There are few things more dangerous than an airline crew on a non-revenue flight, where crews that are normally tightly controlled are unbounded and given an airplane with almost no supervision and no one in back to complain. But one way to make that dangerous situation even more so is to give that crew the responsibility for a functional check flight.

— James Albright

image

Updated:

2016-07-13

image

A320 D-AXLA,
from Bjorn (Creative Commons)

This crew was given an airplane that had been recently painted and washed, leaving their angle of attack systems corrupted, and this would make their Airbus A320 behave differently than they would expect under "normal law." For a primer on understanding how an Airbus flies, see: Airbus Control Laws.

The next problem was they didn't have formal training or a specified test procedure to follow, encouraging them to improvise. The result was a low altitude stall that wasn't survivable. The lessons write themselves: Functional Check Flights.

1 — Accident report

2 — Narrative

3 — Analysis

4 — Cause

image

1

Accident report

  • Date: 27 November 2008
  • Time: 16:46
  • Type: Airbus A320-232
  • Operator: XL Airways Germany
  • Registration: D-AXLA
  • Fatalities: 2 of 2 crew, 5 of 5 passengers
  • Aircraft Fate: Destroyed
  • Phase: Approach
  • Airports: (Departure) Perpignan Airport, France (PGF/LFMP); (Destination) Perpignan Airport, France (PGF/LFMP)

2

Narrative

  • The A320-232 registered D-AXLA operated by XL Airways Germany had been ferried to Perpignan aerodrome on 3 November 2008 for maintenance and painting work. EAS Industries had issued the approval for release to service on 27 November 2008. The aeroplane, leased from Air New Zealand, was at the end of its leasing agreement and was to be returned to its owner. The leasing agreement specified a programme of in-flight checks; to this end, a flight had been planned for the afternoon. The crew consisted of a Captain (PF) and a copilot (PNF) from the airline XL Airways Germany.
  • A pilot and three engineers from Air New Zealand, as well as a representative of the New Zealand Civil Aviation authority were on board. The pilot had taken a seat in the cockpit. The estimated departure time in the flight plan was 12 h 30 for a total planned flight time of 2 h 35 over the west of France with a return to Perpignan. The flight identification was GXL888T. At the end of the flight, the aeroplane was supposed to return to Frankfurt/Main.
  • The takeoff took place at 14 h 44. A 14 h 47 min 20, as soon as he received the strip relating to flight GXL888T, the CRNA south-west controller contacted the Perpignan approach controller by phone. He wanted to ensure that the crew had the necessary authorizations to undertake what he described as a "disguised test flight". He thought that this flight had not been the subject of an appropriate request by the operator, as had already been the case five hours previously for XL Airways Germany Boeing 737-800, flight GXL032T. The crew of this flight, whose flight plan was identical to that of GXL888T, had contacted this CRNA south-west sector and had also asked on several occasions to be able to perform manoeuvres that had required coordination on several occasions between the different control sectors.
  • Thirty-six seconds later, the copilot contacted the CNRA south-west controller. At 14 h 49 min 20, the Captain asked to perform a "360". The controller explained to the crew that this type of flight could not be undertaken in general air traffic and that the flight plan filed was not compatible with the manoeuvres requested. The crew announced that they wanted to continue on the route planned in the flight plan and asked to climb to FL310.
  • At 14 h 54 min 25, the Captain proposed to the Air New Zealand pilot to delay the check on the flight controls to the approach.

Source: BEA, ¶1.1

We all have this "can do" attitude but items on a functional check flight should be designed in advance and if any flexibility needed, designed into the test plan. Ad hoc ideas are not advisable since it is not easy to think things through in the moment.

  • At 15 h 00 min 54, the crew was cleared to climb to FL 320, which they reached two minutes later. Between 15 h 04 and 15 h 06, angle of attack sensors 1 and 2 stopped moving and remained blocked until the end of the flight at almost identical local angles of attack and consistent with the cruise angle of attack, without the crew noticing it. The local angles of attack 1 and 2 were recorded at 4.2 and 3.8 degrees respectively.
  • At 15 h 05 min 30, the crew began the descent to FL 310, reached one minute later. At about 15 h 12, the crew turned back towards Perpignan. The Captain performed the flight control checks in normal law that had been delayed. While the crew was waiting for clearance to climb to FL 390, the Air New Zealand pilot said that there were not many checks remaining to be performed during the descent. He said that in case of a refusal by the controller, the check on the APU at FL 390 could be done during the flight back to Frankfurt, as the Captain had suggested. The Air New Zealand pilot also described the low speed Configuration Full check without however mentioning the values of VLS and Vmin indicated relative to the weights in the check programme. He added that this check should not be performed on approach but at FL 140. The crew was cleared to climb to FL 390 which was reached at 15 h 22.
  • The descent towards Perpignan was initiated at 15 h 26.
  • At 15 h 33 min 04, the crew was cleared to descend to FL130. Eighteen seconds later, while the plane was descending through FL 180 with a speed close to 300 kt, the Air New Zealand pilot asked the crew if they were ready to talk. He told them that the following check was that of the flight controls in alternate law, scheduled in the programme during descent above FL 140. The Captain decided to perform it at FL 130. The Air New Zealand pilot accepted this.

Source: BEA, ¶1.1

For a discussion about "alternate law," see: Airbus Control Laws.

  • At 15 h 33 min 34, descending to FL 130, the crew contacted Perpignan Approach. They were cleared to descend to FL 120 towards the PPG VOR. The approach controller asked them to reduce speed to 250 kt and to plan a hold at the PPG VOR. They were number two on approach. One minute later, the crew requested radar vectoring. The approach controller asked them to turn left on heading 090 and reduce speed to 200 kt. The check on the ECS supply from the APU was performed before a turn to the 090 heading.
  • At 15 h 36 min 47, when the aeroplane was level at FL120, the Captain asked "you want alternate law" and the New Zealand pilot answered "okay alternate law".
  • Eleven seconds later, the approach controller asked the crew to reduce speed to 180 kt and to descend to FL 80. The crew performed the check on the flight controls in alternate law before beginning the descent.
  • At 15 h 37 min 52, flight control laws in pitch and roll returned to normal law, indicating the end of the check on the flight controls in alternate law, and autopilot 1 was reconnected. The New Zealand pilot then said "Low speed flight is now probably next" then described the sequence of events for the check at low speed Configuration Full mentioned previously. The Captain asked if his intention was to go down to VLS and alpha prot. The Air New Zealand pilot confirmed that and said that, on reaching VLS, it would be necessary to pull quite hard to go as far as the alpha floor function. The Captain answered that he knew. The New Zealand pilot continued, saying: "then you need to pitch forward and err... you're happy with disconnect and reengage. And out of alpha floor". He mentioned neither the altitude nor the limit speeds indicated in the programme.
  • At 15 h 38 min 52, the approach controller asked the crew to descend to FL 60. The aeroplane was then slightly below FL 100 and its speed was 214 kt. Five seconds later, the copilot asked to engage the wing anti-ice system when the aeroplane's altitude was about 9,900 ft. The Captain agreed to this.
  • At around 15 h 40, the approach controller asked the crew to turn to the right on heading 190 and to maintain 180 kt. The aeroplane speed was 215 kt. About a minute later, the approach controller cleared the crew to the LANET- ILS approach for runway 33 and to descend to 5,000 ft. At the request of the crew, she repeated the message. While the Copilot read back, the Captain told the Air New Zealand pilot that the low speed check in landing configuration would have to be done later or during the flight to Frankfurt. He also mentioned the possibility of not performing it.
  • The aeroplane reached 5,000 ft altitude at 15 h 42. Its speed was then around 210 kt. At 15 h 42 min 14, the approach controller asked for the aeroplane's speed. The Copilot answered initially that it was falling, then at 15 h 42 min 25 said that it was 180 kt. The aeroplane speed was then slightly above 190 kt and the selected speed went from 180 kt to 157 kt. The approach controller asked them to maintain 180 kt and to descend to 2,000 ft. The slats and flaps configuration control was placed in position 2.
  • At 15 h 42 min 23, the lateral AP/FD mode changed from HDG to NAV. A few seconds later the aeroplane began to descend.
  • At 15 h 42 min 46, the Captain said that the approach was not in the database. Thirty-six seconds later, the copilot performed the approach briefing. Between 15 h 43 min 20 and 15 h 43 min 55, the spoilers were deployed.
  • At 15 h 43 min 37, the Captain disengaged the autopilot. He said "Down below the clouds so you want what?" The New Zealand pilot answered in a questioning manner "to go slower you mean". The Captain and the copilot responded affirmatively. At 15 h 43 min 41, the Captain positioned the thrust control levers on IDLE and auto thrust disengaged. The New Zealand pilot added "We need to go slow with err recovery from... recovery". The altitude was 4,080 ft and the speed was 166 kt.
  • The Captain called for landing gear extension then at 15 h 43 min 48 said "we do the err the...". The New Zealand pilot answered "low speed yeah".
  • At 15 h 43 min 51, the Captain asked for speed values from the Air New Zealand pilot, who answered "just... to come right back to alpha floor activation".
  • During this time the approach controller twice asked the crew for their intentions. The Copilot answered by saying that it would be a go-around and a departure towards Frankfurt.
  • At 15 h 44 min 30, the Captain stabilised the aeroplane at an altitude of 3,000 ft. The aeroplane was in landing configuration. Flight directors 1 and 2 were still active and the vertical mode changed from OP DES to V/S +0000. The speed was 136 kt.
  • At 15 h 44 min 44, the aeroplane was at an altitude of 2,980 ft and a speed of 123.5 kt (VLS).
  • At 15 h 44 min 57, while the aeroplane was near LANET, a "triple click" was recorded and the AP/FD lateral mode changed from NAV to HDG. The selected heading was the current heading of the aeroplane. One second later the aeroplane was at an altitude of 2,940 ft and a speed of 107 kt (Vmin). Between 15 h 44 min 30 and 15 h 45 min 05, the stabiliser moved from -4.4° to -11.2° corresponding to the electric pitch-up stop. It stayed in this position until the end of the recording.
  • At 15 h 45 min 05, the aeroplane was at 2,910 ft altitude and a speed of 99 kt. Pitch angle was 18.6 degrees. The stall warning sounded. In the second that followed, the thrust control levers were moved to TO/GA. A symmetrical increase in engine RPM was observed up to N1 values of about 88%. The aeroplane began to roll to the left. The Captain countered this movement.
  • At 15 h 45 min 09, the bank angle reached 8° to the left and the speed 92.5 kt. The Captain gave a lateral input to the right and a longitudinal movement forwards on his side-stick.
  • At 15 h 45 min 11, the aeroplane wings straightened up and began to roll to the right. The Captain made a lateral input to the left stop. The rudder pedal began to move in the direction of a left turn.
  • At 15 h 45 min 12, both flight directors disengaged. At 15 h 45 min 14, auto thrust disarmed.
  • At 15 h 45 min 15, the flight control laws, which were in normal law, passed to direct law. Bank angle reached 50° to the right. The Captain's lateral input was still at the left stop. The rudder pedal reached a 22 degrees left position. At the same moment, the Captain's longitudinal input changed to the forward pitch down stop position. Pitch was 11 degrees, the speed 100 kt and the altitude about 2,580 ft.
  • A 15 h 45 min 19, the stall warning stopped. The Captain's longitudinal input was still at the forward stop position. The elevators reached their maximum nose-down position of about 11.6°. The bank angle reached 40 degrees to the left and the Captain progressively cancelled his lateral input. One second later, the aeroplane's pitch was at 7 degrees. Its wings were close to horizontal and its speed was 138 kt. The Captain cancelled his longitudinal input. At 15 h 45 min 23, the pitch and the altitude then began to increase. The altitude reached about 2,250 ft. The Captain immediately made a longitudinal input to the forward stop.
  • This did not stop the aeroplane from climbing, with speed dropping. The stall warning sounded again at 15 h 45 min 36. Three seconds later, the crew retracted the landing gear. At 15 h 45 min 40, the pitch was 52 degrees nose up, the bank angle reached a maximum of 59 degrees to the left and the normal load factor dropped below 0.5 g. The flight control law passed to abnormal attitudes. The Captain's roll input was practically at neutral; the longitudinal input was still forward but was not constantly at the stop. The yaw damper orders were nil and remained so until the end of the flight.
  • At 15 h 45 min 42, the speed dropped below 40 kt.
  • At 15 h 45 min 44, the maximum values recorded were: pitch 57degrees nose up, altitude 3,788 ft.
  • At 15 h 45 min 47, the stall warning stopped. It started again about five seconds later.
  • Between 15 h 45 min 49 and 15 h 45 min 53, the Captain made a longitudinal pitch up input. The elevator reached values of about 30° pitch-up. At 15 h 45 min 50, the normal load factor exceeded 0.5 g. The thrust control levers were placed in the CLIMB position for one second then repositioned on TO/GA.
  • From 15 h 45 min 52, the speed passed above 40 kt.
  • At 15 h 45 min 53, the pitch reached 7 degrees nose down. The bank angle was less than 10 degrees, to the left.
  • Between 15 h 45 min 55 and 15 h 45 min 58, the Captain made a lateral input to the left stop; the aeroplane began to roll to the right. The bank angle went from 3 degrees to 97 degrees to the right. At the same time, the pitch went from 3 degrees nose down to 42 degrees nose down.
  • From 15 h 45 min 57, the Captain's longitudinal input was nose up; the elevator was at 14.5 degrees nose down.
  • At 15 h 45 min 58, the flaps and slats were selected to position 1, then to position 0 two seconds later. The Captain made inputs on the flight controls and thrust levers.
  • At 15 h 46 min 00, the stall warning stopped and was followed by a CRC warning that corresponded to a Master Warning, which stopped two seconds later. At 15 h 46 min 01, the pitch reached a maximum of 51 degrees nose-down. The bank angle was 45 degrees to the right, the speed was 183 kt and altitude about 1,620 ft. From this moment, the Captain's longitudinal pitch input was at the rear stop.
  • At 15 h 46 min 02, the thrust levers were moved back to a position close to IDLE then advanced to the CLIMB position. The EPR on the two engines dropped towards 1.2 before increasing to 1.25.
  • At 15 h 46 min 04, an EGPWS TERRAIN TERRAIN warning was recorded, followed one second later by another CRC warning (Master Warning). The recordings stopped at 15 h 46 min 06.8. The last recorded values were a pitch of 14 degrees nose down, a bank angle of 15 degrees to the right, a speed of 263 kt and an altitude of 340 ft.
  • Between the time the stall warning sounded for the first time and the moment the recordings stopped, sixty-two seconds had passed.

Source: BEA, ¶1.1


3

Analysis

  • Local angle of attack is measured by the sensor and corresponds to the value of the locally existing angle between the relative wind and the reference axis of the sensor. As the fuselage disturbs the flow of air, these measurements have to be corrected to obtain the aeroplane angle of attack.
  • The aeroplane angle of attack (also called true, real or corrected angle of attack) is defined by the angle between the relative wind infinitely upstream and the longitudinal axis of the aeroplane. It is generally noted as α (alpha). Thus we call aeroplane angle of attack the angle of attack deduced from the local angle of attack from the configuration of the slats and flaps.
  • Angle of attack, though significant for the study of the aerodynamic situation of the aeroplane, is not a piloting parameter.

Source: BEA, ¶1.6.6.

This is a strange philosophy for an airplane that relies so heavily on angle of attack to determine flight characteristics.

For more about this, see: Airbus Control Laws.

  • The Air New Zealand pilot had never performed this type of flight. Alone, he carried out two simulator A320 flight sessions by following the OFC document programme before his departure for Perpignan. Between the two simulator sessions, he had discussed Airbus demonstration flights (and in particular the test engineer’s role) with an Air New Zealand pilot who had taken part in several of these flights.
  • The XL Airways Germany crew did not follow any specific training for this flight.

Source: BEA, ¶1.17.8.3.

  • No flight check procedure is defined in the Airbus A320 Maintenance Manual or in the other documents available to operators.

Source: BEA, ¶3.1

This is a violation of the first rule of conducting functional check flights: "no winging it." Procedures should be specified ahead of time with exact entry parameters, limitations, and abort procedures.

  • The programme of checks specified by the contractual leasing agreement was developed by Air New Zealand based on the manual used by Airbus for customer acceptance flights, which are performed by test crews.
  • The crew had licences and qualifications to undertake the flight but did not have the technical skills, the experience, and the methods of a test crew to use this flight programme, even if it was not a test flight.
  • The Airbus Customer Acceptance Manual specifies performing the low speed check in landing configuration at FL 140.
  • The programme of checks developed for the leasing of D-AXLA did not reproduce in an identical manner the altitude range at which the low speed check should take place.
  • The maintenance work was performed or checked in accordance with the approved maintenance programme and by part 66 qualified personnel.
  • The stripping and cleaning procedures for the aeroplane, which include rinsing, specify protection of the angle of attack sensors.
  • In order to eliminate the dust on the fuselage, a rinse with fresh water was performed on Monday 24 November 2008, without following the rinsing task procedure in the aeroplane cleaning procedure, and notably without any protection for the angle of attack sensors.
  • During the rinsing, the angle of attack sensors were not protected. Water penetrated inside angle of attack sensors 1 and 2 and remained there until the accident flight, three days later.
  • The AIP France specifies that flights of a specific nature must be subject to a specific request. Without an advance agreement, the flight can be subject to modifications in real time or possibly be refused if the circumstances require it.
  • The Captain asked the Perpignan ATC service, on the morning of the accident, if the planned flight required specific airspace. The Perpignan controller indicated that it was not necessary as the crew of XL Airways Germany flight GXL032T had been able to follow a flight plan identical to that of the D-AXLA flight without any problems that morning.
  • The crew consisted of two XL Airways Germany pilots. An Air New Zealand pilot, present in the cockpit, participated in an active manner in following the programme of checks.
  • The CRNA southwest controller refused the request for manoeuvres by the Captain given that the flight plan that was filed did not include them.
  • The crew adapted the programme of checks in an improvised manner, according to the constraints of the flight plan and the air traffic control service.
  • Angle of attack sensors 1 and 2 blocked during cruise due to frozen water present inside the casing of these sensors. The system surveillance did not warn the crew of this blockage, which was more or less simultaneous and at identical local angle of attack values.
  • The application of a jet of water onto an aeroplane without following the recommended procedure can allow penetration of a small quantity of water into the inside of an AOA sensor which is enough, when solidified, to block the sensor.
  • The AOA sensors are not designed to be subjected to jets of fluids such as those encountered during de-icing, washing and cleaning operations.
  • The CHECK GW message displayed on the MCDU, the consequence of the gap between the weights calculated by the FAC, on the one hand, based on the angle of attack, and on the other hand by the FMS, based on the takeoff weight and the consumption of fuel, was not detected by the crew.
  • The crew decided, without preparation, and in particular without a call-out of the theoretical minimum speeds indicated in the OFC, to undertake the check of the low speed protections at an altitude of about 4,000 ft.
  • The almost simultaneous blockage of the angle of attack sensors 1 and 2 at identical local angle of attack values rendered the angle of attack protections inoperative in normal law.
  • The limit speeds corresponding to angle of attack protections displayed on the strip (Vαprot and Vαmax) were underestimated and were directly proportional to the computed airspeed, due to the blockage of the angle of attack sensors.
  • The crew waited for the triggering of these protections while allowing the speed to fall to that of a stall.
  • The auto-trim system gradually moved the horizontal stabilizer to a full nose-up position during the deceleration. The horizontal stabilizer remained in this position until the end of the flight.
  • The triggering of the first stall warning in normal law, at an angle of attack close to the theoretical angle of attack triggering the warning in landing configuration, indicates that angle of attack sensor 3 was working at that moment.
  • When the stall warning triggered, the Captain reacted in accordance with the approach to stall technique.
  • The flight control law passed to direct due to the loss of the normal law operating conditions. The auto-trim system was thus no longer available. The changes of law that followed did not allow the auto-trim system to move from the nose-up position.
  • No crew member reacted to the USE MAN PITCH TRIM message.
  • The Captain did not react with any input on the trim wheel at any time or to reduce engine thrust in any prolonged manner.
  • Due to the position of the stabilizer at full pitch-up and the pitch-up moment generated by the engines at maximum thrust, the crew lost control of the aeroplane during the increase in thrust.
  • The aeroplane was completely destroyed on impact with the surface of the sea.

Source: BEA, ¶3.1


4

Cause

  • The accident was caused by the loss of control of the aeroplane by the crew following the improvised demonstration of the functioning of the angle of attack protections, while the blockage of the angle of attack sensors made it impossible for these protections to trigger.
  • The crew was not aware of the blockage of the angle of attack sensors. They did not take into account the speeds mentioned in the programme of checks available to them and consequently did not stop the demonstration before the stall.
    • The following factors contributed to the accident:
      • The decision to carry out the demonstration at a low height;
      • The crew’s management, during the thrust increase, of the strong increase in the longitudinal pitch, the crew not having identified the pitch-up stop position of the horizontal stabiliser nor acted on the trim wheel to correct it, nor reduced engine thrust;
      • The crew having to manage the conduct of the flight, follow the programme of in-flight checks, adapted during the flight, and the preparation of the following stage, which greatly increased the work load and led the crew to improvise according to the constraints encountered;
      • The decision to use a flight programme developed for crews trained for test flights, which led the crew to undertake checks without knowing their aim;
      • The absence of a regulatory framework in relation to non-revenue flights in the areas of air traffic management, of operations and of operational aspects;
      • The absence of consistency in the rinsing task in the aeroplane cleaning procedure, and in particular the absence of protection of the AOA sensors, during rinsing with water of the aeroplane three days before the flight. This led to the blockage of the AOA sensors through freezing of the water that was able to penetrate inside the sensor bodies.
    • The following factors also probably contributed to the accident:
      • Inadequate coordination between an atypical team composed of three airline pilots in the cockpit;
      • The fatigue that may have reduced the crew’s awareness of the various items of information relating to the state of the systems.

Source: BEA, ¶3.2

References

(Source material)

Bureau d'Enquêtes et d'Analyses (BEA), Accident of 27 November 2008 off the coast of Canet-Plage (66) to the Airbus A320-232 registered D-AXLA operated by XL Airways Germany.