Tailstrike incident (case study)

Citation

(2006), "Tailstrike incident (case study)", Aircraft Engineering and Aerospace Technology, Vol. 78 No. 4. https://doi.org/10.1108/aeat.2006.12778dab.009

Tailstrike incident (case study)

Tailstrike incident (case study)

Keywords: Aircraft components, Air safety

A Boeing 757-200 aircraft was departing from Runway 06L at Manchester Airport, on a flight to Las Palmas in Gran Canaria. The co-pilot was the pilot flying (PF) and during the taker-offroll the commander noticed that he had introduced what he considered to be an excessive amount of into wind aileron for the prevailing conditions. The commander stated that, initially, the co-pilot set about half of the full control wheel deflection to the right but he reduced this as the aircraft's speed increased. On the commencement of rotation the commander watched for any signs of roll but the aircraft appeared to remain wings level. He considered that the rate of rotation was normal until the aircraft had reached 88 nose up but, thereafter, it increased rapidly and he was unable to check the control input. The co- pilot felt that the aircraft was a little “nose light” and the rate of rotation was too high up to 10° nose up. Passing the 10° pitch up attitude he continued to pull the control column back at a rate that he considered was about 2.5°of pitch/s.

The staff in the visual control room of the airfield's air traffic control tower observed that the rate of rotation was somewhat sharper than usual for a Boeing 757-200. They also saw a significant amount of smoke emanate from the rear of the aircraft as it lifted off the runway. At the same time the commander heard a loud bang from the back of the aircraft and the co-pilot stated that he felt a slight bump as the aircraft rotated through an attitude of 12° nose up. The noise was also heard by the cabin crew. The crew of another aircraft which was stationary at holding point DZ1 adjacent to the mid point of the runway, reported over the radio that they too had seen smoke coming from the aircraft's tail, which had seemed close to the runway's surface as the aircraft took off (Figure 1).

Figure 1 Diagram of tail scrape

ATC inquired of the crew as to whether all was well. The commander replied that they thought that the aircraft had suffered a tailstrike and that they intended to return to the airfield. He requested radar vectors and advised ATC that they did not wish to fly above 10,000 ft amsl. The crew completed the abnormal procedure for a tailstrike and, as part of the that drill, depressurised the cabin. By this stage the aircraft was flying level at 5,000 ft amsl. The commander informed the cabin crew of the nature of the problem and of the decision to return to Manchester. He instructed them to prepare for a precautionary landing and told the passengers that they were returning to the airfield because the cabin could not be pressurised.

The aircraft landed on Runway 06R at a weight which was 12 tonnes above the normal maximum landing weight of 89,811 kg.. The aircraft touched down gently and the commander who had taken over the role of PF, was able to use the full length of the runway and minimum braking in order to reduce the load on the brakes.

Aircraft examination

Following the incident the aircraft was taken to a local maintenance facility, where it was later examined by the AAffl. It had been fitted with a Tail Scrape Limiting Device (TLSD) which consisted of an inverted section made of nickel alloy, enclosed and sitting proud of a composite fairing. The TSLD was mounted underneath the aircraft, on its centreline, at structural frame 1743.85 (the first frame aft of the rear pressure bulkhead) and deliberately located so that it would be the first point of contact during a tail scrape.

The TLSD had extensive contact damage, with the inverted section worn down so that it was flush with its fairing. Around the device, the airframe akin had buckled and rivets attached to the frame had pulled away from the external skin. Internally, frame 1743.85 had been buckled in two diametrically opposite areas where stringer 29L and stringer 29R were attached to the frame. This buckling was consistent with an extensive upward force on the TSLD with the load being transferred into the frame, pushing this upward and causing the plastic deformation of the frame and aircraft skin.

Moving aft from the TSLD, the next contact point was at the APU fire extinguisher access door access door located between frames 1862 and 1885. Light scrapes were evident 150 mm aft of frame 1862, with these worsening towards the rear of the aircraft. 420 mm aft of the frame, the paint on the access door was worn away in line with the centreline of the aircraft and over an area measuring 160 mm wide and 90 mm in length. The APU access doors were mounted just aft of frame 1885, with a deflector strip mounted on the frame. The centre of the deflector strip was totally worn away, with scuffing of the airframe akin underneath over a width of about 100 mm either side of the aircraft centreline. The scrape damage continued into onto the two then two APU access doors up to a point 470 mm aft of frame 1885. The APU doors contained several proud roundhead rivets and those along either side of the centre line had been completely worn down, coupled with additional wear of the door down to its metal skin. This damage was worse on the right hand door, with the damage at its widest point some 65mm to the right of the centreline. The left door also suffered similar damage but this only extended 20mm to the left of the centreline.

Flight recorders

The aircraft was equipped with a 25- hour duration FOR, a 30-minute cockpit voice recorder (C VR) and a quick access recorder (QAR), utilised by the operator to support its flight data monitoring (FDM) program. When the CVR was replayed the take-off, approach and landing phases were found to have been overwritten as the C VR power had not been isolated in sufficient time to preserve information relating to the incident.

The FDR was downloaded and data for the entire flight was successfully recovered. Data from the QAR was also recovered. All times quoted are from the commander's clock. At 1,644 h the aircraft taxied onto a magnetic heading of about 0618 and came to a stop with the engines at idle, flaps were at 15° and the horizontal stabiliser was at about 4.6 units, where it remained until the aircraft was airborne. The recorded gross weight was 100,624 kg at the time. The aircraft remained stationary for about 2min before the engine thrust was gradually increased. EPR for both engines stabilised at about 1.63 and the aircraft started to accelerate. During the majority of the take-off role the control wheel position was about 188 to the right and a small amount of let rudder was also applied. During the take-off the airspeed was between 10 kt and 30 kt greater than the groundspeed.

At 1,647.18 h at an airspeed of about 144 kt and a groundspeed of about 120 kt, the control column started to move aft and about 2s later the nose squat switch indicated that the nose gear was no longer compressed. About one second later the airspeed reduced to about 142 kt, however, the ground speed continued to increase. About 4 s after the control column had started to move aft the pitch attitude was at about 10° nose up and the airspeed was about 146 kt, at that tune the pitch rate was about 5.7° per second. The control column continued to move aft and the pitch rate continued to increase. When the pitch attitude was at about 12.5° nose up, a normal acceleration of 1.22 g was recorded. At that tune the main undercarriage truck tilt parameters indicated that the aircraft was on the ground and the elevators were at about 16° trailing edge up. For a short duration the control column continued to move aft and the elevators moved to about 17.6° trailing edge up, before the control column was then moved forward. About half a second later the ah^* ground parameter indicated that the aircraft was airborne, airspeed was about 148 kt and the groundspeed was about 137 kt. During the take-off and rotation phase the aircraft had remained predominately wings level. The aircraft continued to climb until it reached FL047, where it remained until about 1,704 h when the aircraft started to descend and was then configured for landing. The approach and landing were uneventful with touchdown occurring at about 171 Ihrs.

Discussion

The results of the investigation indicate that the tailstrike was a result of the excessive rate of rotation during the take-off, one of the four take-off risk factors for a tailstrike that have been identified by the manufacturer. Rotation was initiated at the correct airspeed but at a low rate. The it increased rapidly, so that 4 s after the control column had started to move aft the pitch rate peaked at about 5.7° per second. At that point the aircraft's pitch attitude was about 10° nose up and its airspeed was about 146 kt. This compared with the recommended rotation rate of 2.5° per second over a 4s period and a lift off pitch attitude of 9.5° nose up. However, having exceeded the recommended pitch rate, the aircraft continued to rotate faster than the manufacturer's and operator's manuals advised. Also the aircraft's airspeed was less than would be expected at that stage of the take-off, by some 10 kt. The FDR data indicated that this was because of changes hi the headwind component which varied 10 kt and 30 kt and caused a non- uniform airspeed acceleration. The manufacturer gives guidance, in his Boeing 757 Flight Crew Training Manual, on the procedure to use during take-offs hi gusty wind and strong crosswind conditions to cater for this situation. At the time, the operator did not include this advice in his procedures but this has been addressed and that guidance has since been added to the operator's Operations Manual.

The commander had been unable to intervene hi tune to prevent the tailstrike when he noticed the rate of rotation increase. The recorded flight data indicated that there had been a cue from the stagnating airspeed hi the last few seconds before lift off which might also have alerted PNF to the gusty conditions, albeit at a very late stage hi the take-off run The operator subsequently arranged for him to receive some simulator training to address the situation that he had been faced with. The co-pilot also received further training.

The crew's reaction to the tailstrike reflected well on ability to handle the consequences of the event and, having never flown together before, to co- operate together and with the cabin crew. The aircraft returned to the airport for an uneventful overweight landing. Appropriate precautions were taken by the crew and airport authorities to guard against the possible danger of overheated brakes before the passengers disembarked from the aircraft.

Operator's actions

The operator implemented the following changes to their procedures:

1. 1.

   In the first three months following the final line check, new co-pilots are to be rostered for sufficient sectors to ensure consolidation of their training and to allow for close monitoring.

2. 2.

   Commanders are to be encouraged to give feedback and appropriate advice to new co-pilots.

3. 3.

   During training, training pilots are to explain the rotations self timing technique and encourage its use.

4. 4.

   Examine the possibility of obtaining trends from FDM recordings and providing continuation training for pilots where necessary.

5. 5.

   The operator has amended his operations manual to include the advice given hi the Boeing 757 Flight Crew Training Manual on the subject of takeoffs hi Gusty Wind and Strong Crosswind Conditions.

Reference AAIB Bulletin 2/2006