B3/2008L In-flight structural failure at Taipalsaari on 15 August 2008

An aircraft accident occurred in Taipalsaari on Friday 15 August 2008 at about 13.14 (all times are Finnish local time), when a Piper PA-28R-200 Arrow II, registration OH-PDY, crashed into lake Saimaa near the southern tip of an island named Peräluoto. The aircraft was owned by Malmi Air Ltd and operated by Patria Pilot Training Ltd, a company based at Helsinki-Malmi Airport. The accident site is located about 20 km north of Lappeenranta Airport.

Accident Investigation Board Finland (AIB) appointed investigation commission B3/2008L for this accident. AIB Investigator Esko Lähteenmäki was named investigator-in-charge, accompanied by Air Accident Investigator Tii-Maria Siitonen and investigators Ismo Aaltonen and Hans Tefke as members of the commission.

Mr. Heikki Aalto, Senior Teacher of aeronautical engineering at Tampere University of Applied Sciences, and Major Riku Lahtinen, Chief or Aircraft Maintenance at Karelia Air Command, were invited as experts to investigate the structural failure of the aircraft. Mr. Markku Roschier, Master of Science in Technology, was invited to investigate organisational functions and Mr. Jaakko Kulomäki, Master of Psychology of the National Defence University, to examine crew actions. Rescue operations were investigated and analysed by Mr. Kari Ylönen, Master of Social Sciences. At his own request, Markku Roschier was relieved of his duties as an expert for the investigation commission on 9 September 2009.

The aircraft had flown an instrument training flight from Helsinki-Malmi Airport to Kuopio in accordance with the training programme. The flight from Kuopio to Lappeenranta was an internal progress flight test arranged by the flight school. As the aircraft was approaching Lappeenranta Terminal Control Area (TMA), the pilot requested permission to leave the cruising altitude of 8000 ft (about 2400 m) and fly some exercises according to the progress check flight programme. The purpose of the exercise was to practise upset recovery. Air traffic control gave 3000 ft (900 m) as the lower limit for the exercise.

At about 13.13, some witnesses heard an unusual noise from an aircraft engine, which ended with a loud noise caused by the impact with water. The aircraft was completely destroyed at impact. All occupants of the aircraft – a flight instructor and two student pilots – were killed instantly.

A person who had seen the aircraft crash into water reported the accident to the emergency response centre of South-East Finland at about 13.17. The wreckage and the aircraft occupants were found quickly. After the crash, some witnesses had seen loose aircraft parts falling from the clouds and down into the lake. When the wreckage was recovered, it was confirmed that approximately one half of both wings and the right side of the flying tail (later referred to as the elevator) were missing.

The samples cut off from the wing spars still attached to the fuselage and the balance weight arm which had torn off from the elevator spar were examined at the Technical Research Centre of Finland (VTT).

The Centre was requested to examine the fracture mechanisms, particularly any signs of material fatigue and the directions of forces which had caused the fracture. No signs of metal fatigue or material defects were found in the samples examined.

Clear signs of upward bend of the wing were found in the left wing spar. However, the right wing spar showed no clear signs from which the direction of the loads having caused the fracture could have been determined. The investigation commission considers it likely that the elevator broke off when it was hit by the wing.

The investigation of human factors is based on the model of accident causation developed by James Reason. On the basis of it, a theoretical model describing the accident which occurred during upset recovery is constructed. The model accounts for both the individual actions and situational and organisational factors contributing to the accident.

The aircraft was not equipped with flight recorders. The fixed GPS navigation system did not have any function to record the route flown. However, the flight and the final stages of it were recorded by radar. Radar recordings were analysed with the assistance of radar experts, and a speed and altitude diagram was drawn for the 120 final seconds of the flight. The last reliable radar observation of the aircraft was made 10 seconds before impact with water. Based on the analysis, the investigation commission carried out a mathematical study of how the wings broke off, using the speed and altitude data for the flight path. The calculation showed that when the wings broke, the airspeed must have been at least 155 kt and the load factor about 7 g.

The accident was caused by the flight instructor’s misjudgment when he made the student pilot practise upset recovery. He steered the aircraft into a flight condition from which safe recovery did not allow for any mistakes. The instructor may thus have overestimated his own ability to respond to the student’s mistakes, and the student’s ability to perform the recovery. As a result of a pilot error, the aircraft reached a high speed, and the flight control inputs used for recovery resulted in an acceleration sufficient to break the aircraft wings.

A crucial factor contributing to the accident was that the instructor had made a student perform the recovery exercises on the aircraft type in question only once before, and the student pilot had never practised upset recovery on that type prior to the accident flight.

There were several other contributing factors, of which the most important are listed below:

- Deficiencies in standardisation training, which contributed to the fact that flight instructors followed different procedures and some flight exercises were possibly conducted in a high-risk manner.

- The instructions on progress check flight were not detailed concerning the upset recovery training.

- The flight instructor may not have wanted to intervene in the student pilot’s performance at an early stage, since the student was undertaking a check flight.

- The flight instructor did not perceive an earlier occurrence on a progress flight test as an incident, and consequently did not change the procedure he used.

- The student’s earlier experience of upset recovery had been gained on an aircraft suitable for aerobatic flight. The instructor’s experience as an instrument flight instructor and check pilot on this aircraft type was limited.

- Large number of students and lack of resources have created a time pressure in the flight training organisation. As a result, de-briefing after flights was not sufficient.

The investigation commission made five safety recommendations. One of them was addressed to EASA, two to the Finnish Civil Aviation Authority and two to flight training organisations

The investigation commission recommends that EASA study the possibility of drawing up a proposal for a standard which would suggest that all GPS devices intended for use in aviation have a function that records the parameters of the route flown. Moreover, the memory of such devices should not require a power source to retain the stored data.

The investigation commission urges the Finnish Civil Aviation Authority to require that the SPIC training system include a sufficient number of conventional instruction flights on an aeroplane to improve students’ instrument flying skills.

The investigation commission recommends Patria Pilot Training Ltd to draw up detailed flight training instructions for progress check flights, and advises all flight training organisations in general to prepare unambiguous training instructions and organise regular training sessions which all instructors listed in the organisation’s instructor records are required to attend. The topics and participants of each training session should be documented in the training record.

B3/2008L Report (pdf, 1.05 Mt)