15 Aug 2019 - Final report published into Korean 737-900 HL7725 737-900 tailstrike on go-around at Kansai on 9 Apr 2018
On 9 Apr 2018 Korean Airlines 737-900, HL7725 (29999/1512), suffered damage on the lower aft fuselage when making a go-around after a bounced landing on runway 06L at Kansai International Airport at around 21:33 JST (12:33Z). There were 99 people in total on board, consisting of the PIC, seven other crew members, and 91 passengers. No one was injured..
RJBB 091230Z 01003KT 9999 FEW030 BKN045 13/08 Q1020 NOSIG=
On 14 Aug 2019 the Japan Transport Safety Board published their final report into the incident. It identified the probable cause as:
"In this accident, it is highly probable that the lower aft fuselage of the aircraft was damaged with contacting the runway because its pitch angle became too high during the go-around following the bounce at the time of the landing. Regarding the pitch angle became too high, it is somewhat likely that because the Captain, who thought the impact after the bounce would become hard and tried to avoid the second touchdown, performed large nose up maneuver."
For details and analysis of B737 tailstrikes follow this link.
*** Updated 23 Nov 2020 ***
Fig 2: FDR Data of tailstrike
History of the Flight
According to the statements of the Captain and the first officer (hereinafter referred to as “the FO”), the records of the flight data recorder (FDR) and the cockpit voice recorder (CVR), and the records of ATC communications, the flight history was summarized below.
At 20:24 Japan Standard Time (JST: UTC+9 hours, unless otherwise stated in this report all times are indicated in JST on a 24-hour clock), on April 9, 2018, Boeing 737-900, registered HL7725 operated by Korean Air Lines Co., Ltd. (hereinafter referred to as “the Company"), as the Company’s scheduled Flight 733, took off from Jeju International Airport (the Republic of Korea) bound for Kansai International Airport (hereinafter referred to as “the Airport”). The Captain sat in the left seat as PF*1 and the FO sat in the right seat as PM*1. The landing briefing commenced at around 20:59 prior to the descent did not include information regarding a tailwind at the time of the landing.
The aircraft was instructed to fly directly to BERRY (see Figure 1) via NALTO when descending the route prescribed in the standard instrument arrival. The aircraft was descending with receiving a tailwind. The wind at an altitude of 4,000 ft where the Aircraft started the final approach on the ILS approach for runway 06L of the Airport was about 20 kt in tailwind. The wind information (see 2.5(1)) provided by an air traffic controller (hereinafter referred to as "the Controller") of the Airport after the aircraft had passed over BERRY was 030 º in wind direction and 30 kt in wind velocity. Both the autopilot and the auto-throttle of the aircraft were disengaged at around a radio altitude of 1,200 ft. In addition, the aircraft continued a stabilized approach in the tailwind afterwards, and it was about 5 kt tailwind at an altitude of 1,000 ft in the Captain’s memory. The Captain, who was assuming that a landing would be made in the tailwind, planned to put the thrust levers to their idle position earlier than usual with performing a flare (nose up maneuver to reduce the rate of descent) in order to prevent that a touchdown would be long down on the runway. The Captain memorized that the wind information provided by the Controller along with issuing the landing clearance had been 3 kt in crosswind (in fact, wind direction was 030 º and wind velocity was 3 kt) and also stated that the wind at the vicinity of the runway threshold had been almost calm.
The FO was feeling that the approach was stabilized, except that the engine thrust of the aircraft had been set lower than usual. The FO thought the stabilized approach would continue afterward because the captain had set the normal engine thrust after the autocall of a radio altitude of 100 ft.
At around 21:32:54, the Captain moved the thrust levers to their idle position along with initiating the flare at 2 º pitch angle at a radio altitude of about 30 ft. Although the captain tried to continue raising the nose and to reduce the rate of descent, the timing of such maneuvers was slightly delayed from the captain’s assumption. Reducing the rate of descent was infeasible because the thrust levers had already been set to their idle position. The Captain tried to reduce the rate of descent of the aircraft by pulling the control column further.
The FO felt that the amount of the flare the Captain was operating was somewhat small. The FO, who felt that the intervals of the auto-call made at every 10 ft at a radio altitude of 30 ft or below were short and the rate of descent was large, pulled the control column to reduce the rate of descent without making any call-out. Having noticed the operation of the FO, the captain kept the control column so as to follow the FO’s operation.
At around 21:32:57, the right main landing gear of the aircraft touched down at pitch angle of about 3.5 º (Fig. 2 <1>), and all spoilers began to deploy when the auto speed brake was activated. Subsequently, after the left main landing gear had touched down, the aircraft bounced. The maximum vertical acceleration recorded in the FDR during this period was 1.87 G.
The captain, who was unable to predict the degree of the bounce and assumed that the impact accompanied by the touchdown after the bounce would be hard, executed a go-around maneuver. The pitch angle of the aircraft immediately before executing the go-around was about 5 º. The aircraft started climbing positively at about 10 pitch angle after its both main landing gears touched down again (Fig.2<3>) from its right main landing gear at about 7 pitch angle (Fig.2<2>) approximately one second after it had executed the go-around (approximately two seconds after its right main landing gear made a first touchdown). The Captain recognized that the FO was controlling the movement of the control column though the Captain had no memory of the pitch angle in this period.
The FO came to notice that the pitch angle was high, and tried to control the movement of the control column uttering something to the Captain, however, the FO did not remember when he did so. Moreover, CVR did not verify the words the FO claimed to have uttered to the Captain.
The aircraft landed on runway 06L conducting the ILS approach again after flying in accordance with the missed approach procedure.
Both the PIC and the FO did not recognize that the aircraft had struck the runway until the scratch marks were found on the lower aft fuselage by mechanics after the aircraft arrived at the allotted parking spot.
This accident occurred on runway 06L (34º 25’ 52” N, 135º 12’ 36” E) of Kansai International Airport at around 21:33 on April 9, 2018.
PIC Male, Age 45
FO Male, Age 33
Damage on the aircraft
Scratch marks approximately 210 cm in length and approximately 36 cm in the maximum width were found on the skin of the lower aft fuselage including cracks. In addition, the tail skid was broken.
Figure 3 ; Damages on the Aircraft
Analysis of Findings
History until Bounce
The Captain started a landing briefing at around 20:59. It is probable that the meteorological information which the Captain referred to at that time was the observation data issued at 20:30. It is probable that the Captain did not perform the briefing regarding a tailwind at the time of landing in view of the wind direction of 140 issued at 20:30. However, it is highly probable that the Captain assumed that the landing would be made under tailwind conditions because the aircraft was descending in the tailwind and was receiving the tailwind continuously during the final approach for the ILS approach. Meanwhile, the wind direction which the Controller provided with the aircraft was 030º. It is probable that the Captain was able to predict that wind conditions were changing as the aircraft was descending because the Captain recognized the tailwind of about 5 kt at an altitude of about 1,000 ft and almost calm wind in the vicinity of the runway threshold. However, it is probable that initiation of the flare along with reduction of the engine thrust the Captain performed, assuming that the landing would be made in tailwind conditions, followed by insufficient raise of nose up made the descent rate higher than the Captain’s assumption. It is probable that the Captain was required to control the aircraft so as to cope with the changing wind conditions. It is probable, at that moment, that the aircraft touched down when its attitude was changing to nose up direction because the FO, who felt that the descent rate was high, pulled the control column. It is probable that the aircraft bounced because it touched down when its descent rate was high and its attitude was being changed to the nose up direction.
It is highly probable that the Captain executed the go-around because the Captain was unable to predict the degree of bounce. Both the POM and the FCTM prescribe that the lower aft fuselage contacts the ground at a pitch angle of 8.2 º or greater at the moment of the touchdown. The FDR records indicate that the pitch angle varied from approximately 7º to approximately 10º during the time from a second touchdown of the right main gear after initiating the go-around to the lifting off. During this period, it is highly probable that the lower aft fuselage of the Aircraft was damaged with contacting the runway because its pitch angle became too high exceeding 8.2 º. The Captain and the FO stated that the FO, who had noticed the high pitch angle after initiating the go-around, tried to restrict the movement of the control column uttering something to the Captain however, it was not possible to verify the words the FO had uttered in the CVR records. Regarding the pitch angle became too high, it is somewhat likely that because the Captain, who thought the impact after the bounce would become hard and tried to avoid the second touchdown, performed large nose up maneuver. The training guide of the Company prescribes that a second touchdown should not be attempted to avoid if a go-around is executed after a high bounce, aircraft is not damaged as far as it maintains its attitude even if the second touchdown has occurred, and a pitch angle is to be verified with the PFD during the recovery. It is somewhat likely that the Captain was unable to apply the training guide information and simulator training experience to actual situation even if he had received Bounced Landing Recovery training during the simulator training to promote to a captain. Moreover, it is somewhat likely that the fact that the go-around was initiated when the attitude of the Aircraft was changing by the nose up maneuver immediately before the touchdown and the spoilers were deploying contributed to the excessive pitch angle.
Response of PM
At the time of occurrence of this accident, it is probable that the FO, who was the PM, judged the descent rate after initiating the flare was large and subsequently pulled the control column immediately before the touchdown without a call-out to avoid the hard landing. The FOM of the Company prescribes that PM calls out the situations to PF in case that aircraft has deviated or will possibly deviate from a flight path, and in case of no response from PF, PM takes appropriate actions including taking over. It is probable that the FO should have called out “FLARE” or “GOAROUND” at first at the very moment the FO noticed that the descent rate after the Captain had initiated the flare was large as prescribed in the FOM and the POM considering it is somewhat likely that ambiguity over either PF or PM is operating independently could lead to a possible threat to the safety of the flight if PM intervened an operation without a call-out as in the case like this accident.