Upset Prevention, Part 1
On Jan. 10, 2008, an Airbus A319-114 operated by Air Canada was en route from Victoria International Airport (CYYJ), British Columbia, to Toronto Pearson International Airport (CYYZ), Ontario, with 83 passengers and five crew members on board.
The aircraft was at FL350 and following a United Airlines Boeing 747-400 in level cruise at FL370 over the state of Washington.
U.S. air traffic control cleared the Air Canada aircraft to climb from FL350 to its flight-planned altitude of FL370. As the Airbus was climbing through FL366, 10.7 nm behind the United 747, sharp jolts were experienced by the Air Canada flight, followed by a series of rolling motions.
The crew regained straight and level flight, declared an emergency and diverted to Calgary International Airport (CYYC), Alberta, where it landed uneventfully at 0728.
Eight passengers and crewmembers received minor injuries and three received serious injuries due to falls and collisions with aircraft furnishings.
The investigation by Canada’s Transportation Safety Board provides far-reaching recommendations that deserve much further consideration, especially when combined with the results of an in-depth flight-test program by Airbus into the effects of pilot reactions during sudden wake-turbulence encounters.
The Air Canada aircraft was in a stable, wings-level climb through FL366. Three sharp jolts described as similar to hitting automotive speed bumps were felt. These were followed by a roll to the right of 5.6 deg., which was countered by the autopilot with aileron input.
The aircraft rolled past wings level to 27.8 deg. left, with the autopilot responding. When this roll was near its maximum, the captain, who was the pilot flying, disengaged the autopilot and autothrottles and attempted to correct the roll.
This was followed by four rolls varying in magnitude from a few degrees to a maximum of 55 deg. The aircraft then returned to level flight with minor oscillations.
The captain reacted to the rolls with a total of nine sidestick roll inputs, accompanied by coordinated rudder-pedal deflections. Five sidestick inputs were to full travel of 20 deg. Seven successive rudder-pedal inputs were made, with six cyclic reversals from left to right. Rudder deflection followed pedal inputs with maximum deflection of 6 deg. left and 7 deg. right.
Post-incident analysis revealed that the pilot’s control inputs were 90 deg. out of phase with aircraft motion.
After the autopilot was disconnected, most of the aircraft motion in the roll axis resulted from pilot inputs and lateral accelerations were due mostly to pilot rudder-control inputs.
Rudder-control inputs, which were coordinated with the out-of-phase roll inputs, had a direct relationship with vertical stabilizer loads.
This resulted in subjecting the rear vertical-stabilizer attachment fittings to loads of 129% of limit load, and the rear fuselage fitting to 121% of limit load. Certification standards (FAR Part 25.303) set the “factor of safety” at a minimum of 150% of limit load.
This incident is similar to the American Airlines Flight 587 accident on Nov. 12, 2001, which killed all 260 people aboard the jet and five people on the ground in the New York suburb of Breezy Point.
Shortly after takeoff from JFK International Airport (KJFK) the A300-600R encountered some mild rolling disruptions caused by the wake by a preceding transport.
The NTSB noted that the pilot applied four full alternating rudder inputs, and after the fourth input, aerodynamic loads on the vertical stabilizer exceeded its ultimate design load, roughly twice the maximum load.
The NTSB’s airplane performance study revealed that the first officer’s cyclic rudder-pedal inputs led to increasing sideslip angles that, along with the continued rudder deflections, produced extremely high aerodynamic loads on the vertical stabilizer.
In fact, the four additional full-rudder deflections (left-right-left-right) allowed the buildup of enough sideslip angle to produce aerodynamic loads on the vertical stabilizer that were about two times the loads defined by the limit load design envelope.
The NTSB’s investigation determined that American Airlines’ advanced maneuvers program may have reinforced the first officer’s tendency to respond aggressively to wake turbulence, encouraged the use of full-rudder-pedal inputs and misrepresented the airplane’s actual response to large rudder inputs.
Specifically, the investigation found deficiencies in the program, including a simulator exercise that provided unrealistic portrayals of an airplane response to wake turbulence and significantly suppressed control input effectiveness to induce a large rolling potential that was unlikely to occur with an airplane as large as an A300-600; and a simulator exercise that encouraged the use of rudder in a highly dynamic situation without portraying the large buildup in sideslip angle and sideload that would accompany such rudder inputs in an actual airplane.
Flight Crew Perceptions Are Seriously Flawed
An in-depth review by the NASA ASRS Multi-Engine Turbojet Uncommanded Upsets working group found numerous instances of pilots’ flawed and exaggerated perceptions of an aircraft’s sudden motion.
For example, the working group evaluated the flight data of an event involving a 737 in June 1995. The incident report was initiated by the flight crew reporting an uncommanded upset that produced an aircraft roll of at least 45 deg., according to their “observation.” However, a thorough review of the flight data recorder determined that the aircraft’s actual roll reached a maximum far less than that, only 18 deg.
The FAA Safety Analysis Branch Office of Accident Investigation conducted an investigation of an event that occurred in August 1995 involving a 737-300.
The crew reported encountering rolls of 30 deg. The rolls “scared the flight crew,” which perceived them to be more like a barrel roll. Analysis of the FDR by the NTSB revealed that the aircraft actually rolled 3 deg. past wings level in one direction, followed by a second roll of 9 deg. past wings level in the other direction, followed by a third roll in the opposite direction of 9 deg. past wings level.
The NTSB’s Office of Research and Engineering issued its report on Feb. 12, 1996, indicating that the subject aircraft encountered wake vortices from a preceding airplane.
The abundance of data reviewed from similar events by the NASA ASRS Multi-Engine Turbojet Uncommanded Upsets working group found that this has occurred on more than just a few isolated occasions. In fact, the consistency of the data is clearly indicative of the limitations for pilots to accurately understand an aircraft’s sudden unexpected motion. Flight crew reactions during the early seconds of these sudden incidents are causing unexpected motions that seriously worsen the aircraft’s motion.
In Part 2, we’ll discuss the lessons learned from Airbus’ wake turbulence flight-test program.