1.2 Historical Maintenance Accidents & Incident Case Studies

1.2 Historical Maintenance Accidents & Incident Case Studies

In aviation maintenance engineering, regulatory standards are written in the aftermath of forensic accident investigations. Studying landmark accidents provides licensed technicians with essential technical lessons, demonstrating how minor cognitive slips, adverse environmental conditions, and organizational deficiencies combine to breach multi-layered safety barriers. This section analyzes three defining accidents in maintenance history that reshaped international human factors training and continuing airworthiness oversight.

British Airways Flight 5390 (BAC 1-11, 10 June 1990)

On the morning of 10 June 1990, a British Airways BAC 1-11-528FL (registration G-BJRT) departed Birmingham International Airport for Malaga, Spain. As the aircraft climbed through 17,300 feet over Didcot, Oxfordshire, the left-hand (Captain's) windscreen blew out of its airframe mounting frame. The explosive decompression sucked Captain Tim Lancaster halfway out of the cockpit aperture. His torso remained pinned against the outer aircraft skin at sub-zero temperatures and high-velocity slipstream while cabin crew members held onto his legs. The First Officer performed an emergency descent and executed a safe emergency landing at Southampton Airport with the Captain miraculously surviving.

Forensic Technical Breakdown

The UK Air Accidents Investigation Branch (AAIB Formal Report 1/92) established the following technical facts:

• Externally fitted windscreen: A plug-type windscreen fitted from the inside is pressed into its frame by cabin pressure. The BAC 1-11 windscreen was fitted from the outside, so cabin pressure tended to push it out and the load was carried by its 90 countersunk attachment bolts.
• Fastener mismatch: The windscreen had been replaced by the Shift Maintenance Manager (SMM) during the night shift of 8/9 June 1990, and the accident flight was the first since the change. The specified bolt was the A211-8D (10-32 UNF thread). Of the 90 bolts fitted:
  • 84 bolts were A211-8C: same thread pitch, but about 0.026 inches (~0.66 mm) smaller in diameter than specified (an 8-32 rather than a 10-32 bolt).
  • 6 bolts were A211-7D: correct diameter, but 0.1 inches (~2.5 mm) too short.
• The old windscreen was also wrongly fitted: 78 of the 80 bolts recovered from the removed windscreen were A211-7D bolts, 0.1 inches shorter than specified. Replacing them like-for-like copied the earlier error.

Latent Conditions & Human Factors Precursors

1. Night Work: The job was done on a night shift, when alertness, visual search and memory are naturally at their lowest. The AAIB also recommended that staff who need prescription spectacles be required to wear them for maintenance tasks.
2. Failure to Consult Technical Data: The SMM did not consult the Aircraft Maintenance Manual (AMM) or the Illustrated Parts Catalog (IPC). Instead, he removed one of the old bolts from the removed windscreen and used it as a physical sample to find replacements.
3. Like-for-Like Replacement From an Unverified Sample: The SMM did not know the old windscreen had itself been fitted with incorrect (too short) bolts. Matching new bolts to that sample carried a latent defect forward, and the smaller-diameter bolts he chose still looked and felt close enough to pass.
4. Supervision and Organisational Oversight: The AAIB found that the SMM had not followed company policies and that local management had not monitored his working practices.
5. No Independent Inspection: Windscreen replacement was not treated as a task needing a duplicate inspection, so the SMM certified his own work. The AAIB recommended that such safety-critical tasks be tested or verified by another person.

Aloha Airlines Flight 243 (Boeing 737-200, 28 April 1988)

On 28 April 1988, Aloha Airlines Flight 243, a Boeing 737-297 (registration N73711) operating an inter-island flight from Hilo to Honolulu, experienced an explosive decompression while leveling off at Flight Level 240 (24,000 ft). An 18-foot section of the upper fuselage skin and structure (from just aft of the cockpit to the leading edge of the wing, encompassing the lap joint at stringer S-10L) detached completely. Chief Flight Attendant Clarabelle Lansing was swept out of the cabin and lost at sea. Despite catastrophic control cable and hydraulic system degradation, the flight crew executed an emergency landing at Kahului Airport, Maui.

Forensic Technical Breakdown

The US National Transportation Safety Board (NTSB AAR-89/03) determined the following technical findings:

• Multi-Site Damage (MSD): The failure was caused by the presence of multiple, microscopic fatigue cracks propagating simultaneously from adjacent countersunk rivet holes along the lower row of the S-10L lap joint. These individual micro-cracks eventually coalesced into a single critical longitudinal crack that ran catastrophically along the lap joint.
• Bonding Delamination: During manufacturing, Boeing utilized a cold-bond adhesive system to bond lap joint skin panels together. In the warm, humid, maritime environment of the Hawaiian islands, moisture ingress caused extensive delamination (disbonding) of the adhesive. Once the adhesive bond failed, all fuselage pressurization hoop stresses were transferred directly and entirely onto the mechanical fasteners.
• Extremely High Cycles: The airframe had accumulated 89,680 flight cycles (the second highest in the global Boeing 737 fleet) across only 35,496 flight hours, representing unusually frequent pressurization and depressurization cycles.

Latent Conditions & Human Factors Precursors

1. Vigilance Decrement in Visual Inspection: Inspecting thousands of rivet heads along fuselage lap joints is monotonous, and defects are rare. Vigilance decrement, the decline in detection performance during sustained searching for rare targets, typically sets in within about 30 minutes.
2. Night Scheduling: The NTSB noted that fuselage inspections were scheduled at night, which made an adequate inspection of the outer skin more difficult.
3. Limited AD Scope and NDI Training: Boeing Alert Service Bulletin 737-53A1039 recommended inspecting all lap joints, but Airworthiness Directive 87-21-08 mandated eddy-current inspection only at the stringer 4 location. After the accident, cracks were found at S-4R that the required eddy-current inspection should have detected, and records could not show whether it had been done properly. Aloha's training records showed little formal training in non-destructive inspection.
4. Management and Regulatory Oversight: The NTSB cited Aloha management's failure to supervise its maintenance force, the FAA's failure to evaluate the airline's maintenance programme and to widen the AD, and the lack of a terminating action for known problems with the early 737 cold-bonded lap joint.

Chalk's Ocean Airways Flight 101 (Grumman Turbo Mallard, 19 December 2005)

On 19 December 2005, Chalk's Ocean Airways Flight 101, an amphibious Grumman G-73T Turbo Mallard (registration N2969) built in 1947, departed the Miami seaplane base for Bimini, Bahamas. Shortly after takeoff, the right wing separated near the wing root. The aircraft plunged into the shipping channel, killing all 20 passengers and crew members.

Forensic Technical Breakdown

The NTSB investigation (NTSB/AAR-07/04) revealed an extreme case of structural degradation and organizational breakdown:

• Multiple-Element Fatigue Damage: The right wing separated at wing station 34 because of pre-existing fatigue fractures and cracks in the rear Z-stringer, the lower wing skin, and the rear spar lower cap. This damage reduced the wing's residual strength until it failed in normal flight.
• Ineffective Repairs: Recurring fuel leaks near the separation area were treated by removing and replacing fuel tank sealant, but the leaks kept returning. A doubler repair to the lower wing skin at wing station 34 did not restore the skin's load-carrying capability and did not address the underlying cause, a cracked or fractured rear Z-stringer.

Latent Conditions & Human Factors Precursors

1. Recurring Defects Not Evaluated: The repeated leaks were signs of structural damage, yet they were treated as isolated sealant problems. The NTSB noted that repair thresholds would help ensure repeated discrepancies get properly evaluated, and that structural damage found on another company aircraft should have prompted a fleet-wide wing inspection.
2. Limited Manufacturer and Engineering Support: The Turbo Mallard was an ageing type with little manufacturer support, and the NTSB found FAA oversight procedures for such operators insufficient.
3. Failure of Regulatory Oversight: The probable cause included the FAA's failure to detect and correct deficiencies in the company's maintenance programme.

Comparative Synthesis: Three Landmark Accidents

Exam Pitfalls / Common Traps

• Trap 1: The 'Bad Apple' Theory. Exam questions often test whether candidates attribute accidents solely to individual mechanic negligence. EASA Part-66 doctrine rejects this simplistic view: while the SMM on BA 5390 made the error, the AAIB also faulted the absence of a second-person check for a safety-critical task and the lack of local management monitoring of his working practices.
• Trap 2: Confusing BAC 1-11 bolt discrepancies. On BA 5390, the 84 bolts were too small in diameter (A211-8C instead of A211-8D, about 0.026 inches undersized), while the remaining 6 were too short (A211-7D, 0.1 inches short).
• Trap 3: Misunderstanding Multi-Site Damage (MSD). Aloha 243 was not caused by a single large fatigue crack that broke suddenly. It was caused by Multi-Site Damage (MSD)—hundreds of adjacent tiny cracks along multiple rivet holes that linked up simultaneously across stringer S-10L once the bonding adhesive failed.
• Trap 4: Treating recurring defects as unrelated. On Chalk's 101, repeated fuel leaks were fixed again and again with sealant. A defect that keeps returning is evidence of an unfixed root cause and must be evaluated, not simply re-sealed.