West Sweden ATP at Guernsey on Feb 14th 2018, frozen flight and engine controls

Last Update: December 6, 2018 / 17:47:41 GMT/Zulu time

Bookmark this article

Photo of West Air Sweden SE-MHE, BRITISH AEROSPACE ATP — West Air Sweden SE-MHE, BRITISH AEROSPACE ATP (Photo credit: Jonathan Winton Photography / Flickr / License: CC by-nd)

Incident Facts

Date of incident
Feb 14, 2018

Classification
Report

Airline
West Air Sweden

Departure
East Midlands, United Kingdom

Destination
Guernsey, United Kingdom

Aircraft Registration
SE-MHE

Aircraft Type
BRITISH AEROSPACE ATP

ICAO Type Designator
ATP

A West Air Sweden British Aerospace BAe ATP, registration SE-MHE performing freight flight PT-22B from East Midlands,EN to Guernsey,CI (UK) with 2 crew, was descending towards Guernsey when the aircraft entered a left slip as normal, the pilot flying attempted to correct the slip by rudder trim, however, the slip indication persisted, the trim caution light illuminated and the autopilot disconnected. The aircraft rolled left 45 degrees and pitched down, the commander took control of the aircraft and returned it to a stable flight path. The commander however noticed that rudder control was "stiff" though elevator control was normal. He tried to apply rudder and aileron trim, but they were "stiff" too. The crew declared Mayday reporting control problems. The commander attempted to apply power but found the thrust levers wouldn't move, the crew upgraded the Mayday reporting they also weren't able to increase power. ATC cleared them to 3000 feet and offered runway 27 at Guernsey as well as Alderney and Cherbourgh as alternates, the crew elected to go for Guernsey. A short time later the commander radioed that everything had returned to normal. The crew continued for a safe landing without further incident.

The United Kingdom's AAIB released their bulletin concluding the probable causes of the serious incident were:

Shortly after top of descent, the aircraft developed a significantly out-of-trim condition which, when the autopilot automatically disconnected, caused a rapid and significant roll to the left accompanied by a nose-down pitch. The evidence supported the likelihood that the crew used aileron trim instead of rudder trim to balance the aircraft in yaw, leading to the autopilot disconnecting automatically and the initial in-flight upset.

The commander corrected the attitude changes with difficulty but found that both power levers were stuck at flight idle. Approximately three minutes later, he successfully advanced the power levers and the engines responded accordingly. The commander considered that the most likely scenario was icing of the engine controls but this was unlikely. Although the cause could not be established definitively, it is possible that the left power lever was restricted because of wear in the roll-over lever locking mechanism, although this would not explain the jamming of the right power lever reported by the pilots.

The AAIB reported that a captain under training was occupying the left hand seat, the training captain occupied the right hand seat. The flight was uneventful until top of descent into Guernsey, when the training captain handed the controls to the captain under training, who now assumed the role as pilot flying.

The AAIB analysed:

The flight progressed normally and without incident until the commencement of the descent into Guernsey. The power was reduced to 55% torque on both engines and, as normal, the slip indicator ball migrated to the left. The PF reported that he began to retrim the aircraft in yaw and he commented about the amount of trim that he was applying because it was having no effect. The FDR data showed no change in rudder position which indicated that the rudder trim was not being used. The data showed the aircraft rolling slowly to the left and, approximately 12 seconds later, the commander said “you will disconnect it now”, shortly followed by “be careful be careful be careful”. The crew recalled that the trim warning light was illuminated, which indicated that the aircraft was out-of-trim in pitch or roll.

Almost immediately after the commander expressed his words of caution, the autopilot disconnected and the CVR recorded a continuous cavalry charge warning. The aircraft rolled left and pitched nose-down before the commander took control and restored a safe attitude with normal use of the flying controls. The continuous cavalry charge warning indicated that the autopilot disconnected automatically, and the movement of the flying controls indicated that they were not frozen.

No faults were found when the autopilot was tested, and the FDR data showed that the aircraft was operating inside the limits of the safety circuit when the autopilot disconnected. It is possible that the autopilot disconnected because the aircraft was being trimmed in roll instead of yaw. In this case, the current demanded by the aileron servomotor would increase as it tried to resist the effect of the increasing trim tab deflection, and the trim warning light would illuminate when the appropriate threshold was exceeded. Further application of aileron trim would cause the autopilot to automatically disconnect and trigger a continuous ‘cavalry charge’ alert. With the autopilot disconnected, the ailerons would immediately deflect causing a rapid roll to the left. This scenario is consistent with the data and the way the autopilot operates, but both pilots believed that they had operated the rudder trim wheel correctly.

Possible action on the incorrect trim control

The following section considers aspects of human performance which might have influenced events had the aileron trim been used instead of the rudder trim. The reasons why the incorrect trim control might have been used, and the reasons why the situation might not have been corrected before the autopilot disconnected, might be explained by considering the performance shaping factors present during the event.

The rudder and aileron trims are in the same part of the flight deck and are separated by only a small distance. They are operated in the same way, and have the same shape, size and texture. They are, however, orientated differently, being set in different planes. The similarities in these design performance factors, perhaps outweighing the orientation factor, might have increased the probability that an incorrect control would be selected.

The PF had only 12 hours flying the aircraft type out of his total 4,100 flying hours, and his workload while undergoing line training in an unfamiliar aircraft was likely to have been high. Although by no means a novice pilot, he could be considered a novice on type and, therefore, more vulnerable in high workload situations to operating an incorrect control. After a relatively short period on type, he might not yet have developed the automatic task performance skills which would normally be relied upon to avoid mistakes in high workload circumstances.

Successful recovery from the inadvertent selection of an incorrect control requires system feedback to the operator. The pilot thought he was using the rudder trim but was not seeing the appropriate response from the rudder trim indicator. This feedback would have been accurate but confusing. He might even have been encouraged to continue turning the trim wheel, as opposed to stop turning it, by the feedback from pilot in the right seat saying: “you need more [trim]”.

Had the pilot been using aileron trim with the autopilot disengaged, the aircraft would have rolled to the left. The pilot would have noticed this feedback and probably realised immediately that he was using the incorrect trim. However, the autopilot was controlling the aircraft and would have been resisting any tendency to roll. The pilot would therefore have been denied this feedback indicating that he was using the incorrect trim wheel.

The commander, as the supervising pilot, would probably have had an elevated readiness to receive system feedback because of the nature of that role, and it appeared from the CVR and his recollection of events that he received enough feedback to understand that the autopilot would soon disconnect. However, he was unable to prevent it from happening in the 15 seconds between the aircraft beginning to roll and the autopilot disconnecting.

With respect to the power levers the AAIB analysed:

The crew reduced engine power prior to the descent and both power levers were retarded to flight idle after the autopilot disconnected. In both cases, the levers moved without restriction. This indicated that the power levers were not frozen or restricted.

The commander reported that the power levers were stuck at flight idle after he regained control of the aircraft and he believed them to have frozen. He stated that he attempted to control the engines by advancing the power levers, which would not move, and by using the standby control system, but neither method was effective. He briefly pulled the roll-over levers into the beta range and the engines responded with speed and torque increasing accordingly.

The roll-over levers use the same control runs as the engine power levers, so the fact that the commander successfully entered the beta range indicated that the control runs were not restricted or frozen. He deselected the beta range and succeeded in advancing the power levers approximately three minutes after retarding them to flight idle.

Examination of the aircraft in Guernsey identified no explanation for the power lever restriction. When the aircraft was subsequently examined at East Midlands Airport, the left roll-over lever had excessive wear and therefore did not always lock down when it was returned to the flight idle position. The effect of this was that the left power lever could not be advanced from its flight idle position. Had the left roll-over lever been unlocked during this incident, it might have felt like both power levers were physically restricted when the pilot tried to advance them together (even though only the left would have been). However, in these circumstances the right power lever should have been free to move had it been operated independently. The commander stated that he was unable to move the levers by hand or control the engine by using the standby control system. The investigation could not explain why the right power lever might have been jammed.