McDonnell Douglas DC-9-87 (MD-87) N987AK

Accident Sequence The captain (who was the pilot flying) initiated the takeoff roll, and the airplane accelerated normally. According to the cockpit voice recorder (CVR) transcript, the first officer made the “V1” and then “rotate” callouts. According to the captain (in a postaccident interview), when he pulled back on the control column to rotate the airplane, “nothing happened,” and the control column felt like it “was in concrete” and “frozen.” The CVR captured that the first officer subsequently made the “V2” callout, then the captain said “come on” in a strained voice. Both pilots recalled in postaccident interviews that they both attempted to pull back on the controls, but the airplane did not rotate. The CVR captured that the first officer called out “abort.” The first officer pulled the thrust levers to idle and applied the brakes, and the captain deployed the thrust reversers. (See “Execution of Rejected Takeoff” for more information.) The airplane overran the departure end of the runway and continued through the airport perimeter fence and across a road, striking electrical distribution lines and trees before coming to rest in a pasture, where a postcrash fire ensued. The pilots, two additional crewmembers, and all passengers evacuated the airplane. Two passengers received serious injuries, and one received a minor injury. (See “Emergency Evacuation” for more information.) Postaccident examination of the airplane and the related flight data recorder (FDR) data revealed no evidence of preimpact malfunction of the engines or thrust reversers that would have precluded their normal operation. Examination of the elevators and a review of FDR data for elevator position determined that both elevators were jammed trailing-edge-down, which prevented the airplane from rotating during the takeoff roll. Jammed Elevator Condition The investigation determined that, at some point during the 6 months since the airplane was last flown, the inboard geared tab linkages for both elevators had moved beyond their normal range of travel into an overcenter position, resulting in the jammed condition of the elevators in the trailing-edge-down position. No evidence of any other mechanical malfunction, elevator or pitch control system failure, structural failure, or actions by the flight crew or maintenance personnel was identified that could have resulted in the jammed condition. Further, the jammed condition was not detectable during the flight control check the first officer performed during taxi; the elevator control system design is such that, even with this type of jammed elevator condition, the control column feel and travel would be normal during taxi (when the aerodynamic forces on the elevator control tabs would be minimal). Since the airplane was last flown, it was parked outside at the departure airport and exposed to two significant high-wind events: the passage of a squall line about 5 months before the accident (with gusts to 46 kts reported at the airport) and a tropical storm about 1 month before the accident (with gusts of 35 to 45 kts reported over a 5hour period). The possibility of elevator jamming on DC-9/MD-80 series airplanes as a result of exposure to certain high-wind conditions while parked is known and evidenced by two previous events – a rejected takeoff event in 1999 in Germany and a runway overrun accident in 2017 in Ypsilanti, Michigan. The NTSB’s investigation of the Ypsilanti accident determined that the airplane’s right elevator became jammed. Although The Boeing Company (the type certificate holder for the airplane) indicated that the MD-80 series airplane was designed to withstand a 65-kt horizontal ground gust from any direction while parked or taxiing, the jamming occurred even though the highest reported wind gust was 55 kts during the time that the airplane was parked. The investigation included a wind simulation study that determined that the airflow at that airplane’s parked location was affected by the presence of a large hangar (located upwind of the airplane) that generated localized turbulence with a dynamic, vertical component. Based on the wind simulation information, the NTSB developed an elevator test plan that determined that the vertical gust loads at the Ypsilanti accident airplane’s parked location were sufficient to enable the inboard geared tab linkages for the right elevator to move into an overcenter position and jam the right elevator. The investigation found that the airworthiness standard for transport-category airplanes specified that the airplanes must be designed for the limit loads generated when subjected to a 65-kt horizonal ground gust; however, the version of the standard that applied to MD-80 series airplanes allowed for the assumption of only static loads and did not require consideration of dynamic, vertical wind components. (In 2019, the NTSB issued a safety recommendation to the Federal Aviation Administration [FAA] related to the standard. See “Previously Issued Safety Recommendations” for more information.) The accident airplane (like the Ypsilanti accident airplane) had been parked near a hangar. Although the highest reported gust or sustained wind at the airport never exceeded 46 kts while the accident airplane was parked, the localized wind conditions in the immediate vicinity of the parked airplane may have differed from the wind conditions detected by the nearest weather sensor in speed or dynamic characteristics, or both. The presence of localized differences is further supported by the fact that another DC-9 (MD-87) airplane that was parked on the ramp near the accident airplane since the accident airplane had last flown did not sustain jammed elevators. Thus, the accident airplane’s jammed elevators resulted from the airplane’s exposure to high-wind conditions while parked, which likely included localized turbulence with a vertical component. Flight Crew Procedures and Training for Exterior Inspections of Elevators In 2019, the NTSB issued Safety Recommendation A-19-2 (as a result of its investigation of the Ypsilanti accident), which recommended that Boeing develop new preflight procedures or other mitigations for DC-9/MD-80 series airplanes that will enable a flight crew to verify before takeoff that the elevators are not jammed. In response to this recommendation, in 2020, Boeing published Operations Bulletin 80-2-017, “ELEVATORS NOT JAMMED VERIFICATION” and Temporary Revision 80-2-153 to the MD 80 Flight Crew Operating Manual (FCOM), both of which included a warning stating that, before every flight, the flight crew must confirm that the elevator surfaces are not jammed in the trailing-edge-down position. According to the warning, which was added as an update to the FCOM’s Exterior Inspection Procedures, confirmation involves visually verifying that the elevators are faired (even) with or above the stabilizer surface. This flight crew verification applied to every flight, regardless of the airplane’s ground wind exposure. The bulletin also explained that, for the previous known jammed elevator events, the control column feel and travel were normal during the control checks the crews performed during taxi. As a result of Boeing’s publication of this bulletin and revised procedures, the NTSB classified Safety Recommendation A-19-2 Closed—Acceptable Action. During postaccident interviews, the captain and the first officer indicated that they were unaware of the elevator inspection procedure. Although both elevators were visibly trailing-edge-down when the first officer performed a preflight inspection of the airplane (as was evident in a photograph he took of the airplane that morning), he did not recognize the condition as anomalous. Although the flight crew accepted the 14 CFR Part 91 accident flight as contract work for the operator, 987 Investments LLC, they were trained by and used the airplane manuals and procedures from their primary employer, Everts Air Cargo, a 14 Code of Federal Regulations (CFR) Part 121 cargo operator. Although Boeing had distributed Operations Bulletin 80-2-017 to Everts through its MyBoeingFleet system, and Everts was required (per 14 CFR 121.141) to keep its airplane flight manuals current, a review of Everts’ MyBoeingFleet activity data showed no evidence that any Everts personnel had viewed or downloaded the bulletin before the accident. As a result, the company had not updated the FCOM with the revised warning or updated its pilot training materials to include the new preflight exterior inspection procedures for visually confirming that the elevators are not jammed. In a postaccident interview, Everts’ director of operations stated that he was unaware of the Boeing operations bulletin until after the accident. Following the accident, Everts updated its manuals and developed a detailed pilot training presentation that included photographs and a video to show the visual difference between the faired and trailing-edge-down elevator positions when viewed from the ground. Execution of the Rejected Takeoff According to Everts’ procedures, the captain was responsible for deciding, declaring, and initiating a rejected takeoff. Before the takeoff, the captain briefed the first officer on the rejected takeoff criteria, stating that they would reject after V1 only if the airplane would not fly. (This procedure is consistent with longstanding FAA and industry guidance indicating that, generally, a takeoff rejected after V1 will result in a runway overrun.) Based on the NTSB’s airplane performance study and a review of the FDR data, the airplane’s lack of rotational response did not become apparent to the captain until after V1. The CVR transcript showed that the first officer made the “rotate” callout at 0959:48.0 (which was about 1 second after the “V1” callout). The FDR data showed that a change in the control column position began about 1 second after the “rotate” callout, consistent with the captain beginning his attempt to rotate the airplane. FDR data from the accident flight and the airplane’s two previous takeoffs showed similar control column position behavior for all three flights. The data showed that, during the two previous takeoffs, the airplane’s nose-up pitch response began about 2 to 4 seconds after the control column movement. However, during the accident flight, the control column response felt abnormal to the captain, and the airplane’s pitch did not increase. Following the first officer’s “rotate” callout, the captain pulled back on the control column (and was joined briefly by the first officer) before the first officer called out “abort” about 4 seconds later. Although, procedurally, the captain should have been the one to call for and initiate the rejected takeoff, the first officer recognized that the airplane was not going to fly and appropriately took action. Human performance research has shown that the average reaction time to an unexpected driving event is about 1.5 seconds. However, stress and increased task demands associated with an unexpected emergency (such as the abnormal control column feel and the airplane’s failure to rotate as usual) can increase a pilot’s reaction time and degrade a pilot’s ability to accurately assess how to respond. About the time that the first officer made the “abort” callout, pulled the thrust levers to idle, and applied the brakes, the airplane was traveling at 150 kts with only 1,500 ft of runway remaining (and a 600-ft runway safety area beyond that). The airplane reached a maximum speed of 158 kts at 0959:55 (about 2 seconds after the “abort” callout) before it began decelerating. Based on Boeing’s calculations, at this speed and position on the runway, an overrun was inevitable; Boeing calculated that it would have taken 2,450 ft to stop the airplane from the maximum speed on a dry, paved runway using maximum braking and reverse thrust. The FDR data showed the left and right thrust reversers momentarily unlocked and the spoilers deployed (consistent with the captain’s deployment of the thrust reversers) about 0959:59 but then the thrust reversers relocked. The airplane performance study determined that the airplane’s speed was about 121 kts when it exited the paved surface at 1000:01; the FDR data became unreliable at 1000:03. Actual thrust reverser positions during the accident sequence could not be determined from the FDR data (the thrust reverser position parameters were invalid for both the accident flight and previous flights). Although a witness stated that he saw the thrust reversers deploy before he lost sight of the airplane, the lack of damage on and debris inside the thrust reversers’ lower doors was consistent with them having been fully stowed by the time that the airplane began striking tree branches and other vegetation. In the absence of any mechanical anomaly or an intentional command by a crewmember to stow the thrust reversers, it is possible that a crewmember may have inadvertently pushed the thrust reverser levers down during the accident sequence. Emergency Evacuation According to the captain, once the airplane came to a stop, he saw flames out the left cockpit window and commanded for everyone to evacuate. The captain stated that his main concern was getting the passengers off the airplane and away from the fire. He inadvertently did not shut down the engines (per the emergency evacuation procedure), and the right engine continued to run throughout the evacuation. The investigation determined that damage to the fuel system sustained during the final seconds of the impact sequence resulted in the left engine’s power loss. The passengers and crew successfully evacuated the airplane despite the running engine and other challenges, including smoke and flames outside the airplane that deterred them from using some exits. The emergency response was timely and effective. Although the airplane’s cabin included two passenger seat positions not identified on the supplemental type certificate (STC) for the airplane’s cabin modification, the additional seats did not hinder the emergency evacuation. Previously Issued Safety Recommendations As a result of the NTSB’s investigation of the Ypsilanti accident, in 2019, the NTSB issued safety recommendations intended to prevent future occurrences. These included Safety Recommendation A-19-1, which recommended that Boeing modify DC-9/MD-80 series airplanes to prevent the possibility of elevator jamming due to exposure to high-wind conditions while parked or taxiing. However, that same year, Boeing responded that, due to airplane structural limitations, neither a physical travel stop on the elevator structure (to prevent a jammed condition) nor a sensor (to provide a cockpit indication of a jammed condition) was feasible. Based on Boeing’s response, the NTSB classified Safety Recommendation A-19-1 Closed—Reconsidered. Thus, the accident airplane was not equipped with any design feature that could prevent the possibility of elevator jamming or provide the flight crew with a cockpit indication that the elevators were jammed. The NTSB also issued Safety Recommendation A-19-3, which recommended that Boeing lower the ground gust criterion for requiring physical inspections and operational checks of the elevators of DC-9/MD-80 series airplanes by maintenance personnel. In response, in 2019, Boeing established a 55-kt ground wind exposure criterion (which lowered the previous inspection criterion of about 65 kts) that would require a maintenance inspection to ensure that the elevators were not jammed. Although the lowered wind exposure criterion was in effect before this accident, the reported wind at the airport never met nor exceeded the 55-kt criterion during the time that the airplane was parked since its last flight. Thus, no maintenance inspection of the elevators was required for the accident airplane before the accident flight. However, had such high wind conditions existed, neither the maintenance manager for 987 Investments nor the Everts personnel who provided him with maintenance information was aware that an inspection would have been required. In addition, NTSB Safety Recommendation A-19-5 (issued as a result of the Ypsilanti accident investigation) asked the FAA to ensure that operators of DC-9/MD-80 series airplanes have procedures that define who is responsible for monitoring the wind that affects parked airplanes and for notifying maintenance personnel when conditions could meet or exceed the specified ground gust criterion. At the time of the accident, the FAA had not yet completed its planned actions in response to Safety Recommendation A-19-5. (However, as stated above, the wind at the airport where the accident airplane was parked never met nor exceeded the ground gust criterion that would have required a maintenance inspection of the elevators.) On June 14, 2022, the FAA issued Safety Alert for Operators 22001, “Recommended Procedures for Operators of Boeing DC-9/MD-80 Series and B717 Model Airplanes When Wind/Ground Gusts Meet or Exceed Criteria Specified in the Applicable Aircraft Maintenance Manual.” According to the FAA (at the time of this report), it was developing a notice that will be included in FAA Order 8900.1, Flight Standards Information Management System, to ensure that operators have procedures within their Continuous Airworthiness Maintenance Program defining who is responsible for monitoring the wind that affects parked airplanes and who is responsible for notifying maintenance personnel when conditions could meet or exceed the ground gust criteria specified in the Aircraft Maintenance Manual. The FAA stated that it anticipated releasing the notice by December 31, 2023. At the time of this report, Safety Recommendation A-19-5 was classified Open—Acceptable Response. In addition, NTSB Safety Recommendation A-19-4 asked the FAA to determine whether the gust load limits specified in 14 CFR 25.415 adequately ensure that critical flight control systems are protected from hazards introduced by ground gusts that contain dynamic, vertical wind components. The FAA responded in June 2022 that the requirements of 14 CFR 25.415 were revised on December 11, 2014, to include consideration of dynamic loads, and the FAA determined that the ground gust limit loads specified for transport-category airplanes were adequate and appropriate for current and foreseeable future designs. The FAA also noted that the requirement for considering dynamic loads did not exist when the DC-9/MD-80 series airplanes were certified. Based on the FAA’s review of 14 CFR 25.415 and other related material and its determination that the current regulations were adequate, Safety Recommendation A19-4 was classified Closed—Acceptable Action.

Quoted verbatim from the NTSB record.