TWA Flight 841 (1979)

TWA Flight 841

N840TW, the aircraft involved, four years before the accident
Accident
Date	April 4, 1979
Summary	Uncontrolled descent; cause disputed
Site	Over Saginaw, Michigan, United States
Aircraft
Aircraft type	Boeing 727-31
Operator	Trans World Airlines
IATA flight No.	TW841
ICAO flight No.	TWA841
Call sign	TWA 841
Registration	N840TW
Flight origin	John F. Kennedy International Airport, New York City, U.S.
Destination	Minneapolis-Saint Paul International Airport, Minneapolis, U.S.
Occupants	89
Passengers	82
Crew	7
Fatalities	0
Injuries	8
Survivors	89

TWA Flight 841 was a scheduled passenger flight from John F. Kennedy International Airport, New York City, en route to Minneapolis-Saint Paul International Airport in Minneapolis, Minnesota. On the evening of April 4, 1979 while flying over Saginaw, Michigan, the Boeing 727-31 airliner began a sharp, uncommanded roll to the right, and subsequently went into a spiral dive. The pilots were able to regain control of the aircraft and made an emergency landing at Detroit Metropolitan Airport.

The aircraft involved was a 13 year old Boeing 727-31, registered N840TW, which had accumulated about 35,421 hours. The captain was Captain Harvey G. "Hoot" Gibson, 44, who had a total of 15,710 piloting hours, 2,597 of them on the 727. The day before he had returned to flying following a three month medical leave.^[1]: 4 The First Officer was J. Scott Kennedy, 40, who had a total of 10,336 total piloting hours, 8,348 of them on the 727. The Flight Engineer was Gary N. Banks, 37, who had a total of 4,186 piloting hours, 1,186 of them on the 727.^[1]: 38

At 8:25 p.m. EST, TWA Flight 841 departed JFK International after a 45-minute delay due to traffic congestion and reached its initial cruising altitude of 35,000 feet at 8:54 p.m. Due to a 100-knot headwind, the pilots requested to climb to 39,000 feet at 9:25 p.m., which was granted. At 9:47 p.m., after traveling nearly 540 miles (870 km),^[2] while cruising at 39,000 feet (12,000 m) near the city of Saginaw, Michigan, Captain Gibson^[3] said the aircraft was operating with the autopilot on "Altitude Hold" mode when he felt a high frequency vibration in the balls of his feet, followed by a buzzing sound and a light buffeting of the airframe. He looked at his instruments and saw the plane banking right before yawing sharply to the right, briefly pausing, and yawing right again into a skidding right roll with the autopilot moving the control wheel to the left to level the wings.^{[4]: 194-195} Gibson disconnected the autopilot and applied full left aileron which had no effect. He got on the rudder pedals and applied full left rudder. As he did so something didn't feel right but he couldn't place it.^[5]: 11 Although Captain Gibson's inputs briefly stopped the roll from increasing, the roll continued and the aircraft spiraled out of control. Gibson shouted something to the effect of "We're going over!"^[1]: 6 After completing one 360° roll, Gibson shouted to the co-pilot "Get em' up!" referring him to deploy the speed brakes, but the co-pilot didn't understand what the Captain meant so Gibson deployed them which also had no effect.^[1]: 2–3

Flight 841 dove about 34,000 feet (10,000 m) in just 63 seconds.^[1]: 2 During the course of the dive, the plane rolled through 360 degrees at least twice and exceeded the Mach limit for the 727 airframe.

At about 15,000 feet (4,600 m) the pilots extended the landing gear in an attempt to slow the aircraft. A few seconds after extending the gear, Gibson managed to regain control and pulled the 727 out of its dive at about 5,000 feet (1,500 m).^[6][7] The plane suffered substantial structural damage with the No.7 leading-edge slat and a flight spoiler having detached, the right outboard aileron hinge fitting broke due to metal fatigue resulting in free-play, and the System A hydraulics ruptured due to the right main gear overextending which also broke the sidebrace and actuator support beam; the Flight Engineer reported they had a fail flag for the lower rudder yaw damper^[1]: 10 but made an emergency landing at Detroit Metropolitan Airport in Michigan at 10:31 p.m. EST without further trouble. After landing and shutting the engines down the pilots turned on the Auxiliary Power Unit (APU) as a mechanic plugged into an intercom to talk to the pilots and reported fuel was leaking from the left side. All 89 people deplaned via the aft airstair with only eight passengers suffering minor injuries.^[1]

Several days after the upset, the aircraft was sufficiently repaired to be ferried to Kansas City for further repairs instead of being preserved for tests and in-depth analysis of components and systems in order to conclusively determine what happened; it returned to service in May 1979.^[1]: 4 The flight crew would be questioned on three occasions during the course of the entire investigation: hours after landing in Detroit, on April 12, 1979, and on January 29, 1980.^[1]: 37 The investigators of the National Transportation Safety Board (NTSB) located the No. 7 slat missing from the leading edge of the right wing amongst other parts that had detached. They requested that Boeing, the manufacturer, inspect the remainder of the slat assembly, including a portion of the slat actuator cylinder (the motor that moves the slat and holds it in position). Boeing determined that the No. 7 slat had failed because the slats had been extended while the aircraft was flying at cruising speed. They also determined that it was "impossible" for the slat to extend without manipulating the controls.^{[8][better source needed]}

The TWA Flight 841 investigation would be at the time the lengthiest accident investigation in the NTSB's history.^[9] After eliminating all known sources of mechanical failure, the NTSB concluded in June 1981 in its final report, that the probable cause of the accident was the isolated extension of the No. 7 leading-edge slat from the flight crew manipulating the flap/slat controls in an inappropriate manner leading to an uncommanded roll to the right and the captain's untimely action.^{[10][1]: 1–2}

The Safety Board determines that the probable cause of this accident was the isolation of the No. 7 leading-edge slat in the fully or partially extended position after an extension of the Nos. 2, 3, 6 and 7 leading-edge slats and the subsequent retraction of the Nos. 2, 3, and 6 slats, and the captain's untimely flight control inputs to counter the roll resulting from the slat asymmetry. Contributing to the cause was a preexisting misalignment of the No. 7 slat which, when combined with the cruise condition airloads, precluded retraction of that slat. After eliminating all probable individual or combined mechanical failures, or malfunctions which could lead to slat extension, the Safety Board determined that the extension of the slats was the result of the flightcrew's manipulation of the flap/slat controls. Contributing to the captain's untimely use of the flight controls was distraction due probably to his efforts to rectify the source of the control problem.^[1]: 1

Investigators believed that 727 pilots (in general, and this crew specifically) were enabling the trailing edge flaps to extend independently of the slats via the flap/slat lever by turning the Alternate flap switch on, pulling the leading-edge slats circuit breaker, turning the Alternate flap switch off and then setting the flaps to two degrees during high altitude cruise. This configuration was rumored to result in increased lift with no increase in drag, thus allowing more speed, higher elevation, or decreased fuel consumption. Flight tests showed that airspeed and performance decreased in this configuration. Flaps and slats are intended to only be deployed at low speeds during take-off and landing.^[1]

The crew, Capt. Harvey G. "Hoot" Gibson (1934–2015),^[11] First Officer J. Scott Kennedy (1939–2017),^[12] and Flight Engineer Gary N. Banks (born 1942), denied that their actions had been the cause of the flaps' extension:

At no time prior to the incident did I take any action within the cockpit either intentionally or inadvertently, that would have caused the extension of the leading edge slats or trailing edge flaps. Nor did I observe any other crew member take any action within the cockpit, either intentional or inadvertent, which would have caused the extension.

— Capt. Gibson^[13][14]

The flight crew testified that they had not engaged the flaps and instead suggested that the actuator to the No. 7 slat had failed, causing its inadvertent deployment. The NTSB concluded that "if the flight crew's recollections are accurate," the slat extension must have been caused by a mechanical failure or defect but ultimately rejected this as improbable and attributed the extension of the flaps to the deliberate actions of the crew.^[1]: 24 The crew maintained that such failures had happened on other 727s prior and subsequent to this incident.^[15] The NTSB report noted seven separate cases involving a single leading edge slat extension and separation were reported between 1970 and 1973, but none of these reports indicated whether or not the slat extension was due to flight crew involvement.^[1]: 18 Records after 1974 did include two report slat extensions between 1974 and towards the end of the NTSB's investigation in 1981, one of which was inadvertently caused by the flight crew. However, in none of these previous cases did the flights experience significant problems.^[1]: 24

Gibson and the Air Line Pilots Association (ALPA) appealed against the NTSB's findings, first to the NTSB itself in 1983, 1990, 1991, 1992, and 1995 and then to the U.S. Ninth Circuit Court of Appeals.^[9][16] Both petitions were rejected: the former for lack of new evidence, and the latter for lack of jurisdiction due to the NTSB's "unreviewable discretion".^{[citation needed]}

The aircraft was equipped with a Fairchild Industries Model A-100 cockpit voice recorder (CVR). However, 21 minutes of the 30-minute tape were blank. Tests of the CVR in the aircraft revealed no discrepancies in the CVR's electrical and recording systems,^[1]: 6 however, according to ALPA this was false. According to them, a TWA representative erroneously told the NTSB that tests had been conducted when none were believed to have been done.^[17]: 3 The CVR tape can be erased by means of the bulk-erase feature on the CVR control panel in the cockpit. This feature can be activated only after the aircraft is on the ground with its parking brake engaged.^[1]: 6 In a deposition taken by the Safety Board, the captain stated that he usually activates the bulk-erase feature on the CVR at the conclusion of each flight to preclude inappropriate use of recorded conversations. However, in this instance, he and the other two pilots did not recall having done so.^[1]: 6 Pilot and author Emilio Corsetti III suggested in his book Scapegoat: A Flight Crew's Journey From Heroes to Villains to Redemption, that the wiring to the CVR could have been damaged by the heavy, 6 G-force pull out.^[5]: 380 The NTSB made the following statement in the accident report:^[1]: 33

We believe the captain's erasure of the CVR is a factor we cannot ignore and cannot sanction. Although we recognize that habits can cause actions not desired or intended by the actor, we have difficulty accepting the fact that the captain's putative habit of routinely erasing the CVR after each flight was not restrainable after a flight in which disaster was only narrowly averted. Our skepticism persists even though the CVR would not have contained any contemporaneous information about the events that immediately preceded the loss of control because we believe it probable that the 25 minutes or more of recording which preceded the landing at Detroit could have provided clues about causal factors and might have served to refresh the flightcrew's memories about the whole matter.

This section may be too technical for most readers to understand. Please help improve it to make it understandable to non-experts, without removing the technical details. (November 2024) (Learn how and when to remove this message)

In ALPA's 1990 petition to the NTSB to reconsider their findings, an analysis conducted by ALPA concluded that the No. 7 slat did not cause the upset and instead extended as a result of the upset. Had the slat caused the upset then it would have detached from the wing at an altitude of around 31,500 feet, not at 8,000 feet.^[17]: 29 ALPA concluded that evidence best supported that while cruising at 39,000 feet, the bolt to the outboard right aileron on TWA flight 841 fractured, causing the aileron to flutter and create the high frequency vibration that Captain Gibson reported. As the aileron floated up, the plane banked to the right and turned off its heading, the autopilot tried to correct for this by moving the control wheel left. Once the control wheel turned more than 10°, the spoilers on the left wing deployed to aid in roll control, creating the buzzing sound. With the plane turning right and the autopilot commanding a left turn, the 727 was in a cross-controlled position. The yaw damper rate gyro and or coupler sensed discrepant rudder inputs which resulted in the lower rudder going into the hardover position, causing the plane to yaw severely right. In this condition the left wing produced more lift as a result. A large sideslip angle on sweptback planes like the 727 produces a large rolling moment.

Although Gibson disconnected the autopilot and applied opposite aileron and upper rudder, with the lower rudder in the hardover position and limited roll control due to the right outboard aileron free-floating, his control inputs were insufficient to prevent TWA 841 from going into an uncontrollable spiral dive. When the crew lowered the landing gear, the over extension of the right main landing gear ruptured System A hydraulics which centered the lower rudder and allowed the pilots to recover from the dive. Analysis of the No. 7 slat by the NTSB showed a lack of wear, it was misaligned, and didn't lock into its locking mechanism, meaning that it was held in place only by hydraulic pressure and aerodynamic forces. With the loss of hydraulics, the slat extended at 8,000 feet and quickly ripped off. The other slats stayed retracted because they locked into their locking mechanisms.^{[17][verification needed]}

In his book Emergency: Crisis In the Cockpit,^[4] Stanley Stewart, a pilot, asserted the following about the NTSB's findings:

• The NTSB conducted 118 tests in a flight simulator with the isolated extension of the No. 7 slat and the aircraft was easily controllable in that condition. According to Steward, there was no correlation between the simulator traces of these maneuvers and Flight 841's FDR traces.^{[1]: 9 [4]: 209}
• Although it was not determined when the outboard aileron bolt fractured, if it did fracture before the upset, it would have permitted free-play which would induce a right roll thereby contributing to the upset.^[4]: 212 The simulator tests did not take into consideration free-play of the right outboard aileron.^[1]: 9
• The extension of the slats would cause more than a moderate buffeting of the airframe while Captain Gibson reported only a light buffeting.^{[4]: 214  [1]: 2, 13}
• The crew would not have been able to erase the CVR as the aircraft had to be completely shut down and on the ground. According to Stewart the damage would have meant that the computers did not recognize the aircraft in a fully landed state.^[4]: 217
• A CVR technical expert interviewed for Popular Mechanics contended that the slow transfer of electrical power from engine power to the APU after it was turned on likely caused the erasure of the CVR.^[4]: 217
• The pilots, along with the flight attendants and 14 passengers including four or five first class passengers, stated that Second Officer Banks had not left the cockpit just before the upset. Instead, Captain Gibson maintained that Banks had only briefly left the cockpit 30 minutes prior to the upset.^[4]: 218
• Stewart found it rather absurd that Second Officer Banks, supposedly returning to the cockpit right before the upset, would notice the pulled circuit breaker behind crew coats in a dark cockpit and pushed it back in without saying anything to the other pilots. And if the practice of extending the flaps to the 2° was widespread then he would not have pushed the breaker back in. Stewart stated: "The NTSB scenario did not fit the facts and, even if the crew had deliberately set 2° of flap as entertained here, the extension of the number seven leading edge slat was highly unlikely to have happened in this manner".^[4]: 218
• The NTSB investigators questioned passengers seated over the wings if they heard any unusual sounds before the upset, such as the high pitched shrill of the hydraulic flap motors as the pilots, supposedly, extended them and then later retracted to rectify the extension of the slats to which they said no. According to Stewart, this alone suggested that the flaps were not extended in cruise flight.^[4]: 221
• Stewart claimed that the entire investigation was conducted with the flight crew absent except for shortly after they landed and two public hearings. After these hearings, during the intervals between the two and the release of the Final Report, the flight crew were never consulted, cross examined, or questioned by any investigator, and all of the Captain's phone calls to the NTSB were refused.^[4]: 222
• Stewart contended that the flight crew knew the aircraft was potentially unstable at 39,000 ft. Stewart believed it would be unlikely that they would "fool around" with the controls and risk the stability of the aircraft.^[4]: 221
• He stated that there were other incidents of 727-200s with uncommanded slat extensions in the years prior to and after the accident.^[4]: 224

Emilio Corsetti III stated in his book, Scapegoat: A Flight Crew's Journey From Heroes to Villains to Redemption,^[5] that the Boeing engineers and NTSB investigators had massaged the data to make it match what they thought happened, not the flight crew's version of events;^[5]: 192 and stated that "had the NTSB not let the supposed erasure of the CVR implant a bias on the investigation, they might have considered possibilities other than crew involvement. They might have talked to the crew in greater detail and would have learned that the plane yawed before the upset".^[5]: 384

This accident was the subject of a CBS News Special titled, "The Plane That Fell From the Sky". The special won a Peabody Award in 1983.^[18]

The accident was featured on season 22 of the Canadian documentary series Mayday, in the episode titled "Terror over Michigan".^[19]

• China Airlines Flight 006 - Another case where a Boeing aircraft entered an uncontrolled nose-dive
• China Eastern Airlines Flight 583 — Another slat extension in-flight accident
• Partnair Flight 394 - Another instance where using the APU as the power source caused a stoppage in recording of the Cockpit Voice Recorder.
• United Airlines Flight 585
• USAir Flight 427
• Eastwind Flight 517

• Accident description at the Aviation Safety Network
• NTSB Investigation
• U.S. Ninth Circuit Court of Appeals, GIBSON v NTSB (appeal dismissed--lack of jurisdiction)