome Level], apart from ensuring the track and gear retraction.”
A study of the safety record of Concorde showed that the rate of tyre problems on take-off had reduced from one in 1,500 to one in 8,000 over the lifetime of the plane due to modifications. But that compared with one in 100,000 for an Airbus A340. And 50 per cent of the incidents were caused by foreign bodies.
An examination of the bogie on the left main landing gear, which had recently been replaced, revealed that the spacer was missing from the axle. This meant that the bogie could move from side to side. The electrical wiring was long enough not to snap. The hydraulic pipes could have ruptured, but that would only have led to the loss of braking and would not have been a factor in the crash. Usually, the forces on the bogie would bring it back into alignment and no cases of bogie “shimmy” had been reported on Concorde. Besides, examination of the bearings and brakes on the other wheels showed them to be in a normal condition, and there was no evidence of damage to the other tyres.
In theory, the absence of the spacer could have caused the aircraft to deviate from its straight-line path, tyre overheating and slower acceleration than normal. Study of the marks on the runway, as well as calculations of the trajectory and acceleration made on the basis of the data from the flight recorders, show that this was not the case. The path of the plane was straight until the loss of thrust on engines one and two. The temperature of both the right and left bogies was the same and nothing abnormal was noticed when the plane was taxiing. The acceleration down the runway for the first 1,310 yards (1,200 m) was perfectly normal and there were no identifiable tyre marks coming from Concorde before that point.
The plane was slightly overweight, starting its take-off weighing 185,880 kg, while its maximum take-off weight is 185,070 kg. But it was concluded that this had no significant effect on the take-off and acceleration distances.
The investigators concluded that at 1,860 yards (1,700 m) from the threshold of the runway, around where the first parts of the water deflector were found, tyre two ran over the metallic strip. In the following half-second, a clean, short noise was heard on the cockpit voice recorder. The noise was thought to be the sound of the tyre running over the strip. This is where the strip and a large piece of the tyre were found.
At 1,980 yards (1,810 m), the first marks from tyre two were noticeable on the runway. The piece of the bottom of tank five and the first kerosene stain were found at 1,990 yards (1,820 m). At 2,025 yards (1,850 m), the first marks of very dense soot were seen. This led to the conclusion that a large quantity of fuel leaked out before the fire broke out. A change in the background noise was heard, resulting from the ignition of the flame, which was then spotted by the controller in the tower. The heading changed slight, probably due to the burst tyre and the aerodynamic disturbance caused by the fuel leak and the fire.
The captain then turned the rudder slightly to starboard to compensate for the plane’s slight move to port. The flight officer said: “Watch out.”
Engines one and two suffered their first loss of thrust as their “go lights” went out in the cockpit. The captain pulled back on the control column at 2,095 yards (1,915 m). There was a slight yaw to port, then the nose gear lifted off the runway. The rudder was moved further to starboard and the plane side-slipped to port. By then, those on board would have felt a strange sensation of lateral movement due to the lack of thrust. There would have been unusual noises, a luminous glow and a strange smell permeating the cabin.
Engine one’s “go light” came back on. The flight engineer announced the failure of engine two and began the fire-control procedure. Engine two briefly sparked back into life, delivering around 15 per cent of its normal thrust. The “go lights” on engines one, three and four went out – a normal reaction to the relaxation of the shock absorber on the left main landing gear.
Shortly afterwards, engine one suffered a second surge, this time caused by the ingestion of hot gases or kerosene, aided by the aircraft’s peculiar angle of attack. As engine two picked up speed, its auxiliary air intake opened, allowing the further intake of hot gases. This caused a further surge, leaving the aircraft powered only by the thrust from engines three and four.
At this point, the aircraft was about 25 yards (22.5 m) from the runway centreline and wheel six ran over the runway edge light. No components of the light were identified in the debris inside the engines, so it seems to have played no part in the crash.
When the aircraft was airborne, the fire alarm sounded, followed by a warning gong. On the radio, someone, probably among the crew of aircraft waiting to take off, was heard to say: “It’s really burning, eh?”Then a few seconds later: “It’s burning and I’m not sure it’s coming from the engine.”
After three seconds in the air, the flight engineer said: “Shut down engine two.”
The captain called for the engine fire procedure. The thrust lever was moved to the stop position and engine two’s fire handle was pulled. Two seconds later, the flight officer drew attention to the airspeed, which was then 200 knots, when V
2
, or the take-off speed, was 220 knots. Engines one, three and four went into “contingency mode”, which automatically boosts the thrust on the remaining engines if one or more loses power. But engine one took seven seconds to respond due to the solid fragments inside it.
Another three seconds passed before the captain ordered the retraction of the landing gear. The speed was still 200 knots. The radio altimeter indicated 100 ft and the rate of climb was 750 ft a minute. In the next few seconds, the controller confirmed that there was a plume of flame behind the aircraft.
Engine one was then producing 75 per cent of its nominal thrust and the reheat kicked in. The flight officer acknowledged the controller’s message, while the flight engineer drew attention to the problem with retracting the landing gear. The smoke detector went off in the forward lavatories. It was thought that smoke from the left engine that ran the air conditioning circulated into the cabin. The sound of the alarm was recorded on the cockpit area microphone, indicating that the cockpit door was open during take-off, then a common practice on Concorde.
The flight engineer repeated: “The gear.” Then a gong was heard alerting the crew to low oil pressure due to the shutdown of engine two. Two seconds later, the flight engineer again repeated: “The gear.” The red wheel light came on indicating a loss of pressure in tyre two. In this case the procedure requires that the landing gear is not retracted. Nevertheless, two seconds later, the captain ordered “gear retraction”. Three seconds after that, engine two’s fire alarm sounded again, along with its associated gong. The flight officer then said: “I’m trying.” This was thought to have been in response to the captain’s order to retract the landing gear.