Figure: Flight 191, from Chicago Tribune.

Eddie Sez:

The aircraft's left engine tore away during takeoff, taking enough of the wing structure to cause the left side leading edge slats to retract and cause an asymmetric stall. The pilots flew the air carrier's mandated procedure to pitch up to V2 and that caused the airplane to depart controlled flight. But none of that is what really caused the mishap. This is a mishap caused by bureaucracy at several levels:

  • The aircraft designers and FAA certification review failed to consider the vulnerability of the pylon structure to damage during maintenance.

  • American Airlines maintenance and engineering personnel evaluated and implemented a procedure for removing and replacing the wing engines and pylons as a single unit, rather than the two step procedure (engine and pylon separately), despite the manufacturer's warning of unacceptable risk. While one airline was able to accomplish this streamlined procedure without damage to components using a hoist, two other airlines (including American) used a forklift which resulted in damage to a parts of the pylon assembly.

  • American Airlines maintenance personnel did not inform engineering or quality control agencies about the difficulties of removing and replacing the engine and pylon combination using a forklift.

  • The pilots flying "by the book" provided them could not have recovered. They were in complete control when the airplane had excess speed, but their procedures required them to pitch up to capture V2. As the speed fell below V2+6, the left wing began to stall and the aircraft started to roll to the left. Simulator tests showed that at V2 + 10 the crash may not have occurred.

So what can we, as operators, learn from this?

  • We do not always have the expertise and resources to reinvent manufacturer procedures. McDonnell-Douglas was sure to have looked at the engine and pylon parts following their recommended procedures. The American Airlines maintenance and engineering teams were unlikely to have applied as much analysis into their revised procedures. We should be very reluctant to reinvent the wheel when it comes to these procedures.
  • Just because your company hands down procedures to you doesn't mean they necessarily have thought of everything. If you are the person turning the wrench you are duty bound to speak up when something doesn't seem right.

  • We pilots accumulate flying lessons from a broad range of sources, including other aircraft types. At the time of this crash the United States Air Force was flying the DC-10 using engine failure during takeoff procedures more commonly found in Boeing aircraft. Specifically, the USAF procedure at the time was to keep the speed at the time of engine failure up to V2 + 10. Had this crew been using these procedures, they probably would have survived. But that isn't to say these pilots were in anyway to blame; they were following American Airlines procedures.

What follows are quotes from the relevant regulatory documents, listed below, as well as my comments in blue.


Accident Report


Narrative

[NTSB Report, ¶1.11] Correlation of the DFDR [Digital Flight Data Recorder] and CVR [Cockpit Voice Recorder] recordings disclosed that the flightcrew had set the flaps and stabilizer trim at 10° and about 5° aircraft nose up, respectively, for takeoff. A rolling takeoff was made, takeoff thrust was stabilized at [?] KIAS, and left rudder and right aileron were used to compensate for the right crosswind. The V1 and VR callouts were made about 2 sec after these speeds were recorded by the DFDR. The elevator began to deflect up at VR. The aircraft began to rotate upward immediately and continued upward at a rate of 1.5° per sec. Flight 191 accelerated through V2 speed during rotation and before it lifted off the runway. The last stable takeoff thrust on the No. 1 engine was recorded 2 sec before liftoff. One second later, the word "damn" was recorded on the CVR, and then the CVR ceased operating.

[NTSB Report, ¶1.11] Flight 191 became airborne about 6,000 ft from the start of the takeoff roll and remained airborne for 31 sec. It lifted off at V2 + 6 KIAS and at 10° pitch attitude. Two seconds after liftoff, the DFDR reading for the No. 1 engine's N1 was zero, the No. 2 engine's N2 speed was increasing through 101 percent, and the No. 3 engine's N1 was essentially at the takeoff setting.

[NTSB Report, ¶1.1] Witnesses saw white smoke or vapor coming from the vicinity of the No. 1, engine pylon. During rotation the entire No. 1 engine and pylon separated from the aircraft, went over the top of the wing, and fell to the runway.

[NTSB Report, ¶1.1] Flight 191 lifted off about 6,000 ft down runway 32R, climbed out in a wings-level attitude, and reached an altitude of about 300 ft above the ground (a.g.1.) with its wings still level. Shortly thereafter, the aircraft began to turn and roll to the left, the nose pitched down, and the aircraft began to descend. As it descended, it continued to roll left until the wings were past the vertical position.

[NTSB Report, ¶1.1] Flight 191 crashed in an open field and trailer park about 4,600 ft northwest of the departure end of runway 32R. The aircraft was demolished during the impact, explosion, and ground fire.

Figure: Engine and pylon assembly, from NTSB Report, figure 2.

[NTSB Report, ¶1.12]

  • The first marks made by engine contact of the No. 1 engine and pylon with the runway began about 19 ft. to the right of the centerline lights and about 6,953 ft beyond the southeast end of runway 32R. Other parts of engine and pylon structure were located in this area; however, no spoiler actuators or hydraulic lines were found.

  • The pylon is attached to the wing using spherical ball joints in three different structural elements. Two of the spherical joints are aligned vertically in a forward bulkhead which is attached to structure in the wing forward of the front spar. Another spherical joint behind the forward bulkhead transmits thrust loads from pylon structure into a thrust link which in turn is connected through another spherical joint to structure on the lower surface of the wing. The third attachment point is a spherical joint in the pylon aft bulkhead which attaches to a clevis mounted on the underside of the wing. The pylon forward bulkhead and portions of the flange from the pylon aft bulkhead either remained with the separated No. 1 pylon or were scattered along the runway. (See figures 2 and 3.) The No. 1 pylon's aft clevis attach assembly and portions of the pylon aft bulkhead, wing thrust angle assembly and thrust link, and pylon forward bulkhead attach assembly remained with the wing.

  • The pylon forward bulkhead was bent forward about 30° and most of the bolts which held the bulkhead upper plates were missing. The upper 12 inches of the forward plate were bent forward an additional 10° to 15°. The aft plate Was broken below the thrust fitting connection, and a Large piece of the upper left corner was missing.

  • The wing's forward support fitting, which attached the pylon forward bulkhead to the wing at the upper and lower plugs and spherical bearings, was found at the main wreckage site. The upper and lower plugs and their attaching hardware were intact, and the upper and lower spherical bearings were attached to the fitting.

  • The pylon thrust fitting remained attached to the forward portion of the pylon's aft upper spar web. The pylon thrust link, which attached the pylon thrust fitting to the wing thrust angles, was found at the main wreckage site attached to a portion of the wing thrust angles. Its forward spherical bearing was cocked to the extreme left, and a segment of the bearing which had broken away was found on the runway.

  • The thrust bushing bolt had broken in two parts, both of which were found in the grass adjacent to the runway. The bolt nut was attached to one of the broken pieces, and the faces of the nut were gouged severely. Except for one lubrication retainer washer, which was not found, the remaining portions of the thrust bushing bolt assembly were found along the runway. One shim spacer from the assembly was crushed severely while the other was relatively undamaged.

  • The upper two-thirds of the pylon aft bulkhead separated from the flanges around its periphery and was found in the wreckage. The top two pieces of its attach lugs had separated from the bulkhead, and the left side of the bulkhead was gouged heavily near the lower edge of the wing clevis lug, which attached the aft bulkhead to the wing. The wing clevis was attached to the wing. The aft bulkhead's spherical bearing was attached to the clevis, and the separated pieces of the aft bulkhead's attach lugs were found on top of the spherical bearing.

  • A 3-ft section of the left wing's leading edge, just forward of the point where the forward part of the pylon joined the wing, was torn away when the engine pylon assembly separated from the aircraft. The No. 1 and No. 3 hydraulic system's extension and retraction lines and the follow up cables for the left wing's outboard slat drive actuators were severed. Thirty-five of the 36 leading edge slat tracks were examined. The examination disclosed that at impact the left wing's outboard slats were retracted, while the left wing's inboard slats and the right wing's inboard and outboard slats were extended to the takeoff position.

Figure: Pylon assembly, from NTSB Report, figure 3.

[NTSB Report, ¶1.16.2]

  • N110AA's pylon aft bulkhead was examined at the Safety Board's metallurgical laboratory. The examination disclosed a fracture of the upper forward flange. The larger part of this fracture was just forward of the radius between the flange and forward bulkhead plane and was about 10 inches long in the inboard-outboard direction. The fracture characteristics were typical of an overload separation.

  • Fatigue cracking was evident at both ends of the fracture.

  • The examination also disclosed that three shims were installed on the upper surface of the forward upper flange.

  • Taking into account the stack up of the forward flange created by the spar web, doubler, and fasteners, the clearance between the bottom of the clevis and the top of the web fasteners could be about .005 to 045 inch. The addition of a shim would narrow the clearance, and taking into account all tolerances in the spherical assembly, there could be an interference.

[NTSB Report, ¶1.16.3]

  • The evidence indicated that the maximum interference that would result from the insertion of the .050-inch-thick shim was .024 inch. The static tests conducted by American Airlines and McDonnell-Douglas showed that a crack would begin at a deflection of about 0.1 inch; thus, in the worst case, an additional deflection would be required to crack the flange.

  • During ground operation of the aircraft--taxiing, landing, and takeoff rolls--the aft bulkhead is subjected to compression loads and the aft end of the pylon is forced upward. During rotation, the loading changes and the aft bulkhead is subjected to tension-type loads. Those loads were found to be significantly lower than the fail-safe design loads.

[NTSB Report, ¶1.16.4] Wind Tunnel and Simulator Tests

[NTSB Report, ¶1.17.1] Air Carrier Maintenance Procedures

[NTSB Report, ¶1.17.3] The basic regulations under which the slats were certified did not require accountability for multiple failures.

[NTSB Report, ¶1.17.6]

[NTSB Report, ¶1.17.8] Flight Crew Procedures


Analysis

[NTSB Report, ¶2]


Probable Cause

[NTSB Report, ¶3.2]


See Also:

Abnormal Procedures & Techniques / Maintenance Malpractice


References

NTSB Aircraft Accident Report, AAR-79-17, American Airlines, Inc., McDonnell-Douglas, DC-10-10, N110AA, Chicago, Illinois, May 25, 1979