Atlantic Airlines ATP near Bournemouth on Oct 7th 2014, inexplicable roll and smoke

Last Update: May 14, 2015 / 14:51:09 GMT/Zulu time

Bookmark this article

Incident Facts

Date of incident
Oct 7, 2014

Classification
Report

Airline
Atlantic Airlines

Departure
Bournemouth, United Kingdom

Destination
Guernsey, United Kingdom

Aircraft Registration
G-BTPF

Aircraft Type
BRITISH AEROSPACE ATP

ICAO Type Designator
ATP

An Atlantic Airlines British Aerospace Bae ATP, registration G-BTPF performing freight flight NPT-22G from Bournemouth,EN to Guernsey,CI (UK) with 2 crew, had departed Bournemouth's runway 26 and climbed to FL070, the captain (31, ATPL, 2,570 hours total, 325 hours on type) was pilot flying, the autopilot had not been engaged with the captain manually flying the aircraft, when shortly after levelling off the control wheel forcefully turned left ripping it out of the hands of the captain. The captain applied opposite rudder to counteract the roll and applied significant force, estimated to be about 40-45 lbs, until the control wheel was turned 45 degrees to the right, at which point the aircraft began to roll level. When the control wheel was turned between -15 and +15 degrees of roll there was no response for about 3-4 seconds, then the situation became normal again and the aircraft reacted normally to aileron control inputs. Another 30 seconds later the control wheel was again ripped out of the hands of the captain, this time turning to the right, the captain again applied opposite rudder to counter act the roll and forcefully, force about 40-45lbs, turned the control wheel to the left, when after reaching 45 degrees left the aircraft began to roll level, another 3-4 seconds the control wheel handling returned to normal again. The captain decided to declare emergency and return to Bournemouth. While on approach the crew carefully applied flaps confirming the handling remained normal but noticed the fluorescent lights at the roof of the cockpit began to flicker. The aircraft landed safely on Bournemouth's runway 26 and taxied to the apron. About 10 mintues after the engines had been shut down the crew observed haze and an acrid electrical smell in the cockpit. A ground engineer, who had entered the cockpit, heard a bang from the lower forward fuselage and opened the hatch releasing acrid white smoke through the hatch.

The British AAIB released their bulletin on page 3 summarizing crew statements, that the flying conditions were not conducive to icing, there was no turbulence and no other aircraft were in the vicinity. SYS1 had been selected on the autopilot flight controller but no autopilot modes had been activated.

The AAIB complained that they were notified of the occurrence only on Oct 15th 2014 after maintenance had already replaced several components of the aircraft. The cockpit voice recorder had been overwritten by then, the flight data recorder storing 30 parameters over a period of 26 hours, contained valid data including the occurrence flight but due to the limited parameters (which do not include autopilot, control wheel positions or status of the electrical system), which made it impossible to confirm the movement of the control wheel.

The AAIB wrote: "The data recorded the aircraft levelling off at FL70 just over four minutes after takeoff. Twenty four seconds later, the aircraft rolled left to -4° and there followed a number fluctuations in the recorded aileron position and roll attitude. The aircraft altitude increased by approximately 80 ft after which a number of fluctuations in the normal acceleration were observed1. When comparing these acceleration fluctuations with the pitch and roll attitudes, it was considered unlikely that they were caused by the control surface deflections and more likely to be a period of turbulence. The maximum aileron deflection throughout this period was 3.8° (maximum travel is ± 15°) and the maximum roll attitude 5.4° to the left."

The AAIB described the maintenance actions: "Following the incident flight, the operator conducted a thorough examination of the control system and did not identify any faults or evidence of their having been a control restriction. The aileron disconnect release unit had not operated, the cable tension unit was serviceable and the cables were assessed as being of the correct tension. A thorough examination was carried out in Zone 130, where the loud bang was believed to have come from and where the smoke was discovered at the end of the flight. The engineers could not identify the source of the smoke, nor could they identify any damage to circuit breakers, relays and wiring in this area. As a precaution, the No 1 TRU and the No 2 Inverter, which are both located in this area, were replaced. Tests of the autopilot and the SCS was carried out in accordance with the AMM and the systems were assessed as serviceable. These tests included the operation of the autopilot, the autopilot disconnect, the aileron servomotor and the servomotor clutch. A rigging check was also carried out on both aileron synchro transmitters and they were found to be within limits. As a precaution the autopilot flight controller and both computer / amplifiers were replaced and the aircraft was released for further flight."

The aircraft entered service again and completed 20 flight hours in 27 cycles following the occurrence before entering scheduled maintenance, during which the aircraft was stripped down. The AAIB reported: "With the aircraft in a stripped condition, a further examination of the control system and autoflight system was carried out. There was no visual evidence of damage to any of the cables or pulleys, nor any evidence that a control restriction had occurred. The aileron release unit and the cable tension regulator were examined in accordance with the AMM and found to be serviceable. The fuselage cable tension was checked and the reading of 6.1 on the tension regulator was found to be lower than the target reading of 7.3 obtained from the cable tension graph in the AMM. The tension of the wing cable circuits were also checked and found to be slightly outside the target tension of 88 lbs. The actual cable tensions were established as: Left wing Right wing Up 80 lbs 96 lbs Down 88 lbs 92 lbs The aileron position synchros were visually examined and one wire, labelled UB39, on the left aileron synchro was found to be chafed down to the bare metal, Figure 3. This wire provides one of the three secondary outputs from the synchro. The terminals on the contacts on both synchros were found to be corroded."

The AAIB analysed however that the chafed wire and the corroded contacts could not explain the encounter.

Visual examination of the components removed from the aircraft did not find any anomaly.

The AAIB analysed that the investigation was unable to determine whether the uncommanded roll and the smoke were related or unrelated events.

The AAIB concluded the analysis: "In summary, the investigation undertaken by the operator, and the tests by the equipment overhaul agencies, could identify no mechanical or electrical fault that could have caused the reported uncommanded input or the smoke seen to emanate from the forward hatch. The aircraft has since flown without a reoccurrence of either fault."