The Ethiopian Aircraft Accident Investigation Bureau released its Preliminary Report in connection with the ongoing investigation into the crash of Ethiopian Airlines Flight 302. A copy of the Preliminary Report can be found here.
The Report raises serious issues with Boeing’s design of the 737 Max fleet, especially with respect to the aircraft’s ability to interpret and properly process airspeed information, the aircraft’s angle of attack sensors and the automatic anti-stall system known as the M.C.A.S. Indeed, given the similarities between the two recent 737 Max aircraft tragedies, namely that improper airspeed indications, angle of attack sensors and the M.C.A.S. played a role in each crash, we believe that the aircraft must remain grounded until the “fixes” proposed by Boeing are thoroughly vetted and undergo stringent independent analysis to ensure that no tragedy like this ever occurs again. Further, we feel it is important to fully investigate and analyze Boeing’s role in the certification process that allowed these defects to make their way to the flying public.
The Preliminary Report into the Flight 302 accident investigation focuses on information obtained from the aircraft’s black boxes, the Cockpit Voice Recorder (“CVR”) and the Digital Flight Data Recorder (“DFDR”). The DFDR indicates that shortly after takeoff the angle of attack sensor located on the left or Captain’s side of the aircraft recorded inaccurate values which caused the Captain’s stick shaker to activate. The Angle of Attack sensor measures the angle between the relative wind and wing (usually defined as the angle between the chord line of the wing or airfoil, and the direction of the relative wind flow that is “not disturbed”). An aircraft, or wing, always stalls at the same angle of attack. The best way to think of angle of attack is that it is the angle between where the wing is “pointing” and where it is “going.” The stick shaker is a stall warning device that causes the control column to vibrate violently to alert the flight crew that the aircraft believes it is about to enter an aerodynamic stall, which always occurs at the same critical angle of attack for a given wing/aircraft. The left side angle of attack sensor immediately increased from approximately 11 degrees shortly after takeoff to 35 degrees and then within ¾ of a second to 74.5 degrees, erroneously indicating that the aircraft was in a near-vertical, severe nose high attitude. During this time, the right angle of attack sensor that is located on the First Officer’s side of the aircraft recorded relatively constant values ranging from 14.9 degrees to 15.3 degrees. In addition, airspeed, altitude and flight director values on the Captain’s flight instrumentation deviated significantly from the corresponding indications on the First Officer’s display.
During this time the auto-pilot was engaged, and flaps were retracted to the 0-degree position. Shortly after the flaps were retracted the auto-pilot automatically disengaged, likely as the result of the continued activation of the Captain’s side stick shaker, and the Captain directed the First Officer to contact air traffic control to advise that they were experiencing flight control problems.
Shortly after auto-pilot disengagement, the M.C.A.S. caused the aircraft to pitch nose down erroneously determining that the aircraft was still in a dangerous nose high attitude and was about to enter an aerodynamic stall. The M.C.A.S. derives its information from the Captain’s side angle of attack sensor only which, as mentioned, erroneously indicated that the aircraft was essentially flying vertically at approximately 74.5 degrees nose up. The M.C.A.S. automatically imposed aircraft nose down forces for approximately 9 seconds after its initial activation. After the M.C.A.S. stopped exerting nose down forces, the flight crew activated the electric trim switch located on the control yoke, causing the aircraft to pitch nose up.
Despite the fact that the crew had caused the aircraft to pitch nose up, the M.C.A.S. again activated, causing an automatic nose down movement. The Captain’s angle of attack sensor was still erroneously indicating an abnormally high nose up attitude so the M.C.A.S. activated to prevent what it incorrectly believed was an imminent aerodynamic stall. After the M.C.A.S. finished exerting nose down forces as a result of this second activation the flight crew manually disconnected the electric power to the aircraft’s stabilizers to prevent the M.C.A.S. from exerting any additional automatic nose down force.
For the next two and a half minutes the flight crew manually maintained back pressure on the control yoke to continue a nose up attitude. During this time, for reasons that are not yet known, the aircraft’s speed was abnormally high. This excess speed was possibly due to a malfunction of an auto-throttle system, which also processes information from the angle of attack sensors and, given the erroneous AOA readings, would have attempted to keep power at a maximum. The crew continued to attempt to manually trim the aircraft but likely as a result of travelling too fast the manual trim was not working.
At this time the crew requested a return to the airport and air traffic control approved their request. While the crew was attempting to return to the airport and approximately 32 seconds prior to impact, the DFDR again recorded automatic aircraft nose down forces as a result of M.C.A.S. activation. It is likely, although still not known, that as a result of being unable to manually trim the aircraft, the flight crew re-engaged the electrical power to the aircraft’s stabilizers which in turn allowed the M.C.A.S. to exert nose down forces from which the flight crew could not recover. The Preliminary Report, however, does not indicate that the flight crew actually re-engaged the electrical system, and this will undoubtedly be a focus of the accident investigation.