Investigation
Accident Investigation Board Norway (AIBN) investigated the disaster. Fifty of the 55 bodies were recovered. The recovered bodies were given autopsies in Denmark. Investigators used side-scan sonar to plot positions of wreckage. The pieces settled over an area 2 kilometres (1.2 mi) wide. This told the investigators that the aircraft disintegrated in the air. The investigators re-constructed 90% of the aircraft.
Some initial speculation in Norwegian press stated that a bomb brought down Partnair. In December 1988 a bomb downed Pan Am Flight 103. In addition the Prime Minister of Norway Gro Harlem Brundtland had used that particular Partnair aircraft on her campaign trips. The Norwegian press believed that the crash was a failed assassination attempt. Witnesses of the crash said that they heard a loud noise as they saw the aircraft fall. The fact that the aircraft had disintegrated in the air gave credibility to the bomb theory.
The cockpit voice recorder (CVR) usually records the final minutes of an accident flight. In the Partnair crash it recorded the start of the flight and stopped shortly before the aircraft took off. From the maintenance records investigators found that, 10 years prior to the accident flight, the cockpit voice recorder was modified so that the CVR would use the aircraft's generator instead of the aircraft's battery if full power was applied for takeoff. Since the generator was inoperative on this flight, this meant that power to the CVR ended as the aircraft took off. An unpaid catering bill and the failed CVR led Norwegian investigators to scrutinize Partnair and the aircraft.
The flight data recorder (FDR), an antiquated analog model, used a rotating metal foil strip and marking pins which scratched the strip. The FDR did not record all of the parameters that it was supposed to record, mainly vertical acceleration readings were missing and heading indications were abnormal. The needle recorded some lines twice, initially confusing the investigators. The team sent the FDR to the American company which manufactured it.
When the aircraft was re-assembled, investigators found small traces of high-powered military explosives. Members of the public believed that the traces could have been from a bomb or from the NATO war exercise "Operation Sharp Spear," which took place on the day of the accident flight near the flight path. Investigators found that the residue was not from a bomb, nor was it from a warhead, as there was not enough of it present. Finn Heimdal, an AIBN investigator, said in an interview that the residue appeared to be more like a contamination than any other possibility. The sea had old munitions as many battles had been fought off the coast of Denmark. Investigators concluded that the aircraft pieces acquired residue from the bottom of the sea. In the AIBN investigations, no signs of explosives were found, and the traces of RDX military explosives were explained by contamination before the accident or due to storage.
Metallurgist Terry Heaslip of the Canadian Company Accident Investigation and Research Inc. examined the aircraft skin from the tail and found signs of overheating, specifically that the skin had been flexed, through a phenomenon known as flutter. This caused investigators to further scrutinize the tail of the aircraft. The investigation team found that the APU, which was in the tail, had melted plastic parts from the cabin present inside the turbine. This indicated that the APU was operating; normally the APU would not be operating when the aircraft was in flight. The mechanic who had inspected the aircraft on the day of the accident flight told the investigators that one of the aircraft's two main generators had failed and that he was not able to repair the faulty generator. The investigators discovered that the APU had been operated to replace one of the main generators, and that one of the three APU mounts was broken. The mount was broken before the accident flight, and people on previous flights had said that they experienced vibration. One passenger of a previous flight said that the accident Convair had vibrated more than the company's other Convairs. Other passengers of previous flights did not mention any vibration. According to AIBN, the decision of having the APU operating in flight, was denoted in the Flight Log.
The two shroud doors on the aircraft tail were not present in the tail. The doors, constructed with an aluminium honeycomb liner, had reflective properties also found in aluminium foil. Heimdal, who had served in the military, knew that the doors would appear on radar. The unidentified objects found at a high altitude by Swedish radar were likely the shroud doors, which had separated from the aircraft tail. The investigators found that the tail failed at 22,000 feet. If the rudder moved in a violent manner, the weights would also move violently and hit the shroud doors. Therefore the rudder had made a violent movement as the accident unfolded.
Partnair said that an F-16 fighter jet was flying at a faster velocity and closer to the Convair than reported in the media. Therefore the jet, which would have broken the supersonic barrier at that point, would have had flown too close and at supersonic speed near the Partnair plane. The pressure wave, resulting from the breaking of the sonic barrier, would have caused the Convair to disintegrate in midair. Flygtekniska Försöksanstalten, a Swedish aviation technology research facility said that there was a 60% chance of this being the cause. The Norwegian F-16 pilot testified that his aircraft was more than 1,000 feet above the Convair. The investigators concluded that the F-16 would have had to have been within a few metres of the Convair to have affected the passenger aircraft; the investigators had no evidence that the two aircraft were that close together. One aspect that had not been written in the final report is that the investigation had doubted radar information that it received and believed that the jet was traveling closer to the Partnair aircraft than the radar information stated; the pilot of the F-16 stated that he believed the radar data was false. After the investigation concluded, the Thoresen brothers filed a lawsuit, but a ruling in the Norwegian lagmannsrett (intermediate court) dismissed this theory in 2004. This F-16 flight has been part of AIBN investigations, and no connection to the accident was found.
The manufacturer of the defective flight data recorder asked an ex-employee, the highest expert regarding the company's flight data recorders, to leave retirement to examine the FDR. The expert concluded that the needle supposed to have been recording the altitude had been shaking so much that it left another mark on the foil. The particular FDR was able to record for hundreds of hours; the expert unspooled the foil and found that the needle had been shaking for months. This told investigators that another component of the aircraft had been vibrating. The investigators charted the vibrations and found that two months before the crash, the vibrations stopped for two weeks. Afterward, the vibrations increased up to the accident flight. Investigators found that during the two-week period in July 1989 the aircraft received a major overhaul in Canada by the airline's previous owner. When the Canadian company made test flights and during the aircraft's first several passenger flights for Partnair, the FDR recorded almost no abnormal vibrations. When the investigators reviewed the maintenance records, they found that, during the overhaul, mechanics discovered wear on one of the four bolts/pins that connected the vertical fin to the fuselage. During the overhaul the mechanic replaced one of the bolts/pins. The vibrations stopped after the bolt/pin was replaced. Investigators found the four bolts/pins that held the tail on to the fuselage. Heaslip found that the three bolts/pins that were not replaced were not authentic parts and were incorrectly heat-treated during manufacture. Therefore, each bolt had 60% of its intended strength. The three unreplaced bolts/pins were too weak for regular usage on the aircraft. After the period without vibration, the tail vibrated for 16 completed flights and the accident flight. The investigation team concluded that the usage of the APU (which had a broken mount) when combined with the vibrating tail (due to the wear and deformation of the substandard bolts/pins and associated sleeves) caused an aeroelastic oscillation that caused the aircraft's primary control surfaces to fail during the accident flight.
AIBN investigations showed the aircraft to have been subject to abnormal vibrations in the empennage for an extended period. During the accident flight, there was with all probability a connection between the operating APU and a rising level of vibrations in the tail section. The investigation documents both the replacement of vertical stabilizer mounting bolts with insufficient properties and the mounting of the APU with a front support of "inferior design and unknown origin".
The disaster investigation team concluded that three of the bolts/pins used to secure the tail section were counterfeit and inferior to the parts that should have been used. The metal in the bolts was not strong enough and failed when resonant vibration occurred in the auxiliary power unit.
Read more about this topic: Partnair Flight 394