ERA14MA060

The flight crew's excessive elevator input during a rapid descent under night lighting conditions, which resulted in the overstress and breakup of the airplane. Contributing to the accident was an initial loss of airplane control for reasons that could not be determined because postaccident examination revealed no mechanical anomalies that would have precluded normal operation.

The captain and first officer were conducting an international cargo flight in the twin-engine turboprop airplane. After about 40 minutes of flight during night visual meteorological conditions, an air traffic controller cleared the airplane for a descent to 7,000 ft and then another controller further cleared the airplane for a descent to 3,000 ft and told the flight crew to expect an ILS (instrument landing system) approach. During the descent, about 7,300 ft and about 290 kts, the airplane entered a shallow left turn, followed by a 45-degree right turn and a rapid, uncontrolled descent, during which the airplane broke up about 1,500 ft over uneven terrain. The moderately loaded cargo airplane was not equipped with a flight data recorder or cockpit voice recorder (CVR) (although it previously had a CVR in its passenger configuration) nor was it required by Federal Aviation Administration (FAA) regulations. There were also no avionics on board with downloadable or nonvolatile memory. As a result, there was limited information available to determine what led to the uncontrolled descent or what occurred as the flight crew attempted to regain control of the airplane. Also, although the first officer was identified in FAA-recorded radio transmissions several minutes before the loss of control and it was company policy that the pilot not flying make those transmissions, it could not be determined who was at the controls when either the loss of control occurred or when the airplane broke up. There was no evidence of any in-flight mechanical failures that would have resulted in the loss of control, and the airplane was loaded within limits. Evidence of all flight control surfaces was confirmed, and, to the extent possible, flight control continuity was also confirmed. Evidence also indicated that both engines were operating at the time of the accident, and, although one of the four propeller blades from the right propeller was not located after separating from the fractured hub, there was no evidence of any preexisting propeller anomalies. The electrically controlled pitch trim actuator did not exhibit any evidence of runaway pitch, and measurements of the actuator rods indicated that the airplane was trimmed slightly nose low, consistent for the phase of flight. Due to the separation of the wings and tail, the in-flight positions of the manually operated aileron and rudder trim wheels could not be determined. Other similarly documented accidents and incidents generally involved unequal fuel burns, which resulted in wing drops or airplane rolls. In one case, the flight crew intentionally induced an excessive slide slip to balance fuel between the wings, which resulted in an uncontrolled roll. However, in the current investigation, the fuel cross feed valve was found in the closed position, indicating that a fuel imbalance was likely not a concern of the flight crew. In at least two other events, unequal fuel loads also involved autopilots that reached their maximum hold limits, snapped off, and rolled the airplane. Although the airplane in this accident did not have an autopilot, historical examples indicate that a sudden yawing or rolling motion, regardless of the source, could result in a roll, nose tuck, and loss of control. The roll may have been recoverable, and in one documented case, a pilot was able to recover the airplane, but after it lost almost 11,000 ft of altitude. During this accident flight, it was likely that, during the descent, the flight crew did regain control of the airplane to the extent that the flight control surfaces were effective. With darkness and the rapid descent at a relatively low altitude, one or both crewmembers likely pulled hard on the yoke to arrest the downward trajectory, and, in doing so, placed the wings broadside against the force of the relative wind, which resulted in both wings failing upward. As the wings failed, the propellers simultaneously chopped through the fuselage behind the cockpit. At the same time, the horizontal stabilizers were also positioned broadside against the relative wind, and they also failed upward. Evidence also revealed that, at some point, the flight crew lowered the landing gear. Although it could not be determined when they lowered the gear, it could have been in an attempt to slow or regain control of the airplane during the descent. Although reasons for the loss of control could not be definitively determined, the lack of any preexisting mechanical anomalies indicates a likelihood of flight crew involvement. Then, during the recovery attempt, the flight crew's actions, while operating under the difficult circumstances of darkness and rapidly decreasing altitude, resulted in the overstress of the airplane.

Pitch Trim Actuator The linear electro-mechanical actuator was taken to the manufacturer for further examination under NTSB oversight. Examination revealed no evidence of preexisting mechanical anomalies that would have precluded normal operation. Pitch Trim Setting The pitch trim actuator rods had been cut during the wreckage recovery process. Remnants of the actuating rods and the saw cuts were subsequently measured, and together indicated a pitch trim actuator extension of 19.25 inches. Interpolated results yielded a stabilizer incidence of 0.62º-1.02º leading edge up, or airplane slightly nose down trim, consistent with the phase of flight. Aileron Controls In September 2014, M7 Aerospace, the type certificate holder of the airplane, issued two service bulletins instructing operators to inspect the aileron bellcranks, link rods, hinge brackets, and rod end bearings for damage and tightness. On September 19, 2014, M7 Aerospace issued a revised service bulletin instructing operators to inspect the outboard aileron hinge attachments for cracking. The left and right aileron bellcranks and attached link rods were retrieved from the wreckage for examination. The left aileron bellcrank was recovered separated from the left wing but with the aileron link rod and a control tube attached. The bellcrank pivot bolt and both rod end bolts were found installed with the cotter pins intact. There was some wood debris embedded in the head of the cotter pin on the control tube attachment bolt. There was also some mechanical damage on the upper and lower surfaces of the bellcrank adjacent to the pivot bolt. The control tube was removed to facilitate shipping of the bellcrank. The control tube rod end bearing moved freely. The threaded portion of the male rod end on the aileron link rod was deformed and the banjo body on the female rod end (normally attached to the aileron) was fractured and deformed. The spherical bearing from the female rod end was not recovered. The fracture faces on the banjo body were examined under a microscope and had features consistent with overstress separation. The bolt attaching the aileron link rod male rod end was disassembled. The male rod end was examined under a microscope and no evidence of cracking was observed in the banjo body. The spherical bearing on the male rod end was clean and free to rotate with no binding evident. The right aileron bellcrank remained attached to the right wing at its mounting location. The control tube and aileron link rod were attached and the aileron link rod was attached to the right aileron portion that remained. The bellcrank pivot bolt and both rod end bolts were installed with the cotter pins intact. The control tube bolt, pivot bolt, and aileron attach bolt were disassembled in order to remove the bell crank and aileron link rod. The control tube rod end bearing moved freely. The threaded portion of the male rod end on the aileron link rod was deformed. The bolt attaching the aileron link rod male rod end was disassembled. The male and female rod ends were examined under a microscope and no evidence of cracking was observed in the banjo bodies. According to DOT/FAA/AR-00/18, "Development of Supplemental Inspection Report for the Fairchild Metro SA226 and SA227 Airplane," 1.2 Aircraft Description, "Structurally there is little difference between the SA226 and SA227. The primary difference is that the SA227 wing is longer by 10 ft to support higher takeoff weights." Prior SA226/SA227 Accidents/Incidents A search of prior SA226/SA226 accidents and incidents involving an inflight loss of control, included: On August 30, 2004, near Bankstown, New South Wales, Australia: an SA226 with one pilot with seven passengers onboard. With a balanced fuel load noted before takeoff, the airplane was manually climbed in instrument meteorological conditions, leveling off at 16,000 feet, when the pilot noticed that the right wing was "slightly low." The pilot applied left rudder trim and engaged the autopilot; about 2 ½ minutes later the autopilot suddenly disengaged, and the airplane rolled right and entered a steep spiral descent. The pilot was able to recover the airplane at 5,200 feet, and when he did, he noticed that the airplane was "very heavy on the right side," and the right fuel tank was reading 350 kg greater than the left. The investigative report noted that that the outcome suggested that the fuel flow cross flow valve had been open during the flight. When the autopilot could no longer trim against increasing fuel load in the right wing, it disengaged without warning. The airplane was not equipped with a CVR or FDR. (Australia Safety Transport Bureau (ASTB) report 200403209) On May 23, 2005, near Stratford, Taranaki, New Zealand: an SA227 with two pilots onboard during a night cargo flight. The airplane was en route at 22,000 feet, and "the night was dark, with no moon, and the aircraft was probably flying above cloud which would have obscured any ground lights." The crew was balancing fuel between tanks, with the rudder input trimmed to an excessive sideslip, and with the autopilot engaged. Data indicated that autopilot capability was exceeded; it then disengaged, precipitating an upset. During descent, the airplane was overstressed and came apart in flight. The airplane had an operating CVR and FDR onboard. (New Zealand Transport Accident Investigation Commission (TAIC) report 05-0006), On February 8, 2007, over Paris, Tennessee: an SA226, with a single pilot was on a night cargo flight. The airplane was en route at 16,000 feet in visual meteorological conditions. The pilot requested from air traffic control (ATC), a 360-degeee turn to the left, and shortly thereafter, requested a 360-degree turn to the right. He then requested vectors to the closest airport, and advised ATC that he had an asymmetric fuel condition. About a minute later, the pilot transmitted six Maydays, and witnesses saw the airplane descend into the ground in a vertical attitude. The impact crater was about 25 feet deep, the airplane was severely fragmented, and the fuel crossflow valve could not be located. The presence/non-presence of a CVR, FDR or autopilot were not noted in the report. (NTSB Report ATL06FA045) On August 28, 2013, Bankstown, New South Wales, Australia, an SA227, with a single pilot on a night cargo flight. During the initial climb from the departure airport, the pilot reported that the right wing dropped "markedly." The pilot raised the wing and opened the fuel cross flow valve to rebalance the airplane. After the airplane was in trim, he closed the cross flow valve. In cruise, the airplane appeared to be in trim and the pilot engaged the autopilot. About 1 hour later, the pilot disengaged the autopilot and ensured the airplane was still in trim. During the approach, the airplane handled "normally" until about 400 feet above ground level, when the right wing dropped again when the final stage of flap was selected. The pilot raised the right wing and landed without further incident. Subsequent ground checks determined that there was a fuel tank imbalance of about 210 liters (about 55 gallons). (Australian Transport Safety Bureau report AO-2013-196) On April 13, 2015, near North Vancouver, British Columbia, an SA226, with two pilots onboard during day cargo flight. The airplane was about 15 nm north of the departure airport, above a cloud layer, over mountainous terrain, about 2,400 meters (about 8,000 feet) above sea level, when it lost altitude rapidly and experienced an inflight breakup. As the State of Design and Manufacture, the U.S. sent a team to assist in the examination the wreckage; however, the investigation is under the jurisdiction of the government of Canada and is ongoing at the time of publication of this report. (Transportation Safety Board of Canada investigation A15P0081) Autopsies were performed on both pilots at the Instituto de Ciencias