AAIB Bulletin: 4/2018

turbulence in the two years he had worked at Gamston. The instructor said that the pilot was proficient in go-arounds and that they might have practised one the day before the accident, but if not, he believed it was at least within the previous two weeks. He said he was instructing about 20 students at this time. Accident site A loss adjustor attended the accident site (Figure 2). He reported that the aircraft appeared to have hit the ground vertically in a steep nose-down attitude. He checked the flight controls and did not find any defects. The left flap was down, the right flap was up, the flap selector was up but the flap indicator showed 30°. However, the flap lever could have easily moved in the impact sequence and the mismatch between the flap indicator and the right flap showed that the flap mechanism was disrupted and therefore this indication could not be relied upon as a valid pre-impact position. If the flaps had been asymmetric 52 © Crown copyright 2018 All times are UTC AAIB Bulletin: 4/2018 G-BHNA EW/G2017/05/13 as found it would have caused a right roll instead of the left roll that occurred. Therefore it is highly likely that the flap mechanism was damaged in the impact. Figure 2 G-BHNA accident site Meteorological information Wind information was not recorded at Gamston. The nearest airfields, Waddington and Scampton, were reporting respectively 12 kt from 240° and 13 kt from 250° around the time of the accident. The reports did not include a value for gust which meant that any gusts were less than 10 kt above the mean speed. The air ambulance pilot who transported the pilot to hospital described the conditions as a “challenging day”. He landed about 100 m from the accident site about 6 minutes after the accident. He said that conditions were “bouncy and blustery” with “turbulence across the runway” due to the trees on the western side. He estimated the wind at 15 kt, gusting to 20 or 25 kt on the approach end of Runway 21. Recorded data The GPS data recovered from the SkyDemon track log revealed that the approach groundspeed from about 500 ft agl until crossing the beginning of the paved surface varied between 56 and 61 kt. With a headwind component of 10 kt this would equate to an airspeed between 66 and 71 kt. The normal approach speed in a Cessna 152 with full flap is 55 to 65 KIAS. The pilot recalled his instructor telling him to use 70 KIAS on final approach due to the turbulent conditions. 53 © Crown copyright 2018 All times are UTC AAIB Bulletin: 4/2018 G-BHNA EW/G2017/05/13 The GPS groundspeed and altitude where the aircraft crossed the beginning of the paved surface were 60 kt and 190 ft amsl, with a displaced runway threshold elevation of 82 ft amsl. Over the runway numbers, the groundspeed was recorded as 59 kt and the altitude 83 ft amsl but the GPS altitude is not accurate enough to show whether the aircraft touched down; it could still have been at 50 ft agl or more. As the aircraft veered left from the runway, the GPS data showed a brief climb and reducing groundspeed, consistent with the eyewitness accounts. Flying training record The flying school provided a copy of the student pilot’s flying training record. In the remarks section, it stated that go-arounds were practised on 10 September 2016, the day before the pilot’s first solo flight, and then practised again on 30 September 2016. In the subsequent 31 dual flights and 8 solo flights, over a period of seven months, there were no remarks of go-around practice. For the entry on 13 May 2017 (the day before the accident) it stated: ‘Lesson spent revising circuits including flapless and go around.’ Go-arounds are performed at the end of practice forced landings away from an airfield. According to the pilot’s flying training record, practise forced landings (Exercise 16, ‘Forced landing without power’) had been performed on one flight, on 14 October 2016. Radar data for the day before the accident flight There was no recorded GPS data for the dual flight on 13 May 2017, but the radar data was obtained. This data revealed that five circuits were carried out to Runway 21 and the approach and climb-out profiles were almost identical for the first four, with the last one ending in a full-stop landing. The lowest recorded altitudes on each of the first four approaches was 47 ± 50 ft (corrected for QNH) and the runway elevation was between 72 and 87 ft. This data indicated that the first four were probably normal touch-and-go landings but it is possible that at least one was actually a very low go-around, from almost touchdown height, although a higher and earlier initial rate of climb would be expected from a go-around than a touch-and-go. Figure 3 Radar data from G-BHNA’s flight on the day before the accident, 13 May 2017 54 © Crown copyright 2018 All times are UTC AAIB Bulletin: 4/2018 G-BHNA EW/G2017/05/13 Additional information The Cessna 152 flight manual states that: ‘in a balked landing (go-around) climb, the wing flap setting should be reduced to 20° immediately after full power is applied.’ 20° corresponds to two stages of flap. In this instance, the pilot’s recollection was that he was landing with flaps 20°. Analysis There were differences in the reports of the three eyewitnesses to this accident. The pilot at ‘Bravo’ and the A/G operator thought that the aircraft’s left wing had dropped first, while the instructor thought that the aircraft had pitched up first before the left wing dropped. The GPS data showed that the aircraft likely rolled left after passing the runway numbers but the altitude data was not sufficiently accurate to determine at what height this occurred or whether it had climbed first before rolling. If the instructor’s account is correct then it would suggest that a go-around was initiated and that the aircraft pitched excessively nose-up. In this aircraft type, applying full power when the aircraft is trimmed with 20° of flap can cause a significant pitch-up if the pilot does not react quickly by pushing the yoke forwards. Applying full power will also cause a yaw and roll to the left if it is not countered with right rudder. It is thus possible that the pilot allowed the pitch-up to continue until the aircraft stalled, causing the left wing to drop and a rotation to the left. Alternatively, it is possible that the aircraft encountered a strong gust first which caused the left wing to drop, as described by the pilot at ‘Bravo’. The air ambulance pilot stated that the conditions were challenging with turbulence near the approach end of Runway 21. A sudden wing drop, low to the ground, may have startled the pilot and caused him to execute a go-around which he was unable to complete that close to the ground. Helicopter wake turbulence was also considered as a trigger for the upset but none of the eyewitnesses could recall a helicopter operating ahead. The pilot did not recall having practised go-arounds since around his first solo, seven months previously. The flying training record supported his recollection, apart from the entry of go-around practice on the day before the accident. The radar data showed that if a go-around had been practised on the day before the accident, it was very close to touchdown height, followed by a delayed or shallow climb-out. However, the pilot was confident that he had never performed a “low-level go-around” before. Conclusion In this accident, the differences between the eyewitness accounts prevented a clear determination of what happened during the final approach, whether the final manoeuvre was initiated by a gust lifting the right wing, or by an initial pitch-up as part of an attempted 55 © Crown copyright 2018 All times are UTC AAIB Bulletin: 4/2018 G-BHNA EW/G2017/05/13 go-around. In either case, this accident is a reminder of the value of well-taught go-around manoeuvres during pilot training as the low-level go-around can be a challenging manoeuvre to perform, especially for a low-time solo pilot who may not be expecting to do one and who has not practised it many times.