Designed for Disaster: The DC-10 Airliner, Part 2

November 30, 2009

The Cargo Door Fiasco - Turkish Air Flight 981

Fifteen days had passed since American Airlines Flight 96 had come within a hair’s breath of crashing; fifteen days that had seen the Western Region of the Federal Aviation Administration begin the process of issuing an Airworthiness Directive (AD) intended to correct the problems with the DC-10 cargo door design and probably the unvented passenger compartment floor as well.

But fate had intervened. The FAA administrator had taken the unusual action of becoming personally involved in the problem thereby short-circuiting the AD issuance. He had spoken with the president of Douglas Division of McDonnell-Douglas and the two men had agreed that the situation didn’t require a far reaching AD since the FAA and Douglas division of McDonnell-Douglas had a great working relationship.

With the involvement of the FAA’s top man, management of the Western Region field office stepped back and allowed McDonnell-Douglas to address the situation with service bulletins. The only problem was a service bulletin didn’t have the same effect as an AD.

Airworthiness Directives and Service Bulletins

An AD is a big hammer. A properly worded AD can require an action be taken before further flight under penalty of law. A service bulletin can only suggest an action be taken before further flight. More likely, it will suggest that the activity be accomplished within a certain time period or perhaps at the discretion of the aircraft operator.

Although an airline exposes itself to legal action if it doesn’t accomplish a service bulletin in a timely manner, (and there’s a crash as a result of a failure to perform a service bulletin), the FAA doesn’t view non-compliance with a service bulletin the same way it reacts to flouting an AD.

Fail to perform service bulletins and you’ll get jawboned. Flagrantly disregard an AD and you’ll be grounded.

Nevertheless, most airline companies take the issuance of a service bulletin, particularly a SB with ALERT status (meaning it’s a safety of flight issue), as an important event. Virtually all airlines jump right on them.

McDonnell-Douglas Service Bulletin 52-35 was sent to all operators of the DC-10 following the American Airlines Flight 96 episode. The SB was printed on Blue Paper and labeled ALERT which meant, in effect, “Listen Up, Pay Attention”.

SB 52-35 suggested the installation of a small viewing port directly above one the locking pins in the cargo door on all existing and all future DC-10 aircraft. It also suggested that decals be placed on the aircraft clearly showing the proper orientation of the locking pin and also showing what an improperly latched locking pin looked like.

The problem was the door was 15 feet above the ground and required a flashlight at night to allow viewing of the locking pin. Also, the window was only 1 inch in diameter and easily fogged. And in some cases, the baggage handler had to move a work ladder into position in order to complete the inspection.  The solution placed quite a burden on a person normally referred to within the industry as a “Ramp Rat” since it made that hapless individual responsible for the lives of hundreds of people.

A second SB, number 52-37, was also issued, but on white paper, which meant it did not have ALERT status. This SB added a stiffener to prevent bending of a torque tube on the cargo door. The action required installing a metal retainer to prevent a baggage handler from being able to force the door’s locking mechanism into a false latched position by not allowing the torque tube to deform when the door’s handle was operated.

dc-10-forced-catch-closure.jpg

Fig 1. The torque tube for the cargo door’s latching mechanism could be bent if the handle was forced into the closed position. A service bulletin was issued to correct the problem but no Airworthiness Directive followed the service bulletin. In addition, the SB was not issued as an ALERT category meaning it wasn’t necessarily a safety of flight issue.

The SB was issued because the torque tube on AA 96 had been bent by the handler when he forced the door shut. The theory was that if the torque tube could not be bent, the door could not be forced closed and give a false lock indication.

A third SB was issued that required the replacement of the undersized wiring used for the automated door closure motor.

Dan Applegate

Two weeks after the American Airlines incident, Dan Applegate, Director of Product Engineering, Convair Division of General Dynamics, had seen and heard enough.

He felt that McDonnell-Douglas had compromised the integrity of the DC-10’s design. He also felt they were setting his company up to take the fall if anything went wrong.

Applegate wrote a memo to his boss, J.B Hurt, Program Manager, DC-10 Support Program, and in the memo he called a spade a spade.

  • He said that the fundamental safety of the cargo door was being progressively degraded.
  • He said that the airplane demonstrated an inherent susceptibility to catastrophic failure when exposed to explosive decompression - i.e. the unvented passenger floor.
  • He recounted the door blowing off the pressure test airframe in July of 1970 and described how the explosion would have rendered control of the tail flight surfaces and number 2 engine impossible because the control cables were routed down the center of the unvented floor.
  • He wrote how the baggage loader in Detroit had forced the cargo door shut which caused the improperly latched door to blow off the airplane, very nearly resulting in its loss.
  • He called McDonnell-Douglas’ service bulletins “Bandaid fixes” and predicted the loss of other aircraft because of their shortcomings and because of the decision to not cut vents in the passenger compartment floor above the aft cargo door.
  • Applegate made a compelling argument that even if the door fixes were fool-proof (which he argued they were not), the decision to not address the floor problems left the aircraft vulnerable to explosive decompression and total loss of control in the event of a bomb, mid-air collision or other event unrelated to the cargo door itself.

Applegate’s memo was far blunter then most engineering memos. There were no intentionally broad statements like: “These data would appear to indicate….”, or “Testing has provided reason to believe….”

Instead, Applegate told his Convair bosses that the DC-10’s design was fraught with peril and they needed to discuss the matter with McDonnell-Douglas ASAP. He said that he expected the loss of airplanes during the 20+ year life of the product if design changes were not made.

At the end of the day though, for all its urgency and concern, Applegate’s memo did not result in any changes.  Convair Division of General Dynamics’ upper management decided it was not their responsibility to make McDonnell-Douglas aware of the DC-10’s shortcomings since the airframe manufacturer was surely aware of the problems already.

Turkish Air Flight 981

The DC-10 that would be used on Turkish Air Flight 981 on March 3rd, 1974 finished the build sequence three months after the three cargo door service bulletins were released following the American Airlines Flight 96 incident. Three more months went by before it was delivered to Turkish Airlines.

The construction logs showed that all three service bulletins had been performed however service bulletin 52-37, the one dedicated to preventing the torque tube from becoming bent, had in fact not been done.

dc-10-981-cargo-door.jpg

Fig 2. Although the construction logs for the airplane sold to Turkish Air indicated that the torque tube service bulletin had been performed, in fact it had not been done. Over time, the torque tube became progressively more and more bent which allowed the baggage handler for flight 981 to mistakenly believe that the door was locked when it wasn’t.

A year and a half after Turkish Air took delivery of the airplane, it would all come undone a couple of miles in the air above Senlis, France.

Because the torque tube stiffener had not been installed, the torque tube had become more and more deformed. The discrepancy had gone unnoticed by airline maintenance personnel and by a succession of flight engineers and ground engineers during their preflight inspections.

March 3rd, 1974

On March 3rd, 1974, the baggage loader servicing flight 981 at Orly airport swung the rear cargo door’s handle to the closed position. He later said he had not felt excessive resistance and he thought the door latched securely.

The individual was from Algeria. He could read neither French nor English so he didn’t understand the words on the decals next to the view port. He also had not been taught to look at the lock pin through the viewing port and didn’t know what it was for.

Flight 981 departed Orly airport for London Heathrow at 12:30 PM local time. It departed to the east and then turned north to avoid Paris. Just after passing the French village of Meaux, controllers heard a transmission from the aircraft. In the background they heard the Crew Advisory System (CAS) reporting problems with the pressurization system and an over-speed condition.  They also heard a crew member saying that the “Fuselage had burst”.

javdc10020030374dioramaie0.jpg

72 seconds later, the aircraft slammed into the Ermenonville forest. The destruction was so complete that some of the 346 victims were never identified.

The circumstances of the loss of Turkish Air flight 981 were virtually identical to the American Airlines flight 96 incident and the cargo door episode that occurred during certification testing.

In the testing event and the Turkish Air accident, an improperly secured rear cargo door had blown off the airframe causing a total collapse of the passenger floor. This had destroyed the control cables going to the rear of both the Turkish Air airplane and the test airframe.

In the American Airlines incident however, there had been a different outcome when the cargo door departed the aircraft because of the interior layout.

The American Airlines configuration had a galley area above the floor failure point but there were several rows of seats above the failure point on the Turkish Air airplane.

Apparently there was an incrementally greater static load on the floor on flight 981 then there had been on flight 96. The increased load had caused a total collapse of the floor and complete destruction of the flight controls on flight 981 whereas the American Airlines incident had caused only a partial destruction of the flight controls.

Aftermath

Dan Applegate had been proven right. A DC-10 had crashed because of the cargo door problems and the unvented passenger compartment floor. His memo came to light during the subsequent lawsuits after the Turkish Air 981 crash and his pre-crash actions are considered to be a study in proper engineering ethics by many.

For his part, except for required dealings with the lawyers, Applegate avoided public comments on the entire episode and kept his post-crash opinions to himself.

After the loss of flight 981, the FAA required a complete redesign and retrofitting of the cargo doors on all existing and future DC-10 aircraft.

Additionally, the FAA required vents be cut into cabin floors to allow the delta to quickly equalize in the event of a blown-out door.

Somewhat later, they further required that airliners be able to sustain the loss of a 20 square foot area of the hull and still remain controllable.

Any one of the three ADs would likely have saved the lives of the 346 people lost northeast of Orly airport that March day in 1974.

Next Monday, in part 3 of this series, we’ll look at the crash of American Airlines Flight 191.

Unsanctioned maintenance practices coupled with more design shortcomings cost the lives of an additional 273 people and resulted in the loss of the DC-10’s airworthiness certificate for a time. It also caused a near total loss of faith in the airplane by the flying public.

John Loughmiller is an Electrical Engineer, Commercial Pilot, Flight Instructor and a Lead Safety Team Representative for the FAA.

(All imagery in this article was obtained either from government accident reports or from Wikimedia Commons media library and is public domain.)

Comments (0)

Please log in or register to post comments.
By submitting this form, you accept the Mollom privacy policy.
  • Oldest First
  • Newest First
Loading Comments...