Aircraft Accidents and Lessons Unlearned XIV: American 191
The relationship between Maintenance and Engineering (M&E) can be represented as two ‘friends’, facing ninety degrees apart with arms crossed over their chests, looking suspiciously at the other over their shoulders. Unlike the pilot-mechanic relationship being akin to the New York Yankees and the Boston Red Sox competition, M&E folks are not rivals for the ‘True Airman’ prize. M&E is a semi-symbiotic bond; they are more dependent on the support of each other than either will admit. At the same time, theirs is an affiliation whose foundation demands trust and communication. In this strange allies’ relationship, the tragedy of American Flight 191 was fostered.
American 191 crashed after taking off from Chicago’s O’Hare airport on May 25, 1979. National Transportation Safety Board (NTSB) Report AAR 79/17 recorded the investigation into the accident. The root cause of the accident was that American Airlines (AA) combined two approved maintenance procedures in an impractical way: removal and replacement (R&R) of an engine and the R&R of a pylon.
Briefly, McDonnel-Douglas (MCD) issued service bulletin (SB) 54-48 in 1975 and SB 54-59 in 1978 for their DC-10s; both SB’s called for replacing the two wing engine (#1 and #3) pylons’ spherical bearings to address cracks found in the aft Attach Spherical Bearings and lubrication problems in the forward Attach Spherical Bearings, which would contribute to seizing.
Both SBs required the engine be removed to access the pylon, a costly undertaking in both manpower and money. Replacing the engine with the pylon attached as one unit would lower maintenance costs as AA complied with the SBs for their DC-10 fleet; the timesaving procedure meant less aircraft down time and less man-hours used. In order to modify the repair procedures, AA would need acceptable data from MCD for the removal and reinstallation of the combined engine/pylon, e.g. center-of-gravity (CG); the combined weights (13,480 pounds) and the dimensions from the lifting device to the end of the pylon.
MCD strongly advised against the procedure; however, MCD could not refuse the requested data nor prevent AA from performing the modified procedures. The two main problems with modifying the maintenance procedures were: 1 – the engine/pylon’s CG is moved well outside the envelope that normal hoisting is accomplished in, and, 2 – the engine/pylon combination would be raised from beneath, not lifted (hoisted) from above.
Each of the DC-10’s three CF6-6D engines’ (11,600 pounds) main purpose: to provide 40,000 pounds of thrust to move the aircraft. The energies each engine produces include shear, torsional, compression and tensile forces. The pylon (1860 pounds) is what the CF6-6D engine mounts to; the pylon translates all loads – the previously mentioned four forces – from the engine into the aircraft, plus the energies e.g. weight, landing stresses, balance, thrust and reverse thrust, which are produced during normal operation. The pylon’s design transfers these forces through its entire structure, which is why the pylon’s integrity must not be compromised.
A CF6-6D engine is normally lowered and raised by four come-along hoists that are hung off the pylon. There are four points, left and right front, left and right rear. With each come-along manned, a fifth person directs the other four to slowly raise and lower as the engine mates to its mounts, otherwise the 11,600 pound engine’s mass can cause considerable structural damage, e.g. cracks, scoring, blunt damage, bending or transmittal load damage. The tight clearances between pylon and engine demand that someone direct the four points closely.
By itself, the 1860-pound pylon is similarly raised, a director controlling the ascent. The CG of the engine or the pylon is managed within a small envelope front to back, the CG movement being limited by the forward and aft come-along hoists. The AA procedure that led to the accident, required a CG movement that exceeded the normal limits. Furthermore, per accident report AAR 79/17, the engine/pylon combination was raised and lowered from underneath by forklift – not from above.
The danger of this approach was from the lack of control and the resulting sustained damage. The tight clearances between pylon and wing allowed no room for error; damage was easily suffered if the engine/pylon was not raised slowly and under complete control. Any forklift hydraulic pressure loss would result in the pylon acting like a twenty-foot long pry bar in its mounts. The twenty-foot separation between forklift and pylon meant that the forklift’s movements became exaggerated, e.g. a one-inch side movement of the forklift’s back wheels translated into the pylon moving many more inches at the wing. Furthermore, the forklift driver was blind to what was happening at the pylon’s end; the driver was solely dependent on the Director’s instructions telling the forklift driver, e.g. to move forward, backward, left, right, slide the forks left, slide right, lean the forks back or forward. Meanwhile the engine/pylon’s CG was well forward of the forklift; the forklift bounced with each movement, making the hoist unsteady while repositioning. The localized forces on the pylon were not translated through its structure; damage resulted in specific local areas, weakening the integrity of the pylon.
The modified procedures should never have gotten this far; the exaggerated movements of the forklift were the equivalent of swinging an 13,480-pound mallet or forcing an 13,480-pound, twenty-foot pry bar to the pylon’s structure. Any damage incurred was not recognizable. What would one look for? How would they identify unexpected damage in a localized area not known for that type of stress cracking? Non-destructive techniques to locate damage were either not employed or were not available to identify the injuries.
This is where communication between Engineering and Maintenance most likely broke down. While Engineering wrote the procedures, Maintenance was on the scene; there should have been reported problems unforeseen during the planning stages. Was Engineering part of the ongoing procedural changes; did Maintenance include them? Was Engineering available on-site during the entire job? Would the Maintenance group have stopped at the first sign of trouble or pushed on with a ‘can-do’ attitude employing unsafe compensations? Did Engineering plan for proper mount bolt torques while taking into account the forklift’s possible hydraulic creep? In other words, if the forks drooped, putting weight on the bolt heads, would the torque be correct when the mount bolts were tightened down?
Human Factor issues included the desire to get the job done in good time. Was there pressure to get the engine/pylon changed out and pushed from the Hangar quickly? This issue is affected by communication or the lack thereof, or simple reasons like inconvenient breaks, shift changes or poor shift turnovers. Were the proper number of personnel assigned to the job? Was any other work being performed around the wing that intruded or interfered with the engine/pylon change?
These are issues that usually never came up as part of this or other NTSB accident investigations. Until 2001, the NTSB did not have an aircraft maintenance experienced accident investigator. This investigator was an engineer, unfamiliar with the maintenance environment. In order to understand a culture, one must have either worked in it or taken the time to become familiar with it; AA’s maintenance culture was never examined to determine where the failure of communication occurred.
To correct the mistakes made in what was done incorrectly, the tangible details must be addressed, e.g. the modified maintenance procedures and the changes made in lifting. However, to properly prevent an accident from recurring, the investigators must determine how and when the errors got out of hand. Factors, e.g. time-constraints and their effects on the work force; M&E assuring two-way clear communications; guaranteeing Quality Control was on hand through the altered procedures to assure safety is maintained; decisions were made by consensus of those involved, not by a senior person or from off-site. Unless the contributing factors and the root causes to the accident were properly identified and addressed, accidents will recur under similar circumstances, especially as Flight 191 is forgotten over time.