Last updated: 20 April 2014
Statuts:Enquête Officielle
Date:vendredi 9 avril 2004
Type/Sous-type:Silhouette image of generic A343 model; specific model in this crash may look slightly different
Airbus A340-313X
Compagnie:Emirates Airlines
Immatriculation: A6-ERN
Numéro de série: 166
Année de Fabrication: 1997
Heures de vol:21800
Moteurs: 4 CFMI CFM56-5C4
Equipage:victimes: 0 / à bord: 14
Passagers:victimes: 0 / à bord: 260
Total:victimes: 0 / à bord: 274
Dégats de l'appareil: Mineurs
Conséquences: Repaired
Lieu de l'accident:Johannesburg International Airport (JNB) (   Afrique du Sud) show on map
Phase de vol: Au décollage (TOF)
Nature:Transport de Passagers Intern.
Aéroport de départ:Johannesburg International Airport (JNB/FAJS), Afrique du Sud
Aéroport de destination:Dubai Airport (DXB/OMDB), Emirats Arabes Unis
Numéro de vol: 764
An Airbus A340-313X passenger plane sustained damage in a takeoff incident at Johannesburg International Airport (JNB), South Africa. There were 260 passengers and 14 crew members on board.
The day before the incident, the crew had operated from Dubai to Johannesburg in an Airbus A340-300 and on that occasion the First Officer was the pilot flying (PF). That flight had been the first occasion either pilot had actually operated, or been in, the Airbus A340-300 apart from their flight simulator experience. The crew were licensed to participate in Mixed Fleet Flying (MFF), but had previously only operated the Airbus A340-500 and the Airbus A330-200. On the take-off from Dubai, both crew members noticed that, whilst the aircraft performed adequately, the Airbus A340-300 seemed to have visibly less performance than other Airbus types they were familiar with.
Emirates night flight 764 back to Dubai was cleared to takeoff from Johannesburg's 4,420m (14,500 feet) runway 21R.
After a normal application of the reduced, flexible, thrust all required calls were made and the aircraft behavior was entirely normal. With the power set, both crew stated that they considered the aircraft acceleration to be slow, in comparison to the other types which they flew (Airbus A330-200 and Airbus A340-500), but that this matched their expectations, in light of their experience the previous day and the Captain's earlier briefing.
The Captain maintained the aircraft on the centreline as it accelerated, with the First Officer monitoring airspeed , the engine instruments and the centreline tracking, etc. During the initial ground run the PF applied a forward stick control input as described in the Flight Crew Operating Manual to counter the nose up effect of setting engine takeoff thrust to about 80 knots. At 75 knots the Side Stick Order Indicator (SSOI) started moving back to the centre position as the PF centralized the sides tick. As the aircraft approached 144 knots the First Officer called "V1" and this was reiterated by the auto call-out. At 152 knots the First Officer called "rotate". The Captain stated that, at this point he went fully onto instruments, to ensure the rotation was as accurate as possible. His understanding of the new technique was to achieve 2/3 back stick, by bringing the stick rearward at a measured rate, over three seconds. This he achieved by cross-referring to the SSOI and finally placing the SSOI at +9° on the Primary Flight Display (PFD) pitch scale, to initiate rotation and to maintain this, to get the aircraft airborne. Thereafter to follow the flight director Speed Reference System (SRS) pitch command.
Recorded data shows that, in initiating the rotation, the Captain applied rearward sidestick such that the SSOI, displayed on the PFD indicated +9° and that, after approximately 4 seconds the aircraft nose started to move upwards at a normal rate.
The Digital Flight Data Recorder (DFDR) data then shows that, over the next 3 seconds and simultaneously with the pitch up, there was a progressive relaxation of the rearward stick input, which had the effect of keeping the SSOI in its original position on the PFD but also decreased the nose up pitch demand. Over the next two seconds, the aircraft pitch attitude stopped increasing with a maximum achieved value of +5.6° and, after a further 4 seconds, the pitch attitude had decreased to +3.5°. With the aircraft main wheels still on the ground, the SSOI remained displayed on the PFD and the Captain continued to control the SSOI at go nose up on the PFD, failing to realize that, to do so, he had moved the stick forward from the original 2/3 back position, thereby lowering the nose and keeping the aircraft on the ground. Keep in mind that the moment the aircraft start to rotate and the nose rise, the horizon indication on the PFD start to sink lower on the PFD display. With the PF attempting to keep the SSOI at a 9° relative to the horizon effectively caused him to relax backpressure on the stick and thus lowers the aircraft's nose.
The next thing that either pilot recalled was that they felt a heavy 'juddering' vibration and noise. Whilst they believed the aircraft was approaching the very end of the runway they thought they were on a rough area of paved surface. Simultaneously the First Officer called "TOGA", which the Captain selected. 2 Seconds after leaving the end of the runway, the aircraft became airborne, the noise and vibration ceased and, with a positive climb indication, the landing gear was selected up. The aircraft had travelled off the end of the runway onto a flat grassy area and had impacted 25 runway approach lights. The lights, mounted on small concrete plinths were approximately 20cm high. The rear wheels of both main landing gear bogies were still on the ground and impacted the edges of several plinths, plus the lights. This immediately caused some, or all, of the tyres on wheels 5, 6, 7 & 8 to burst, throwing tread debris against various parts of the aircraft and damaging some fairings and the flap drive mechanism.
The initial climb out and thrust reduction was normal, but as the aircraft was accelerated and flaps were retracted, an Electronic Centralized Aircraft Monitoring (ECAM) caution annunciated for FLAPS LOCKED. The crew presumed that this had arisen , as a result of the rumbling they had felt on lift off, causing a vibration which caused the flap asymmetry protection system to activate. However, as a precaution they requested ATC to inspect the runway for any signs of damage, or debris. ATC subsequently responded that there was none. In light of this information and in the absence of other warnings, plus no ECAM system for displaying tyre pressure being installed, the crew assumed that the landing gear and tyres were unaffected. The captain contacted the engineering section in Dubai and they requested that the pilots should attempt to move the flaps, but this attempt proved that the flaps were locked although the slats did move. Since the flaps were now immovable and not fully retracted, continued flight to Dubai was not a viable option. A decision was made to return, in the hope that the flap problem could be resolved quickly, to allow the flight to continue to Dubai, without undue delay. ATC were informed and the aircraft was flown to a designated area for fuel dumping down to maximum landing weight. The fuel dumping procedure commenced shortly after the pilots made sure that this was their only option and they had completed the necessary calculations.
ATC now advised the crew that a second inspection had revealed damage to two runway end lights and that the aircraft may have traveled about 150 meters across the grass, at the end of the runway, before becoming airborne. There was however, no sign of any other damage, or debris. During the planning for the approach and landing, even with the report of the second runway inspection, one could hear on the CVR that the pilots was still not comfortable that the aircraft did not suffer tyre damage.
The crew had planned a non-standard configuration (Flaps Locked) for the approach and calculated that, at maximum landing weight a significant safety margin existed, in terms of landing distance required, for any runway at Johannesburg. They requested the use of runway 21L and the ATC cleared them accordingly. The crew was then informed that the wind had changed, which meant that a Northerly landing was preferable and they can plan for a landing on runway 03R, which suited the crew well.
With fuel dumping completed the aircraft was radar vectored for a long final approach to runway 03R. While the crew was briefing for the landing the ATC now advised that a third inspection of the threshold of 03L had revealed some tyre debris in the area of the broken lights. As a result of this information, it became clear that the aircraft would not be able to taxi to the gate, the cabin crew, passengers and ATC were now briefed that the aircraft would stop on the runway, to allow an inspection of the landing gear, prior to taxiing to the parking stand. The ground the aircraft was configured for landing and the approach to runway 03R commenced, with a landing reference speed (VREF) of 148 knots, at a landing weight of 189.5 tons, a calculated landing distance required of approximately 1400 metres, with a runway length of 3400 metres. Low auto brake was selected . As a precaution, and in case any undetected tyre damage existed, the crew requested emergency services on standby for their return .
The touchdown was smooth, but high levels of vibration and noise was immediately apparent and full reverse thrust was applied, with normal spoilers. The vibration was of such magnitude that several covers on the emergency exit lights in the cabin detached and fell to the floor, plus several other lighting units became dislodged. The aircraft decelerated, but at approximately 70 knots normal braking appeared to have
failed. The first officer made a comment about number eight brake and, after he had also attempted using his brake pedals he suggested using the parking brake to stop the aircraft. The Captain did not approve of this suggestion. The aircraft was still moving at 70 knots and, showing no signs of stopping, began to veer towards the left hand side of the runway. To regain some braking, the Captain instructed the First
Officer to select the anti skid/nose wheel steering switch to OFF, thereby allowing accumulator pressure to power the brakes. This also allowed some degree of lateral control to return, using differential brakes, thereby keeping the aircraft on the runway.
With the upwind end of runway 03R rapidly approaching, full braking, without antiskid, was successfully applied. As the aircraft slowed down the crew observed shards of rubber and dust being blown forwards past the windscreen and cancelled the reverse thrust. The aircraft came to a halt approximately 250 meters from the end of the runway. The cabin crew was alerted to prepare for an evacuation. The fire services were in immediate attendance and seven of the main gear tyres were reported as damaged and deflated with several of the wheels also substantially damaged. There was no sign of fire and it was decided to revert to normal cabin operations to await stairs and buses for disembarking the passengers. The passengers were disembarked via the stairs and returned to the aerodrome building.

During the take-off roll the captain applied an improper rotation technique by referencing the Side Stick Order Indicator to the 9° position on the Primary Flight Display. This caused the aircraft to de-rotate and not to lift-off as expected. When the aircraft over-ran the end of the runway with the associated noise, the aircraft was rotated further and became airborne with the application of full thrust.
The training and mindset of the pilots could be considered as a significant contributory factor to this incident. They were "programmed" by tailstrike avoidance information, aircraft difference information and certain expectances to perform the way they did.
A further contributing factor to the cause of this incident could be considered as the differences in take-off performance of the different Airbus aircraft of the operator's fleet.
A further contributing factor to the cause of this incident was that this was the second flight in the actual aircraft of this variant of the Airbus A340 for both flight crew members. However it was the first flight for the captain as pilot-flying.

» SA CAA Serious Incident Report CA18/3/2/0330
» Air Transport Intelligence news


Add your photo of this accident or aircraft

Ce plan montre l'aéroport de départ ainsi que la supposé destination du vol. La ligne fixe reliant les deux aéroports n'est pas le plan de vol exact.
La distance entre Johannesburg International Airport et Dubai Airport est de 6392 km (3995 miles).

Les informations ci-dessus ne représentent pas l'opinion de la 'Flight Safety Foundation' ou de 'Aviation Safety Network' sur les causes de l'accident. Ces informations prélimimaires sont basées sur les faits tels qui sont connus à ce jour.
languages: English Français Nederlands Deutsch Espanol