The Bolts That Were 0.026 Inches Too Small: What British Airways Flight 5390 Still Teaches About Matching Fasteners to Spec

On the morning of 10 June 1990, British Airways Flight 5390 left Birmingham for Málaga with 81 passengers and six crew aboard a BAC One-Eleven. Climbing through about 17,300 feet over Didcot, England, the captain's left windscreen panel left the airframe. Cabin pressure blasted the panel clear of the fuselage. Captain Tim Lancaster was pulled halfway through the opening and held there by the slipstream while cabin crew clung to his legs. First Officer Alastair Atchison flew the jet alone to a safe landing at Southampton. No one died. The injuries were serious enough: frostbite, fractures, shock.

The UK Air Accidents Investigation Branch (AAIB) final report (1/1992) did not leave the cause in doubt. The windscreen had been replaced about 27 hours earlier. Of the 90 bolts that held the panel to the frame, 84 were the wrong diameter. They were roughly 0.026 inches (0.66 mm) too small. Under the differential pressure of climb, those bolts simply pulled through the countersinks. The panel left the aircraft. So did most of the hardware.

That is not a story about exotic metallurgy. It is a story about a fastener that looked close enough, threaded into the hole, and was still the wrong part.

What the bolts actually were

The BAC One-Eleven captain's windscreen is an external-fit panel. Cabin pressure works to push it off the frame, not seat it. Every retaining bolt has to carry that load in tension and shear through the countersunk holes. Spec called for British Standards A211-8D bolts: correct diameter and correct length for the stack-up (#10-32 class under the Unified Thread Standard, about 0.8 inch long in this application).

What went on the airplane was different:

• 84 bolts were A211-8C: same general family, correct length, but #8-32 instead of #10-32. About 0.026 inch undersize on diameter.
• 6 bolts were A211-7D: correct diameter, but about 0.1 inch (2.5 mm) too short.

The previous windscreen had already been fitted with the shorter A211-7D hardware. The night-shift maintenance manager removed those bolts, read them as A211-7D, and set out to replace like with like. The stores did not have enough A211-7D stock. Working from worn labels, poor light, and without his glasses, he selected what looked like a match from a self-service carousel. He took 84 of the thinner A211-8C bolts, kept six of the old short ones that still looked serviceable, and installed the mix. The thread pitch was close enough that the screws started. An electric screwdriver clutch slip felt normal from the awkward angle on the lift. He signed the work. The job was not listed as a "vital point," so no second person had to verify the installation before the next passenger flight.

AAIB investigators recovered the panel near Cholsey with bolts still in some of the holes. Diameter and length told the whole story. Undersize shanks had never developed the bearing and thread engagement the countersinks needed. When cabin differential pressure climbed with altitude, the joint failed all at once.

Why "it threaded in" is not a specification

Anyone who has stood in front of a fastener bin has felt the pull of a near-match. Same head style. Same rough length. Same pitch feel when you start the first turn. On Flight 5390 that near-match was catastrophic because three separate properties all matter on a pressurized external windscreen:

1. Diameter (major and minor). Undersize bolts reduce shear area, reduce thread engagement area in the frame, and sit poorly in a countersink designed for a larger head and shank. A #8 in a #10 hole is not "almost right." It is a different load path.
2. Length and grip. A bolt that is 0.1 inch short may not put full thread into the nut plate or insert. Clamp load drops. Fatigue and pull-through risk rise.
3. Thread series and fit. Unified #8-32 and #10-32 share a pitch of 32 TPI, so the wrong screw can start in the right hole and still fail under load. Feeling the first few turns is not a gauge check.

The correct A211-8D choice was not optional decoration. It was the only combination of diameter and length the joint design assumed. Cabin differential at climb is a known, calculable load. The fastener stack had to meet that load with margin. Visual matching from a dim carousel does not meet that standard.

The right fastener for the job

Match the fastener to the load, the environment, and the application, then prove it with the paperwork, not with memory.

• Read the callout. Part number, standard (here, British Standards A211 series), diameter, length, material, and finish. If the last install used the wrong part, copying it repeats the error. Flight 5390's previous windscreen had already been on short bolts. "What came off" was not the drawing.
• Verify diameter and length with a gauge or micrometer, not by eye. A 0.026-inch difference is easy to miss under fluorescent light and fatigue. It is obvious under a go/no-go check.
• Confirm thread engagement. Correct length puts full, specified thread into the insert. Correct diameter fills the countersink and the hole as designed.
• Treat pressure-retaining external panels as critical joints. Second-person verification, torque or installation check where required, and no self-cert of the same work by the same person who installed it. AAIB made those process points explicit after this accident.
• Stock discipline. Minimum stock levels, clear labels, and no "close enough" substitutions from adjacent bins. The store supervisor on the night had already flagged that A211-8D was the normal windscreen bolt. That signal was ignored.

None of this is aircraft-only advice. Fleet shops, plant maintenance, and field service run the same risk every time a bin is unlabeled, a drawing is skipped, or a "temporary" bolt stays in service. Cabin pressure is dramatic. A hydraulic flange, a steering shaft pinch bolt, or a suspension mount fails for the same reason: the hardware never matched the load the joint was designed to carry.

What still holds

Flight 5390 is remembered for the image of a captain half outside a jet at altitude and a first officer who brought everyone home. The fastener lesson is quieter and more useful. Eighty-four bolts failed because they were 0.026 inches too small in diameter. Six more were too short. The joint looked assembled. It was not built to spec.

Right fastener for the job means the right diameter, the right length, the right thread, and the right verification. Close enough is not a grade. It is a failure mode waiting for the next climb.

Sources

• AAIB Aircraft Accident Report 1/1992, BAC One-Eleven G-BJRT, 10 June 1990
• Aviation Safety Network summary of G-BJRT / BA5390
• British Airways Flight 5390 overview (with AAIB-sourced bolt dimensions)