German investigators studying the cause of an uncommanded pitch-down of a Lufthansa AirbusA321-200 near Pamplona, Spain, in November 2014, hope to discover the probability of similar events linked to frozen angle-of-attack (AOA) sensors as the investigation continues. In the interim, both theEuropean Aviation Safety Agency (EASA) and Airbus have published directives alerting pilots of the issue and the workarounds.
In a new update on the A321 incident, the German Federal Bureau of Aircraft Accident Investigation (BFU) says data from two of the three A321’s AOA sensors locked into a climb attitude following takeoff from Bilbao, Spain, on the flight to Munich. Airbus says the two sensors, which have heated, moveable vanes on the exterior of the aircraft and are used in the automatic stall-protection system, froze about 8 min. into the climb, at approximately 19,500 ft., and remained frozen for more than 90 min. The sensor manufacturer, UTC Aerospace Systems, in a post-incident inspection found no “indications of contamination, damage or defects” either in or on the sensors that would explain the blockage, the BFU says.
Each AOA sensor on A320-family aircraft is connected to an Air Data Reference (ADR), and the three ADRs feed two Elevator Aileron Computers that use voting logic to eliminate any “out-of-family” sensor. Neither ADR1 nor ADR2—the electronics tied to the frozen sensors—were rejected, as values were similar. The measured AOA is compared to an “alpha protection” threshold that decreases as speed, or Mach number increases, and also when the autopilot is disconnected.
The BFU says the co-pilot on the incident flight disconnected the autopilot when he noticed an irregularity in the speed characteristics on the airspeed scale on his primary flight display. When he began to level out and accelerate when approaching 31,000 ft., the static AOA values and increasing Mach number tripped the fly-by-wire aircraft’s AOA-protection limit and automatically injected a nose-down command into the elevator control, a reaction designed to prevent aerodynamic stalls.
The stall-protection action pushed the aircraft’s nose from 4.5 degrees above the horizon to 3.5 deg. below the horizon, resulting in a 4,000-ft.-per-minute plunge, despite the co-pilot pulling the sidestick rearward through more than 60% of its available travel. The aircraft dove from 31,000 ft. to 27,000 ft., at 4,000 feet per min., before a full-rearward stick input by the captain resulted in a net elevator command that leveled the nose. The captain continued to hold “more than 50%” rearward stick in stable flight for a period, but with help from technicians on the ground, the crew was able to reconfigure the automation into the aircraft’s alternate control law, rather than its normal “direct” law. The action removed the alpha-protection checks and canceled the nose-down input. The aircraft then continued to its destination.
Given the unexplained descent that preceded the crash of Germanwings flight 4U 9525 on March 24, there is some speculation that the same problem may have played a part. The sequences of events have notable differences, however. The Germanwings aircraft leveled off in cruise flight for several minutes before the ensuing dive. In addition, Germanwings upgraded the protection system software, and a source with knowledge of the situation says the carrier also replaced the AOA sensors on its fleet.
Regardless, the crew should have been familiar with a workaround to the problem. Airbus in early December issued a Flight Operations Transmission and Operations Engineering Bulletin to all A320-family aircraft operators noting that “if two or three AOA probes are blocked at the same angle, an increase of the Mach number may activate high-angle-of-attack protection,” which results in “continuous nose-down pitch rate that may not be stopped with backward sidestick inputs, even in the full backward position.” The recommended fix is to turn off two of the three ADRs, which puts the aircraft in alternative control law, deactivating the high AOA protection. Information was to be added to the Flight Crew Operations Manual, Aircraft Flight Manual and Quick Reference Handbook. EASA one day later issued an airworthiness directive on the same issue, calling for making the changes to the flight manual prior to the next flight.
Meanwhile, BFU continues to collect and analyze data to help quantify the problem’s scope. The agency says that during the investigation, it received “data analysis results” collected by airlines as part of regular flight-operations data analysis. “These analyses should show how often in the past individual sensors provide inappropriate constant values during a flight,” says the BFU, noting that more work will be required to understand the “algorithms and context” of the data.
“Based on these values it is part of the further investigation to determine the probability of occurrence of a similar serious incident,” the BFU says.