History of the Flight
The aircraft, a Beech Super King Air B200 (registration G-BPPM), was conducting a scheduled public transport flight from Aberdeen to Humberside on 13 June 2001. During taxi, the flight director annunciator remained dim regardless of switch position, but was corrected after recycling the switch three times. A normal departure followed.
Shortly after takeoff, air traffic control reported that the landing gear had remained extended. The crew observed that despite the gear selector being in the UP position, three green indicator lights showed the gear was down-and-locked. Recycling the selector and checking the circuit breaker had no effect.
The commander handed control to the first officer and consulted the emergency checklist while the aircraft orbited offshore. Pulling and resetting the landing gear circuit breaker did not resolve the issue. While orbiting with the autopilot engaged, the red master warning and autopilot fail lights illuminated, followed by autopilot disengagement. About one minute later, the commander's attitude indicator and altimeter failed, accompanied by an electrical ozone smell. The instruments were restored after switching electrical supply to the No 2 inverter, though the No 1 inverter had not indicated a failure.
The commander returned to Aberdeen, where a fly-past at 100 feet confirmed the gear appeared locked-down. After a go-around, a normal circuit and approach were flown, with passengers briefed. The touchdown was normal, but during rollout, the gear slowly collapsed. The first officer shut down engines and feathered propellers. Passengers exited via rear doors; no injuries were sustained.
Examination of Aircraft at Aberdeen
The operator's chief engineer examined the aircraft on jacks at Aberdeen. The gear selector was found in the DOWN position. Landing gear doors were fully closed with little damage, and actuator supports were undamaged, indicating the gear had been driven to the retracted position rather than collapsing from failure to lock. Circuit breakers had not tripped, and the gear UP limit microswitch was activated. No obvious damage, foreign objects, or loose articles were found in the selector, motor/gearbox assembly, or wiring. The gear could be extended manually and then operated normally electrically.
Investigation focused on the safety (squat) switch, as similar components had caused problems on this type. Moving the switch could interrupt the retraction cycle; tapping it caused intermittent resumption. Internal soldered joints were extensively corroded from moisture ingress, with wires breaking away when handled. This condition was considered to fully explain the gear failing to retract after takeoff.
Electrical problems before and during the flight were investigated but not apparent afterward. The flight director dimming was attributed to a transient switch fault. Instrument failures were thought due to temporary frequency/voltage issues with the inverter output. These events were an apparent coincidence but presented a considerable distraction.
Further Examination
Maintenance replaced all landing gear control components, including squat switches, motor/gearbox assembly, drivetrain, selector assembly, and wiring between selector, motor/gearbox, and right main gear. Removed wires showed minor abrasions but no exposed conductors. The selector assembly and relay control box were examined internally without finding damage, bridged contacts, or foreign objects. Bench testing verified correct operation of relays and dynamic brake relay. The UP relay closed at 20 volts and de-energised at 5 volts; dynamic brake relay latched correctly. Resistance values were measured for UP and DOWN relays. The selector handle latch coil resistance was 112 ohms.
Possible Failure Conditions
Two possibilities for uncommanded gear retraction were considered. The first, spurious power to the UP motor winding, was discounted due to lack of arcing evidence. The second remained unclear; however, no definitive cause for the gear retraction during rollout was established beyond the safety switch condition.
Probable Cause
Not stated.