On 5 March 2013, the aft-stationed cabin crew of an Airbus A330-300 (D-AIKJ) being operated by Lufthansa on a scheduled international passenger flight from Chicago O'Hare to Munich advised the flight crew after the night normal visibility take-off that they had heard "an unusual noise" during take-off. Noting that nothing unusual had been heard in the flight deck and that there were no indications of any abnormal system status, the Captain decided, after consulting Company maintenance, that the flight should be completed as planned. The flight proceeded uneventfully but on arrival in Munich, it became clear that the aircraft had sustained "substantial damage" due to a tailstrike on take-off and was unfit for flight.
The NTSB delegated the Investigation to the German BFU. The FDR and CVR were downloaded but relevant data from the latter had been overwritten during the subsequent flight. Based on the severity of the damage found, the occurrence was classified as an accident by the BFU.
It was noted that the 57 year-old Captain had a total of 17,963 total flying hours which included 2,729 hours on type. The 39 year-old First Officer, who had been PF for the departure from Chicago, had a total of 9,332 total flying hours which included 5,431 hours on type. The augmenting crew member, a 36 year-old Senior First Officer, was occupying a supernumerary seat in the flight deck for take-off and he had a total of 8,967 total flying hours including 4,186 on type.
It was established in accordance with Standard Operating Procedures (SOPs) that as the TOM was 7½ tonnes below the maximum permitted, a flex take-off thrust setting would be used. Although any take-off below MTOM is normally made with the configuration which gives the best thrust reduction for take-off, flap configuration F2 in this case, since "the crew feared the flaps could be hit by raised snow slush and ice", they chose the alternative 1+F configuration for the Chicago departure. The Investigation "understood" the reasoning for the chosen configuration - less flap deployment - because it would reduce exposure of the lift surfaces to frozen deposits and also noted that guidance on configuration choice is "very general" as "There is no quantification or assessment, e.g. between the different configuration options. Therefore it is difficult for a pilot to assess the risk.". It was also noted that "from the flight crew’s point of view entering the alternative configuration 1+F into the flight performance software did not result in changed speeds (from an F2 configuration) and therefore difficulties during take-off would not have been expected."
The take-off was perceived as normal by the flight crew and the speed profile was accurately followed but examination of the FDR data showed that a tailstrike had occurred "about one second prior to the main landing gear lifting off completely from the runway". It was noted that "the energy of the strike (would have been) transported by struts and cross beams to the cabin floor of the aft galley" where cabin crew were seated.
The Investigation noted that once the decision to continue the flight as planned, the Captain had personally "explained to the cabin crew the measures taken and the decision to continue the flight". Because of the assumption that there had not been a tailstrike, ATC were not advised of the possibility that there might be debris or marks providing evidence of fuselage contact on the departure runway.
It was also noted that “The damages on the tail section of the airplane could be associated to the tailstrike. The area of the pressurised cabin was damaged but not degraded in its stability. There was no pressure loss during the flight.”
On arrival structural damage was found to include:
- damage to frames and stringers over a 4.5 metre length of the lower fuselage
- deformation of the web plates on one of the stringers
- deformation of the cross beams of three frames supporting the aft cargo compartment
- damage to or severance of parts of the mounting elements between the inner fuselage structure and the outer skin
It was noted that "past experience has shown that the A330-300 is generally not very prone to tailstrikes" so a detailed examination of relevant FDR data was carried out. The focus of this work was the period just before and just after the aircraft had become airborne. It was found that exactly at Vr, the PF begun to pitch up to approximately 9° and "approximately one second later, rotation had begun and the nose landing gear lifted off". During this change, the maximum rotation rate reached was 4.2°/second and it was concluded that this was when the tailstrike had occurred. Once the aircraft pitch angle had reached 11.5°, the rotation rate decreased to around 2.8°/second and approximately one second later, with the airspeed at 164 knots and the aircraft pitch angle at approximately 13.5°, the main landing gears lifted off.
In the case of the A330-300, the aircraft pitch angle which will produce tail contact during rotation was found to range from 10.1° with complete compression of the main landing gear oleos to 14.4° with full 'rebound' of the landing gear. Within these limits, there are various factors which are capable of increasing or decreasing the relative risk of a tail strike during take-off. All were considered in relation to the accident take-off and the following were found to have influenced the tailstrike:
The rate of pitch change during rotation recommended by the manufacturer was found to between 2°/second and 3°/second - somewhat less than the rate which had been used and it was evident that this had been a significant contributor to the tailstrike. It was noted that there is no flight deck indication of rotation rate - it is the response of the aircraft which shows the pilot what the consequences his control inputs are.
It was noted that the reduced flap extension of the 1+ F configuration compared to that of the F2 configuration means that less lift surface is available for take-off so the aircraft must have a slightly steeper angle of attack to lift the aircraft into the air from a rotation speed which is almost the same. This then means that there is less distance between tail section of the fuselage and the runway surface which reduces the margin available before a tailstrike occurs. It was concluded that the configuration used had also been a factor which increased risk.
- Shock Absorber Oleo inflation
The nominal increase in tail clearance as the aircraft rotates will be delayed if the main landing gear shock absorbers do not extend normally which requires that the respective oleos are correctly inflated in accordance with AMM procedures. After arrival in Munich, checks found that although both main landing gears contained at least the required amount of hydraulic oil, both shock absorbers were low on gas and gas pressure so that the expected rebound performance was not met which resulted in a decreased in tail clearance.
The aircraft was loaded within the allowable limits for take-off, but both the MACTOW and consequent take-off trim position were indicative of a centre of gravity towards the aft end of the allowable range.
The manufacturer’s analysis of FDR data during the investigation process showed a downward wind gradient of 6kt in 2.5s and a tailwind of 6 kt in 5 s. These gradients were assessed to have produced a lift reduction during rotation
The Cause of the Accident was formally documented as a reduction of clearance due to the following factors:
- Flaps configuration
- Rotation rate dynamic
- Position of centre of gravity and pitch trim
- Reduced rebound of the main landing gear shock absorbers
It was concluded that it must be pointed out that it was the concurrence of these factors which had resulted in the tailstrike - the complete system would have tolerated each individual factor by itself.
Safety Action taken by the aircraft operator included informing all A330 pilots that compared to flap configuration 2, flap configuration 1+F poses an increased tailstrike risk.
Two Safety Recommendations were made as a result of the Investigation as follows:
- that Airbus should improve the indications of a possible tailstrike risk in the documentation, especially in combination with the configuration of the lift surfaces during take-off (and that) the design of the flight performance program should (provide an) alert (to) the cockpit crew whenever the provided flap configuration is not one with the highest level of tailstrike tolerance. [04/2016]
- that Airbus should develop a strategy which allows the crew to unambiguously detect if a tailstrike has occurred once airborne. [05/2016]
The Final Report was completed and published in December 2016.